J01 1 <#FLOB:J01\>5.3 THE EXPERIMENTAL APPROACH TO NUCLEAR J01 2 REACTIONS

J01 3 Having defined the quantities that are normally measured in a J01 4 nuclear reaction we here outline the typical experimental J01 5 procedures which are followed for studying the symbolic reaction J01 6 A(a, b)B. No details are given of the apparatus other than to J01 7 mention very briefly the underlying physical principles. Details of J01 8 low energy nuclear physics apparatus are given, for example, in J01 9 Burcham (1988) and of high energy elementary particle apparatus in J01 10 a book in this series by Kenyon (1988).

J01 11 Referring to Fig. 5.3, charged ions of the particle a are J01 12 produced in some form of accelerator (described later in this J01 13 section) and, by use of bending magnets for example, will emerge J01 14 with a particular energy. These ions then pass through a collimator J01 15 in order to define their direction with some precision and strike a J01 16 target containing the nuclei A. As the beam particles move through J01 17 the target they will mainly lose energy by ionizing target atoms J01 18 and so, if precise energy measurements are to be made, a thin J01 19 target must be used. This, however, increases the difficulty of the J01 20 experiment since few interactions will take place. Choice of target J01 21 thickness is clearly a crucial decision in planning an J01 22 experiment.

J01 23 diagram&caption

J01 24 The reaction product particles b move off in all directions and J01 25 their angular distribution can be studied by detecting them after J01 26 passage through another collimator set at a particular angle J01 27 <*_>theta<*/>. Various types of detector are used (discussed later) J01 28 - sometimes in combination - and these can determine the type of J01 29 particle as well as its energy. But experimenters have to contend J01 30 with many complications of interpretation, impurities in targets J01 31 and, not least, the stability of their apparatus. In the end, J01 32 detailed information becomes available about <*_>sigma<*/>, J01 33 d<*_>sigma<*/>/d<*_>OMEGA<*/> and their energy dependence for the J01 34 reaction under study.

J01 35 5.3.1 Accelerators

J01 36 Most important for nuclear reaction studies are Van de Graaff J01 37 accelerators in which ions are accelerated in an evacuated tube by J01 38 an electrostatic field maintained between a high voltage terminal J01 39 and an earth terminal, charge being conveyed to the high voltage J01 40 terminal by a rotating belt or chain. In early forms of this J01 41 accelerator, positive ions from a gaseous discharge tube were J01 42 accelerated from the high voltage terminal to earth. But, in modern J01 43 'tandem' accelerators, negative ions are accelerated from earth to J01 44 the high voltage terminal where they are then stripped of some J01 45 electrons and the resultant positive ions are further accelerated J01 46 down to earth potential. The effective accelerating potential is J01 47 thus twice the potential difference in the machine. High flux J01 48 proton beams with energies up to around 30MeV can be produced in J01 49 this way. The machines can also be used to accelerate heavy ions J01 50 such as 16O.

J01 51 At higher energies use is generally made of orbital J01 52 accelerators in which charged particles are confined to move in J01 53 circular orbits by a magnetic field. At non-relativistic energies J01 54 the angular frequency of rotation <*_>omega<*/>, known as the J01 55 cyclotron frequency, is constant depending only on the J01 56 strength of the field. In a cyclotron, the particles rotate in J01 57 a circular metallic box split into two halves, known as Ds, between J01 58 which an oscillating electric field is maintained. Its frequency J01 59 matches <*_>omega<*/> and so the particle is continually J01 60 accelerated. In a fixed magnetic field the orbital radius increases J01 61 as the energy increases and, at some maximum radius, the particles J01 62 are extracted using an electrostatic deflecting field. However, as J01 63 the energy becomes relativistic (remember formula), J01 64 <*_>omega<*/> decreases with energy and it becomes necessary to J01 65 decrease steadily the frequency of the oscillating electric field J01 66 with energy to preserve synchronization.

J01 67 Such a machine is known as a synchrocyclotron and protons J01 68 with energies in the region of 100 MeV have been produced in this J01 69 way.

J01 70 For energies higher than this gigantic magnets would be needed J01 71 and so the approach is to accelerate bunches of particles in orbits J01 72 of essentially constant radius using annular magnets producing J01 73 magnetic fields which increase as the particle energy increases: J01 74 This energy increase is provided by passing the particles through J01 75 radio-frequency cavities whose frequency also changes J01 76 slightly as the particles are accelerated to ensure J01 77 synchronization. Such devices are called synchrotrons and can J01 78 be physically very large. For example, the so-called Super Proton J01 79 Synchrotron (SPS) at CERN (Geneva) has a circumference around 6 km J01 80 and can produce protons with energies up to around 450 GeV. LEP J01 81 (the Large Electron-Positron Collider) has a circumference of 27 km J01 82 and accelerates electrons (and positrons in the opposite direction) J01 83 to energies of <*_>approximate-sign<*/>60 GeV or more. Finally, the J01 84 Superconducting Super Collider (SCC), which uses superconducting J01 85 magnets, and which is being built in the USA, has a circumference J01 86 of 87 km and will produce proton and antiproton beams with energies J01 87 <*_>approximate-sign<*/>20 000 GeV!

J01 88 Electrons can also be accelerated in synchrotrons but, because J01 89 of their small mass, large amounts of energy are radiated J01 90 (synchrotron radiation) owing to the circular acceleration. At J01 91 energies beyond a few GeV this loss becomes prohibitive and use has J01 92 to be made of linear accelerators in which electrons are J01 93 accelerated down a long evacuated tube by a travelling J01 94 electromagnetic wave. The Stanford Linear Accelerator (SLAC) in the J01 95 USA, for example, is around 3 km long and can produce pulses of J01 96 electrons with energies up to 50 GeV.

J01 97 5.3.2 Detectors

J01 98 Although in the early days much use was made of ionization J01 99 chambers, for example the Geiger counter (section 1.4), the J01 100 detectors currently in use for nuclear physics experiments are J01 101 usually either scintillation counters or J01 102 semiconductor detectors or some combination. The former J01 103 are developments of the approach of Rutherford, Geiger and Marsden J01 104 (section 1.3) using the scintillations produced in a ZnS screen to J01 105 detect <*_>alpha<*/>-particles. Various scintillators are in J01 106 current use such as NaI activated by an impurity (usually thallium J01 107 for detection of <*_>gamma<*/>-particles), or some organic material J01 108 dissolved in a transparent plastic or liquid. The scintillations J01 109 are detected by a photomultiplier tube producing a pulse of J01 110 photo-electrons. The size of the pulse - the pulse height - J01 111 gives a measure of the energy of the incident particle.

J01 112 Semiconductor detectors depend on an incident particle or J01 113 photon exciting an electron from the valence band to the conduction J01 114 band. The resultant increase in conductivity - a conduction pulse - J01 115 then produces a signal which is processed electronically and which J01 116 enables the energy of the incident radiation to be measured.

J01 117 In the field of very high energy physics, considerable use is J01 118 made of bubble chambers and wire chambers. The J01 119 former follows on from the Wilson cloud chamber and consists J01 120 essentially of a large chamber, possibly several metres in J01 121 diameter, containing liquid (e.g. hydrogen, helium, propane, ...) J01 122 near its boiling point. The chamber is expanded as charged J01 123 particles pass through it, leading to the formation of bubbles, as J01 124 a result of boiling, along the particle tracks which can be stereo J01 125 flash photographed. The lengths of the tracks and their curvature J01 126 in a magnetic field enable particle lifetimes, masses and energies J01 127 to be deduced.

J01 128 Wire chambers consist of stacks of positively and negatively J01 129 charged wire grids in a low pressure gas. An incident charged J01 130 particle ionizes the gas and acceleration of the resultant J01 131 electrons near the anode wires leads to further ionization and an J01 132 electrical pulse. The physical location of the pulse can be J01 133 determined electronically so that track measurements can be made. J01 134 Using an applied magnetic field to bend the tracks again enables J01 135 information to be obtained about the properties of the detected J01 136 particle.

J01 137 5.4 NUCLEAR REACTION PROCESSES

J01 138 In the previous chapter some understanding of nuclear structure J01 139 has been achieved in terms of a nuclear model in which nucleons J01 140 move around fairly independently in a potential well. To give some J01 141 intuitive understanding of nuclear reaction processes we stay with J01 142 this description of the nucleus and follow a very illuminating J01 143 discussion given by Weisskopf (1957).

J01 144 An incident particle, a, approaching a nucleus, A, will, if it J01 145 is charged, first experience the long-range Coulomb potential and J01 146 if its energy is low will be elastically scattered by this before J01 147 coming within the range of the nuclear force. In this case it J01 148 undergoes Rutherford or Coulomb scattering as J01 149 described in section 1.3. For higher energies, or if the particle J01 150 is uncharged, it will come within the range of the nuclear force J01 151 due to the nucleons. This can be represented by a potential energy J01 152 curve of the form already discussed and shown in Fig. 4.2 for an J01 153 incident neutron or proton. As a result of this interaction the J01 154 incident particle may again be elastically scattered without J01 155 colliding directly with a nucleon in the nucleus. The form of this J01 156 scattering will obviously depend on the shape and size of the J01 157 nucleus and its associated potential well, and is referred to as J01 158 shape elastic scattering. All of these processes are J01 159 symbolized by

J01 160 formula

J01 161 If a direct collision with a nucleon takes place then there are J01 162 various possibilities. The nucleon may be excited to a higher J01 163 (unoccupied) state and the incident particle leaves the nucleus J01 164 with reduced energy. This in an inelastic scattering J01 165 process and the nucleus is left in an excited state. Another J01 166 variant of this is that, instead of exciting a nucleon, the J01 167 incident particle excites a collective mode - a vibrational or a J01 168 rotational state (section 4.3). Such processes are symbolized by

J01 169 formula

J01 170 where A* signifies an excited state of the nucleus A.

J01 171 Alternatively the incident particle may give enough energy to J01 172 the nucleon with which it collides so that this nucleon, b, is J01 173 knocked out from the nucleus (Fig. 5.4a). There are two J01 174 possibilities here, depending on how much energy is lost by the J01 175 incident particle. If the incident particle, a, retains enough J01 176 energy to escape from the nucleus after the collision we have the J01 177 process

J01 178 formula

J01 179 where B is the residual nucleus remaining after the nucleon b J01 180 has been knocked out of A. However, the incident particle may lose J01 181 so much energy that it is captured, resulting in the formation of a J01 182 nucleus B', i.e.

J01 183 formula

J01 184 Reactions of the various kinds just discussed are referred to J01 185 as direct reactions since there is direct interaction J01 186 with a single nucleon rather than with the nucleus as a whole. J01 187 Other variants illustrated in Figs 5.4b and 5.4c are stripping J01 188 and pick-up reactions. In the former a composite incident J01 189 particle, usually a deuteron (2H), is stripped of one of J01 190 its component nucleons which remains in the target nucleus and the J01 191 remaining nucleon(s) escape. Conversely, in the latter, the J01 192 incident particle, usually a nucleon, picks up another nucleon from J01 193 the target nucleus and carries it away, emerging as a deuteron.

J01 194 diagram&caption

J01 195 The next possibility is that the incident particle collides J01 196 with a nucleon in the target nucleus, perhaps one lying in a very J01 197 low shell model level, and neither has sufficient energy to escape. J01 198 There will then be a series of further random collisions in the J01 199 nucleus (Fig. 5.5) until eventually enough energy is concentrated J01 200 by chance on one particle to enable it to escape; or the nucleus J01 201 may lose its energy by emitting electromagnetic radiation. This J01 202 state of the nucleus after it has captured the incident particle J01 203 and in which many internal collision processes are occurring was J01 204 first discussed by Niels Bohr in 1936 and is referred to as the J01 205 compound nucleus. Whereas a direct reaction, for an J01 206 incident particle of several MeV, takes place in a time of the J01 207 order of that taken by a nucleon to cross a nucleus J01 208 (formula) the compound nucleus exists for a much longer J01 209 period and we shall see in section 5.8 that it can exist for times J01 210 in the approximate region from 10-14s to J01 211 10-20s. A compound nucleus process can thus be J01 212 represented as taking place in two stages - formation of the J01 213 compound nucleus and, after a considerable time, its decay. J01 214 Symbolically,

J01 215 formula

J01 216 where C* represents the excited compound nucleus. Because of J01 217 the long life of the compound nucleus little information about its J01 218 mode of formation is carried forward to influence the way in which J01 219 it disintegrates.

J01 220 diagram&caption

J01 221 Finally, one other approach to coping with the complexities of J01 222 nuclear reaction processes should be mentioned, namely, the J01 223 optical model which was developed by Feshback, Porter and J01 224 Weisskopf in 1954 and which is useful in giving a broad J01 225 understanding of nuclear reactions. J01 226 J02 1 <#FLOB:J02\>Logs of wood may have become partly compressed prior to J02 2 silicification (or calcification). Wood tends to be compressed J02 3 parallel to the cell files. Autochthonous roots, upright stems and J02 4 trunks are generally preserved as impressions of the bark, and J02 5 compressions of what tissues remained (mainly xylem) after decay J02 6 and infilling of the space by sand and mud. In a lycopod (Figs 4.7, J02 7 4.11), the woody outer cortex becomes compressed, and the small J02 8 woody stele is often displaced. Particularly difficult to unravel J02 9 in fossil plants are the different appearances that can occur J02 10 between different modes of preservation of the same species.

J02 11 Most aerial parts of fossil plants are found as fragmented and J02 12 dispersed remains. How can these be reassembled as a reconstruction J02 13 of the original plant? Palaeobotany is much concerned with this J02 14 detective work, and it is important to realize, in this respect, J02 15 the significance of any material one comes across in the field. J02 16 There are three methods of reconstruction:

J02 17 1. Locating material where, by good chance, two or more parts J02 18 (organs) are joined, e.g. root and stem, cone with pollen, and J02 19 dispersed pollen, leaves and stem, leaves and reproductive organs. J02 20 This is the only positive method, and it depends mainly on field J02 21 observation.

J02 22 2. Identifying particular anatomical features that are common J02 23 to two or more organs. Frequently-cited characters are particular J02 24 glands, type of stomata, cell outline: features that cannot readily J02 25 be identified in the field.

J02 26 3. Commonly occurring associations of organs (e.g. on bedding J02 27 surfaces) that would be difficult to explain if not from the same J02 28 plant.

J02 29 4.2.2 Coal and oil shales

J02 30 Coal is described in terms of rank (thermal grade - Fig. 4.12) J02 31 and type (composition). The end members (anthracite and peat) can J02 32 be fairly readily identified in the field. Very bright, splintery J02 33 coal is likely to be of high rank (low volatile bituminous J02 34 coal-to-anthracite). The tectonic - sedimentary setting of the J02 35 site, and the grade of the associated mudrocks, will provide a clue J02 36 to rank. Is there a root bed present? Is the coal autochthonous or J02 37 drifted? How 'pure' is the coal? It may be no more than a J02 38 carbonaceous mudstone. Try to identify the components of a J02 39 bituminous coal:

J02 40 Vitrain, representing coalified logs or pieces of bark, is J02 41 glassy (vitreous) and usually closely jointed.

J02 42 Fusain generally forms lenses and has a silky sheen, but J02 43 is soft and leaves a black mark.

J02 44 Durain is dull and tough, with megaspores often evident in J02 45 Palaeozoic coals.

J02 46 Clarain (attrital coal) is finely-laminated vitrain and J02 47 durain.

J02 48 This classification of coal is tedious to follow for logging. J02 49 Try using mm, cm or dm thicknesses of vitrain, fusain or attrital J02 50 coal. Further detail can be introduced by making estimates of J02 51 vitrain:attrital ratio. What one is after is an indication of the J02 52 overall vitrain:fusain ratio, and the percentage of mud present.

J02 53 figure&caption

J02 54 Oil shales and cannel coals have formed in aquatic anaerobic J02 55 environments. Cannel coal (dull with conchoidal fracture) J02 56 represents drifted, finely-divided terrestrial plant material. It J02 57 is often found at the top of a coal, following a flooding event. J02 58 Torbanite, lamosite, tasmanite and marinite are oil shales J02 59 distinguished by their principal components (identifiable J02 60 microscopically): torbanite, Botryococcus and allied algae; J02 61 lamosite, planktonic algae (both are lacustrine); tasmanite with J02 62 Tasmanites (marine); and marinite with various planktonic J02 63 marine algae (Hutton, 1986).

J02 64 4.2.3 Calcareous algae and stromatolites

J02 65 The undulose, arched or planar laminations of stromatolites J02 66 (Fig. 4.13) are generally readily recognized in the field. The J02 67 laminations reflect discontinuities in formation (e.g. storm J02 68 action) rather than short (e.g. diurnal) periodicities. Nodular J02 69 stromatolites (oncolites) are distinguished by their irregular J02 70 lamination around a shell or lithic clast. The clotted texture of J02 71 thrombolites is not easy to prove in the field.

J02 72 Growth form is probably the best indicator of red, calcified J02 73 algae with forms corresponding to the shapes of most types of J02 74 breakfast cereal. It may just be possible to make out, with a hand J02 75 lens, the cellular structure of a solenopore, or the larger, J02 76 reproductive cells of more advanced rhodophytes. Rhodophytes will J02 77 probably have an irregular, knobbly outline attributable to J02 78 truncation and branching.

J02 79 Although the stem of the green alga Chara sensu lato J02 80 may become encrusted, it is the calcified walls of the cells J02 81 enclosing the oogonia, the gyrogenites, (often replaced by silica, J02 82 or seen as moulds) that are so useful as environmental indicators J02 83 (essentially fresh water). Use a hand lens to identify the spiral J02 84 cells. Marine, calcareous (aragonite) green algae such as J02 85 Halimeda readily breaks up into segments, as does J02 86 Corallina (articulated, red alga) (see also J02 87 <*_>section<*/>3.6.1).

J02 88 4.3 Animal fossils

J02 89 4.3.1 Soft tissue preservation

J02 90 Soft tissue preservation is rare (see also J02 91 <*_>section<*/>6.3). In the field consider what types of sediment J02 92 may yield information on such material. There are five modes of J02 93 preservation: (1) mineral coats, where tissues are outlined by a J02 94 mineral, typically pyrite; (2) permineralization, actual J02 95 replacement of the tissue by phosphate (best), pyrite or silica (as J02 96 with plant cells, above); (3) casts or impressions of the tissues, J02 97 where the adjacent sediment was stabilized prior to decay and J02 98 infilling of the void: or, for example, by concretion formation in J02 99 the Silesian Mazon Creek Formation, but probably by organic J02 100 stabilization in the late Precambrian Pound Quartzite (Fig. 4.14); J02 101 (4) encasement in decay-inhibiting material; for example, scorpions J02 102 in Carboniferous coal (peat), insects in amber; and (5) J02 103 encrustation of soft tissue (e.g. hydroids, algae) by skeletal J02 104 organisms, to leave a mould of the soft tissue (bio-immuration).

J02 105 Evidence of bioturbation, scavenging and early reworking almost J02 106 always excludes soft tissue preservation of types 1-4 (except where J02 107 concretion initiation was so early as to inhibit these factors J02 108 locally). Although transport per se prior to deposition J02 109 does not appear to be a major factor, aerobic decay during J02 110 transport can be important. Following deposition, it is early J02 111 diagenesis that is important since anaerobic decay may be complete J02 112 (Fig. 4.15).

J02 113 Note that:

J02 114 1. In freshwater sediments, pyritization of soft tissues is J02 115 relatively unimportant due to relatively low levels of sulphate in J02 116 the water. In marine sediments with sufficient iron, pyritization J02 117 is favoured by high burial rate. With rapid burial more reactive J02 118 (less decayed) organic matter is added to the sediment, leading to J02 119 a rapid uptake of sulphate ions from pore water and therefore J02 120 increased diffusion of sulphate ions from the overlying sea water. J02 121 In euxinic conditions organic carbon is generally too high to J02 122 favour good pyritization and pyrite is disseminated.

J02 123 figure&caption

J02 124 2. Phosphatization of soft tissue is generally associated with J02 125 low burial rate (but without reworking) and high organic input. J02 126 Phosphatic materials (bone, arthropod carapace) often act as J02 127 nuclei, but the best phosphatization seems to have taken place, or J02 128 at least to have been initiated, at the sediment - water interface J02 129 (Martill, 1988; see also <*_>section<*/>4.5).

J02 130 3. Early carbonate encasement, where calcium carbonate was J02 131 precipitated as concretions, is favoured by high organic input and J02 132 burial rate. Where terrigenous input is low (marine or freshwater), J02 133 rapid carbonate precipitation is favoured by high carbon dioxide J02 134 production by algal and cyanobacterial mats e.g. the bedded J02 135 limestones of Solnhofen (Upper Jurassic, Germany) and the Green J02 136 River Formation (Eocene, USA) where fossils always seem to be J02 137 compressed.

J02 138 4.3.2 Univalves and groups with skeletons that remain more J02 139 or less intact

J02 140 Graptolites (V - revised) As with ammonoids (below), the J02 141 mode of preservation of graptolites in a sequence of mudrocks may J02 142 change appreciably from bed to bed, indicating small variations in J02 143 facies. Graptolites had a pliable organic (collagenous) skeleton. J02 144 Three-dimensional preservation in early-formed concretions J02 145 and certain limestones is uncommon. Material can be collected for J02 146 later acid-treatment. The most common mode of preservation is as J02 147 carbonized side-down (profile) compressions, or compressions where J02 148 the organic skeleton has been replaced by white-weathering, J02 149 phosphatic material (Fig. 4.16). Diplograptids (Fig. 4.17), J02 150 however, commonly landed on the sediment with one series of thecae J02 151 being entombed aperture down (scalari-form view). If the J02 152 form of the thecae is clear, identification is generally possible. J02 153 Locate material with discrete rather than crowded specimens and J02 154 with the proximal morphology distinct. Take care not to J02 155 misinterpret apparent branchings for Cyrtograptus or J02 156 Nemagraptus. Commonly, pyrite filled the thecae at an early J02 157 stage. Only internal moulds may remain but these have the advantage J02 158 that they are undeformed.

J02 159 table&caption

J02 160 Where mudrocks contain distal turbidites (striped mudstones), J02 161 the graptolites therein tend to be richer and better preserved than J02 162 in more oxidized sediment. Preservation was enhanced by rapid J02 163 sedimentation.

J02 164 Conularids (F) are generally highly compressed in J02 165 mudrocks. External and internal impressions occur in sandstones.

J02 166 Archaeocyathids (E - revised; Fig. 4.18) Skeleton J02 167 generally with morphology distinct. The calcareous skeleton is now J02 168 usually of calcite or has been silicified, but may have originally J02 169 been Mg calcite. Geopetal sediment and sparry fill to chambers is J02 170 common. Check for epitaxial algae (Epiphyton) as dark micritic J02 171 clots on autochthonous material. Mouldic preservation should show J02 172 evidence of regularly porous walls.

J02 173 Sponges (E) It is exceptional for a fossil sponge to be J02 174 found relatively perfect. The best preserved are those with fused J02 175 mineralized spicules giving a rigid skeleton. (Peripheral, loosely J02 176 bound spicules will still be missing.) Judging by the abundance of J02 177 loose spicules in many sediments (e.g. the needle-like spicules in J02 178 many Jurassic and Cretaceous rocks, or the ball-bearing-like dermal J02 179 spicules (Rhaxella) in the Upper Jurassic) sponges were much J02 180 more widespread than is now apparent. Opaline spicules of J02 181 demosponges and hexactinellids are readily dissolved. Look out for J02 182 moulds (in sand and mudrocks), or carbonate replacement in J02 183 limestones. Sclerosponges have a basal calcareous skeleton with J02 184 embedded siliceous spicules. In fossil representatives (e.g. J02 185 Chaetetes) the skeleton may now be entirely calcitic.

J02 186 Rugose and tabulate corals (F - revised) In J02 187 sandstones, preservation is nearly always poor. If the skeleton is J02 188 present then it will be almost certainly replaced. Sediment may J02 189 have infilled the calyx and infiltrated more deeply, giving J02 190 variable perfection to moulds. Preservation in mudrocks can be good J02 191 especially if they are somewhat bituminous. In wackestones and lime J02 192 mudstones, compaction is common. Select material suitable for J02 193 transverse and longitudinal sectioning. The manner in which J02 194 colonial forms increase is important.

J02 195 Scleractinian corals (F - Fig 4.19) Due to the J02 196 originally aragonitic skeleton, mouldic or replacement modes of J02 197 preservation are widespread. Neomorphic preservation J02 198 (calcitization) is mostly associated with massive colonies (because J02 199 they were less permeable). Colonial and solitary forms can be J02 200 completely dissolved and the resulting cavity spar-filled. J02 201 Preservation is generally poor in reefs, becoming better laterally J02 202 in muddier facies. For full identification it is essential to have J02 203 good preservation of the skeleton for transverse and longitudinal J02 204 sectioning. Distinct banding (probably emphasizing original J02 205 seasonal growth variation) is often evident.

J02 206 Bryozoans (G and G - revised) The stony bryozoans and the J02 207 Palaeozoic had a calcitic skeleton and are like miniature tabulate J02 208 corals, showing similar preservation modes. Preservation is often J02 209 clear in mudstones with specimens weathering out. Good preservation J02 210 of external morphology and, then, sectioning is necessary for close J02 211 identification. Fenestrate bryozoans are typically exposed with the J02 212 apertures (smaller than the fenestrules) facing down (Fig. 4.20), J02 213 so it is necessary to search for material with apertures facing up. J02 214 Slender, encrusting cyclostomes are easily damaged. In many J02 215 cheilostomes the frontal surfaces include delicate sculpture, J02 216 ovicells and avicularia, costae and spines. Some loss of detail may J02 217 have taken place because of abrasion or dissolution. Also, J02 218 replacement usually leads to loss of detail, though silicified J02 219 bryozoans within hollow flints may be well preserved externally. J02 220 Impressions of the external surface of Palaeozoic forms are of J02 221 little significance.

J02 222 Larger foraminifera (C) Larger foraminifera are J02 223 common at certain times and in certain facies and are generally J02 224 well preserved in fresh rock. Fusulines (Upper Palaeozoic), J02 225 alveolinids (Mesozoic onwards), and rotalids (e.g. Nummulites, J02 226 orbitoids of the Tertiary) all had calcitic skeletons. In J02 227 limestones, preservation is generally good. Individuals may weather J02 228 out, or blocks may be collected for later slabbing. In sand and mud J02 229 rocks, weathering may lead to dissolution. Break some specimens J02 230 open to check that skeletal detail is present.

J02 231 Stromatoporoids (F) are a heterogeneous group. Break off a J02 232 piece of rock and examine the fresh, wetted surface with a hand J02 233 lens to spot the skeletal laminae and pillars (never present in J02 234 stromatolites). Thin-sections are required for identification.

J02 235 Gastropods (I) Most modern gastropods have an entirely J02 236 aragonitic skeleton. The operculum, when present, is horny or J02 237 aragonitic. J02 238 J03 1 <#FLOB:J03\>Doubtless, as understanding of the structure and J03 2 function of the world's ecosystems increases, and as monitoring of J03 3 their degree of degradation improves, new worries will arise. Since J03 4 1973, the 'oil-crisis' has attracted attention, yet the potentially J03 5 much more serious problem of soil-loss has generated far less J03 6 interest in spite of the problems it caused in the mid-west of the J03 7 USA in the 1930s (Brown, 1977: 18; 1978).

J03 8 Different processes of degradation could act synergistically J03 9 and some have a cumulative effect, so recognition of the degree of J03 10 threat may not be easy. The management of land degradation is J03 11 therefore an art, with a huge palette of responses or avoidance J03 12 procedures, the application of which involves value judgements, J03 13 wisdom and scientific skill.

J03 14 2 Why is land degradation occurring?

J03 15 CAUSE AND PROCESS

J03 16 LAND DEGRADATION is commonly blamed on 'acts of God' or 'acts J03 17 of the peasantry', and there is typically little attempt to assess J03 18 the real causes. Another approach is to argue that, as land J03 19 degradation is widespread in environments exploited by Man, then J03 20 human actions must be the main cause. A wide range of human J03 21 activities can trigger or exacerbate land degradation; there are J03 22 also environments which are very vulnerable to degradation, and J03 23 natural catastrophes from time to time degrade virtually all J03 24 environments. Also, it would be rash to assume that human activity J03 25 always has a negative effect on rate of erosion, species survival, J03 26 etc.

J03 27 Sometimes the process of land degradation can be followed, but J03 28 the cause(s) may be obscure. Commonly, a chain-of-causation J03 29 stretches away in space and/or time from the site where land J03 30 degradation is manifest (see Fig. 2.1) (Eckholm, 1976; Darkoh, J03 31 1987: 25; Blaikie & Brookfield, 1987: 4).

J03 32 Sometimes causes are local and relatively simple, sometimes J03 33 land degradation results from, possibly complex, global changes J03 34 some of which are at least partly caused by human activity. While J03 35 the scale of global processes may be vast, they may be in a state J03 36 of dynamic equilibrium, easily upset by human actions. Chadwick & J03 37 Goodman (1975: 4) recognized three types of bio-geochemical cycle: J03 38 natural; perturbed (upset by Man) and recycling, i.e. managed by J03 39 Man to be sustainable. The second type is increasingly common, the J03 40 last type is often elusive.

J03 41 Land is a 'stage' within, upon, or above, which, a number of J03 42 resources may be exploited. Where there is exploitation of more J03 43 than one resource, this may be mutually compatible, or there may be J03 44 damaging interactions. Resources vary in character and some are J03 45 more difficult to manage than others (Ramade, 1984). The following J03 46 classification of types of resource is generally accepted and gives J03 47 some indication of 'manageability':

J03 48 1. Continuous resources include: solar energy, wind, J03 49 gravity, tidal energy, geothermal energy. These continue to be J03 50 available, and, with the possible exception of solar energy, the J03 51 receipt of which could be affected by atmospheric pollution, cannot J03 52 be degraded, even with gross mismanagement.

J03 53 2. Renewable resources (flow resources) include: J03 54 clean water, flora, fauna, soil, clean air. Rees (1985: 224), J03 55 defined renewable resources as "... those capable of J03 56 natural regeneration into useful 'products' within a timespan J03 57 relevant to man." These resources are potentially renewable J03 58 and could be indefinitely available, provided their capacity to J03 59 regenerate is not damaged by natural catastrophe or human J03 60 activities. Once degraded beyond a certain critical point, a J03 61 renewable resource may never recover.

J03 62 table&caption

J03 63 3.Non-renewable resources (stock resources) include: J03 64 many minerals and some groundwaters. These are available only in J03 65 finite quantities, or else the rate of renewal is so slow that they J03 66 must be regarded as available only in fixed quantities.

J03 67 4. Extrinsic resources include: human skills, J03 68 institutions, management abilities, etc. They can be fickle and J03 69 prone to breakdown or degradation, yet can be continuous resources J03 70 if well managed (Riddell, 1981: 23).

J03 71 In categorization of resources, one should not lose sight of J03 72 the fact that non-renewable resources and continuous resources are J03 73 opposite ends of the same continuum and that the above classes J03 74 overlap. The demand for a resource can vary because attitudes, J03 75 tastes, willingness or ability of people to use or purchase the J03 76 resource alter, or because a substitute has been found. J03 77 Technological advances and altered world circumstances affect J03 78 resource demand, people: "... live and act at a local J03 79 level, but their alternatives and opportunities are shaped J03 80 significantly by events at the regional, national and international J03 81 levels" (Dorner & El-Shafie, 1980: 8).

J03 82 caption&table

J03 83 Land must be described as a non-renewable (fixed stock) J03 84 resource, although it has a renewable capacity to support most J03 85 forms of biological life (Rees, 1985: 224). Exploitation of the J03 86 world's land has been described by some as conforming to a J03 87 "lollipop model", in that, with each J03 88 "lick", there is less for the future (Chisholm & Dumsday, J03 89 1987: 354). On the other hand, with good management, it might be J03 90 possible to sustain indefinite usage of land, even improve its J03 91 utility.

J03 92 Alternative ways of classifying Earth resources include:

J03 93 1. a division into - those that can be safely 'stretched' by J03 94 Man;

J03 95 - those that can be safely 'stretched' only if carefully J03 96 managed;

J03 97 - those that cannot/should not be 'stretched'.

J03 98 2. a division into - resources with actual value;

J03 99 - resources with option value (possible use perceived);

J03 100 - resources with intrinsic value (no obvious practical value, J03 101 but there is a will to maintain it).

J03 102 The processes of land and other environmental degradation are J03 103 not adequately understood, so predictive models are mainly poorly J03 104 developed (Chisholm & Dumsday, 1987: 322). Table 2.1 groups various J03 105 theses which seek to explain why land degradation takes place and J03 106 Figure 2.1 shows causes or combinations of causes that may result J03 107 in land degradation.

J03 108 Natural hazards as a cause of land degradation

J03 109 There are few, if any, regions so blessed that they will not at J03 110 some point experience a natural disaster. However, some J03 111 environments are less stable than others or are more likely to J03 112 suffer disruptions. Areas predisposed to disaster include:

J03 113 - steeply sloping areas;

J03 114 - easily damaged soils;

J03 115 - drylands and localities where soils drain fast;

J03 116 - lowlands close to the sea, particularly on exposed coasts and J03 117 in areas prone to glacier 'calving', submarine eruption or J03 118 submarine <}_><-|>mudlslides<+|>mudslides<}/>;

J03 119 - regions where rainfall is intense;

J03 120 - drought-risk areas where rainfall is mainly due to airmass J03 121 movements which can be fickle ('monsoon' rainfall);

J03 122 - parts of the Earth where hurricanes or similar storms J03 123 occur;

J03 124 - areas prone to sudden frost or cold winds;

J03 125 - areas of earthquake or volcanic activity;

J03 126 - areas subject to periodic invasion by destructive insects.

J03 127 There is evidence of periodic natural catastrophes (infrequent J03 128 on the human time-scale, but not if judged by the geological J03 129 time-scale) which caused widespread and severe land degradation. A J03 130 number of scientists argue that massive extinctions of organisms at J03 131 the end of the Permian and Cretaceous Periods reflect such J03 132 catastrophes. Huggett (1988, 1989) provided a fascinating study of J03 133 catastrophic tsunamis, arguing that huge waves have been quite J03 134 common. The UK he suggested has been, and probably will be, struck J03 135 every 0.8 to 1.4 million years by 13 to 130 m tsunamis. Worse, J03 136 there are indications of waves several hundred metres in J03 137 height caused in the past by meteorite/comet strike, eruptions or J03 138 other Earth processes. There is evidence of climatic disruption, J03 139 sufficient to seriously affect human affairs following the Hekla-3 J03 140 volcanic eruption (Iceland c. 1120 BC) and the Santorini J03 141 volcanic eruption (Crete c. 1645 BC).

J03 142 Storms, cyclones and hurricanes are common causes of land J03 143 degradation. For example, in 1986, Guadalcanal in the S. Pacific J03 144 was devastated by Cyclone Namu, and in 1988 the Atlantic coastlands J03 145 of Nicaragua were struck by Hurricane Joan destroying around 7000 J03 146 km2 of tropical forest (The Times, 29/11/88: J03 147 24). In the short term, these events cause degradation, but, over J03 148 the longer term, without such damage, the natural regeneration of J03 149 some tropical forests would be upset, for there would be no J03 150 clearings and less chance for fresh growth.

J03 151 The onset and/or severity of a natural disaster may owe J03 152 something to human activity - Man may have 'triggered' the disaster J03 153 or may have 'sensitized' the environment. If development has J03 154 exceeded environmental limits, a natural disaster might speed up J03 155 what may otherwise have been a virtually inevitable decline. J03 156 Progress sometimes comes out of adversity - in reacting to a J03 157 disaster Man may develop new strategies for using land or resources J03 158 which may be less degrading; indeed, much development has been J03 159 spurred on by natural and human disasters; notably scientific J03 160 advances in times of war.

J03 161 The worst possible scenario is for Man to practise a land use J03 162 which makes land liable to degradation in an area prone to J03 163 disasters on land that is especially vulnerable. At the other J03 164 extreme, there is good management of land that is not vulnerable in J03 165 a region seldom subject to disasters. It should be possible to map J03 166 natural hazard risk and so be prepared for, or to avoid, some J03 167 disasters.

J03 168 Population change as a cause of land degradation

J03 169 Population increase has been one of the most frequently cited J03 170 causes of land degradation since Malthus drew attention to it. The J03 171 Malthusian, or more recently Neo-Malthusian, view is that J03 172 increasing demographic pressure results in overuse of reasonable J03 173 quality land and/or the misuse of marginal, often easily degraded J03 174 land. If population increase has that effect, the impact is double J03 175 edged: a simultaneous increase in demand made upon the environment J03 176 in order to support growing numbers of people, and a destruction of J03 177 the resource base (Clark & Munn, 1986: 8-12).

J03 178 There is a need to treat the argument that environmental J03 179 degradation arises whenever population grows, even when it exceeds J03 180 a region's 'carrying capacity', with caution, for there are areas J03 181 with large numbers of people and relatively little damage and there J03 182 are areas with very few people, a short settlement history and much J03 183 damage. In sensitive areas, or with certain types of exploitation, J03 184 population need not be high to cause problems. Population J03 185 stagnation or decline can also be a cause of land degradation.

J03 186 Globally, human population has shown three periods of J03 187 relatively rapid increase, if a logarithmic plot is used, followed J03 188 by periods of relative stability. Arithmetic plotting of the J03 189 statistics has reinforced the idea that growth has been J03 190 exponential. There are some who hope that there might be another J03 191 period of reduced growth in the late twenty-first-century. Whether J03 192 or not that happens, the world population will have risen above J03 193 6000 million by AD 2000. The world today supports roughly J03 194 three times the human population and roughly 100 times the J03 195 industrial activity it did a century ago. Although they comprise J03 196 only about 30% of the world's population, those in developed J03 197 countries consume roughly five times as much food and commodities J03 198 as people in developing countries and produce more industrial J03 199 pollutants.

J03 200 A number of researchers have tried to establish where J03 201 population has exceeded the limits of food production (Higgins J03 202 et al., 1982; Mahar et al., 1985). However, the J03 203 indications are that population - poverty, and possibly J03 204 degradation, correlations can be spurious. People, on the whole, J03 205 have children in response to their economic, political and J03 206 historical situation - a simplistic Malthusian perspective misses J03 207 this (Hecht, 1985: 665). Redclift (1987: 30) suggested that it is J03 208 not so much net global population increase that matters, but the J03 209 rate of change in population in critical regions.

J03 210 There has been debate on the relationship between population J03 211 density, growth rate and agricultural development, particularly the J03 212 intensity of farming (Boser<*_>u-umlaut<*/>p, 1965; Carlstein, J03 213 1982). The relationship is by no means a simple one; put crudely it J03 214 has been claimed that, if a population does not grow enough, or J03 215 grows too fast for agriculture to respond, then production is J03 216 likely to remain extensive, and may cause land degradation. If J03 217 population grows, but not too fast, then intensive agriculture, J03 218 possibly causing less land degradation, may result.

J03 219 Marginalization as a cause of land degradation

J03 220 People may become marginalized - forced or attracted onto poor J03 221 quality, possibly easily degraded land, because better land is J03 222 settled; because of unrest; because large landowners monopolize the J03 223 better land, and possibly make poor use of it; because of the J03 224 creation of reserves. It is not always unrest or social J03 225 differentiation which causes people to settle marginal land. The J03 226 land may be available and people perceive it offers them a chance J03 227 for a living or a profit (Reining, 1978: 75).

J03 228 J04 1 <#FLOB:J04\>Wolfe [14] and Olson [42] have produced generalized J04 2 vegetation maps for this period. Figure 9.4 is based on these maps, J04 3 and shows clearly the vegetation changes that are presumed to have J04 4 taken place since the Middle Eocene some 30 Ma earlier (Figure J04 5 9.2). The wide-ranging, large-leaved, evergreen, tropical and J04 6 paratropical rain-forests of the Eocene had contracted J04 7 through about 35<*_>degree<*/> of latitude, and been replaced at J04 8 higher latitudes by smaller-leaved seasonal forests and woodlands J04 9 with a variable (but often high) proportion of deciduous taxa. The J04 10 greater diversity of vegetation types in the Middle Miocene (nine, J04 11 as against five Eocene types) was largely the result of adaptation J04 12 to strongly seasonal climates at all latitudes: deciduous woodlands J04 13 in the tropics and subtropics of Africa and India, sclerophyll J04 14 woodlands between latitudes 20<*_>degree<*/> and 40 <*_>degree<*/> J04 15 in both hemispheres, and mixed coniferous forest at higher J04 16 latitudes and altitudes in the northern hemisphere. Nevertheless, J04 17 the tenuousness of the evidence on which this Miocene map is based J04 18 must not be overlooked. The vegetation reconstructions for almost J04 19 the whole of the southern hemisphere rely on a very small number of J04 20 fossil sites, and rather different reconstructions, based on the J04 21 same very limited data, have been proposed for South America [43], J04 22 Africa [9] and Australia [44]. The extent to which non-forest types J04 23 (savanna, grassland, or even desert), in particular, had developed J04 24 in continental interiors in the subtropics is quite unclear.

J04 25 The best documented feature of the Middle Miocene vegetation is J04 26 the wide-ranging occurrence of temperate broadleaved forests in the J04 27 northern hemisphere. These forests contained taxa now J04 28 characteristic of the present-day forest vegetation of western and J04 29 eastern North America, non-Mediterranean Europe and East Asia, but J04 30 were more diverse than any single modern forest-type in these J04 31 regions. Parallels have been drawn with the modern "mixed J04 32 mesophytic forests" of China [45], but a better analogy J04 33 might be to regard them as an amalgam of at least four of Wolfe's J04 34 forest-types [16] (types 6-9 in Figure 9.1). Although there must J04 35 have been regional variation in these Miocene forests, many taxa J04 36 apparently occurred over almost the entire range. Among these taxa J04 37 were: cold-temperate hardwoods such as Alnus, Betula, Corylus, J04 38 Nyssa, Salix and Ulmus; cold-tolerant conifers such as J04 39 Picea, Pinus and Tsuga; warm-temperate trees such as J04 40 Carya and Liquidambar (still widespread), J04 41 Ceridiphyllum and Glyptostobus (now restricted to eastern J04 42 Asia), and Sequoia (now restricted to western North America); J04 43 and subtropical taxa such as Engelhardtia. All these (and many J04 44 other) taxa had also been widespread earlier on, in the Middle J04 45 Eocene Epoch, but in rather different forest-types. Consequently, J04 46 to regard the middle-latitude Miocene forests as derived solely J04 47 from southward displacement of an Eocene Arctotertiary Geoflora J04 48 would be erroneous. The middle-latitude Miocene flora certainly J04 49 contained many temperately-adapted taxa of that geoflora J04 50 that had migrated southward with deteriorating climate, but it had J04 51 also drawn in taxa from the Early Tertiary paratropical forests - J04 52 taxa that had been 'left behind', and had evolved in situ J04 53 to produce new lineages adapted to colder conditions than those J04 54 favoured by their Eocene ancestors [45].

J04 55 Similar considerations must also apply to the flora of the J04 56 sclerophyll woodlands at lower latitudes (the Madrean-Tethyan J04 57 Geoflora), associated particularly with the subtropical J04 58 high-pressure belt and its margins in both hemispheres. That belt, J04 59 with its seasonal or more prolonged drought conditions, remained J04 60 much more nearly stationary as vegetation zones were displaced J04 61 progressively equatorwards through the Tertiary Period. J04 62 Consequently, new lineages were continually being drawn in as J04 63 temperate taxa migrated equatorwards. Some adapted to the drier J04 64 climates, and persisted, so that recruitment to the Madrean-Tethyan J04 65 Geoflora was continually occurring. The concept of geofloras as J04 66 originally proposed by Chaney (section 1.3) can thus have only J04 67 limited signficance in the context of Tertiary vegetation history J04 68 [45,46].

J04 69 9.3.3 The spread of non-forest communities

J04 70 Following the thermal optimum of the Middle Miocene Epoch J04 71 (section 4.4), renewed orogenesis affected most parts of the globe, J04 72 and continued, often with increasing intensity, into the subsequent J04 73 Pliocene Epoch (section 4.2). As a result, the range of climatic J04 74 and topographic conditions over the earth's surface was enhanced, J04 75 and global ecospace was further diversified. Ocean temperatures J04 76 (and presumably land temperatures also) fell sharply between about J04 77 15 and 13 Ma ago, and again between 7 and 5 Ma ago, leading first J04 78 to enlargement of the East Antarctic Ice Sheet and then to icing-up J04 79 of West Antarctica [47]. The ensuing fall in global sea levels, J04 80 coupled with intensification of high pressure in the subtropics and J04 81 increasing rainshadow effects in the lee of newly-uplifted mountain J04 82 ranges, combined to make the Late Miocene the driest part of the J04 83 Tertiary Period. This arid episode was manifested most J04 84 spectacularly in the almost complete drying-up of the Mediterranean J04 85 Sea [48]. In northern Israel as a result [49], pollen spectra with J04 86 50% tree pollen in the Middle Miocene (and including such J04 87 mesophytic taxa as Alnus, Corylus, Engelhardtia, Juglans, J04 88 Platycarya and Pterocarya) were replaced in the Late J04 89 Miocene by spectra with almost no tree pollen, and with 'arid' J04 90 indicators such as Artemisia and Chenopodiaceae prominent.

J04 91 There have been speculations that a wide range of dryland J04 92 communities had developed earlier, in the drier conditions of the J04 93 Oligocene 25-35 Ma ago [9,50], with centres of aridity in South J04 94 America in the rainshadow of the rising Andean Cordillera, in the J04 95 low-latitude Sahara and Kalahari Deserts, and in the remote J04 96 interior of Middle Asia [51]. However, the major evidence of J04 97 diversifying dryland vegetation comes from the Late Miocene and J04 98 Pliocene. Certainly, the environmental conditions necessary for the J04 99 creation of some present-day desert regions did not exist before J04 100 the Miocene: the Middle Eastern desert region was largely submerged J04 101 below the Tethys Sea [52], the American Southwest was still open to J04 102 the influence of rain-bearing westerly winds, and parts of Middle J04 103 Asia were not yet isolated from the south-west monsoons by the J04 104 mountain barrier of the Himalayas. All that changed with the J04 105 frenzied tectonic activity of the Miocene and Pliocene, and during J04 106 the last 5-10 Ma of the Tertiary Period non-forest communities J04 107 became widely established over the same general areas occupied J04 108 today by desert, steppe and savanna vegetation.

J04 109 However, although floras containing non-woody taxa similar to J04 110 those predominant today in desert, steppe and savanna areas were J04 111 undoubtedly present in the Late Tertiary, the component taxa might J04 112 have been assembled into communities very different from modern J04 113 dry-climate ones. Thus pollen floras of Late Miocene and Pliocene J04 114 age, even from present-day desert areas, almost invariably contain J04 115 tree pollen, often of mesophytic taxa. For example, pollen of J04 116 Betulaceae, Fagaceae, Juglandaceae and Ulmaceae accompanies that of J04 117 Pinus, Artemisia and Chenopodiaceae in Miocene pollen J04 118 spectra from the Gobi Desert region of north-west China [53]. J04 119 Whether this pollen was transported from faraway deciduous forest J04 120 sources to the south, or whether it was derived from local woodland J04 121 patches growing in damper situations within the region, is not J04 122 clear. Most palaeoecologists would probably favour an explanation J04 123 based on local sources of pollen in arid regions - either from J04 124 continuous scrubby woodland with an understorey of desert plants, J04 125 or from local woodland patches intermixed with more open J04 126 vegetation.

J04 127 Nevertheless, the macrofossil, pollen and faunal evidence for J04 128 the Late Tertiary supports the existence of a wide range of dryland J04 129 communities. True desert conditions may have been present in the J04 130 Sahara and the Namib [54], and semi-desert in southern Africa and J04 131 the Near East [9,49]. The extensive saline deposits that formed as J04 132 the remnants of the Tethys Sea evaporated could have triggered the J04 133 evolution of many desert halophytes [15]. Some sort of J04 134 sclerophyllous woodland was almost certainly present in the J04 135 American Southwest, whilst fossil floras of Pliocene age from the J04 136 Great Plains region of the United States attest to a gradual J04 137 restriction of forest to river valleys there as grassland J04 138 communities developed on the drier interfluves. The existence of J04 139 extensive areas of open grassland is suggested also by the presence J04 140 of grazing and browsing faunas (horses, camels and osteodonts) in J04 141 Kansas and Oklahoma after 10 Ma ago [55]. The presence of grazing J04 142 faunas has similarly been used as evidence of grassland conditions J04 143 in South America, but here aridification occurred rather earlier - J04 144 by 15 Ma ago at least [13] - and by the Pliocene thorn scurb and J04 145 grassland probably occupied much of southern South America east of J04 146 the Andes [13,43]. More open conditions probably developed within J04 147 the Amazonian forest too in the Pliocene, to form a major 'savanna J04 148 avenue' extending north from the dry woodlands of the Brazilian J04 149 Chaco, through the Mato Grosso of Brazil, to the highlands of J04 150 Guyana and Venezuela [13,56].

J04 151 In East Africa, grazing faunas diversified later, about 2.5 Ma J04 152 ago, at a time when pollen evidence indicates a contraction of J04 153 rainforest [57]. Savanna and grassland may have spread sufficiently J04 154 then to separate the formerly continuous rainforest into eastern J04 155 and western segments [9]. In India the dominant rainforest of the J04 156 Middle Tertiary retreated gradually eastwards and southwards in the J04 157 Pliocene, as arid conditions developed in Pakistan and to the south J04 158 of the rapidly-rising Himalayas [8,58]. Arid conditions almost J04 159 certainly developed too in the interior of the Australian J04 160 continent, but adequate fossil documentation is lacking J04 161 [2,59,60].

J04 162 In areas outside the tropics where seasonally dry climates J04 163 developed in the Late Tertiary, the major dry period initially J04 164 occurred during the cool season, and 'mediterranean' climates with J04 165 summer drought did not develop until the accumulation of polar ice J04 166 allowed the formation of cold ocean currents affecting the western J04 167 sides of the major land masses [61]. For some areas, full summer J04 168 drought conditions may not have been realized until the glacial J04 169 stages of the Pleistocene [9], but in the northwest Mediterranean J04 170 Basin at least there is some evidence that dry summers developed in J04 171 the Late Pliocene. About 2.8 Ma ago here, pollen of woody taxa J04 172 requiring year-round moisture (e.g., Engelhardtia, Hamamelis, J04 173 Nyssa and Symplocos) declined to near-extinction, and J04 174 pollen of taxa characteristic of present-day shrublands around the J04 175 Mediterranean Basin (Cistus, Olea, Phillyrea, Pistacia, J04 176 Quercis ilex-type) became a constant component of the J04 177 sediments [62]. However, it must not be forgotten that adaptation J04 178 to summer drought is a secondary evolutionary feature that J04 179 developed in plants already adapted to seasonal drought; the J04 180 adaptation necessitated a shift in the time of seedling germination J04 181 and establishment from late spring/early summer to late J04 182 winter/early spring, together with concomitant adjustments in the J04 183 timing of flowering and seed maturation [8]. Hence the presence of J04 184 'mediterranean' genera in the fossil record does not necessarily J04 185 imply the existence of a mediterranean climate. Arbutus, J04 186 Chamaerops, Juniperus, Pistacia and Quercus J04 187 coccifera, for example, characteristic members of present-day J04 188 shrublands in southern Europe, are recorded from European sites J04 189 back to the early Oligocene, when they were part of the developing J04 190 Madrean-Tethyan Geoflora [63].

J04 191 Spread of non-forest plant communities late in the Tertiary J04 192 Period occurred also at high altitudes, as mountain ranges were J04 193 thrust up above the attainable treeline. Documentation of this J04 194 spread of herbaceous plants relies largely on the pollen record, J04 195 and is consequently rather non-specific. However, it is clear both J04 196 from the pollen record and from present-day distribution patterns J04 197 that the colonizing species of these newly-created habitats were J04 198 drawn both from local sources and from more distant ones. Thus J04 199 pollen analyses from the high northern Andes [43] show developing J04 200 montane and alpine floras in the Pliocene Epoch that were derived J04 201 mainly from neotropical elements, but drew in some North American J04 202 taxa as well (Alnus, Aster, Cruciferae, Ericaceae, J04 203 Erigeron, Sambucus and Viburnum) and at least one J04 204 Antarctic one (Lagenophora). Present distribution patterns of J04 205 alpine plants on the East African mountains [9] similarly indicate J04 206 an input of exotic genera in the past - from southern Africa and J04 207 from temperate regions of the northern hemisphere.

J04 208 Adaptation to conditions of extreme cold almost certainly J04 209 occurred in these habitats and not in high-latitude (arctic) ones. J04 210 Tundra vegetation dominated by herbaceous plants was almost J04 211 non-existent before the Pleistocene Epoch, with boreal forest and J04 212 spruce parkland reaching north to the shores of the Arctic Ocean J04 213 [64]. At a few sites only is there pollen evidence of dwarf-shrub J04 214 tundra in the Late Pliocene: on Seward Peninsula, Alaska [39], and J04 215 in the Kolyma lowlands of north-east Siberia [65].

J04 216 9.4 THE TERTIARY VEGETATION OF WESTERN NORTH AMERICA

J04 217 J05 1 <#FLOB:J05\>Framboidal Pyrite Nucleation and Crystal J05 2 Growth: a progress report

J05 3 by David Rickard and Ian Butler

J05 4 Framboidal pyrite represents a simple two component inorganic J05 5 system that readily reaches a remarkable degree of organizational J05 6 sophistication. Apart from their aesthetic attributes, framboids J05 7 are interesting because they may tell us something about the J05 8 fundamental processes involved in the development of organized J05 9 systems and about the physico-chemical conditions of ancient J05 10 environments. Framboids also have a close relationship with J05 11 biological materials and processes: (a) they are intimately J05 12 involved in biomineralization, especially in pyritization of plants J05 13 (b) microbial sulphate reduction is the major source of framboidal J05 14 sulphide and (c) organic matter forms the matrix between J05 15 microcrysts in some framboids and enclosing outer sacs in J05 16 others.

J05 17 Framboid characteristics

J05 18 Any model for the formation of framboids must account for the J05 19 following properties: (1) the dominant mineral form in framboids is J05 20 pyrite (2) up to 107 pyrite mycrocrysts (usually < J05 21 1<*_>mu<*/>m in size) occur in spheroids usually less than J05 22 100<*_>mu<*/>m in diameter (3) the even size of the microcrysts J05 23 within any framboid (4) the varying habits of the microcrysts in J05 24 different framboids (5) the occurrence of ordered patterns J05 25 involving either ccp or hcp organization (6) the occurrence in some J05 26 framboids of organic material between the microcrysts and/or around J05 27 the framboids.

J05 28 Pyrite formation

J05 29 The formation of pyrite from aqueous solutions at low J05 30 temperatures proceeds via three known pathways:

J05 31 (1) Reaction of precursor metastable iron sulphides (MIS) with J05 32 polysulphides (Rickard 1975). Schoonen (1989) examined the reaction J05 33 in some detail and Luther (in prep) has confirmed Rickard's J05 34 kinetics.

J05 35 (2) Reaction of Fe3S4 (greigite) with J05 36 sulphur species to form pyrite (Sweeney and Kaplan 1973). This J05 37 reaction can lead to the formation of framboids.

J05 38 (3) Reaction of Fe (II) sulphide with H2S to form J05 39 pyrite with the production of H2 gas (Taylor et al 1979; J05 40 Dobner et al., 1990). We have reproduced this reaction and detected J05 41 H2 gas. We are at present running a programme to examine J05 42 the kinetics and mechanism of the process.

J05 43 In each case, pyrite formation at low temperatures proceeds J05 44 through a MIS precursor. It appears that this precursor pathway for J05 45 pyrite formation is related to its propensity to form the J05 46 framboidal texture.

J05 47 Direct precipitation of pyrite from solution

J05 48 The direct precipitation of pyrite has not been achieved J05 49 experimentally.

J05 50 One of the problems of experimental work in these systems is J05 51 the enormous supersaturations necessary in most experimental work. J05 52 If the kinetics of nucleation of MIS was far faster than that for J05 53 pyrite, then these supersaturations, within the MIS stability J05 54 zones, might precipitate the metastable phase before pyrite: i.e. J05 55 the pyrite requirement for a precursor phase could be merely J05 56 kinetic and pyrite would precipitate directly from solution if the J05 57 solutions were undersaturated with respect to the MIS phases and J05 58 oversaturated with respect to pyrite. Schoonen (1989) arranged J05 59 experiments below saturation for the MIS but saturated with respect J05 60 to pyrite, but was unable to precipitate pyrite directly. Dales and J05 61 Rickard (unpublished) have arranged such an experimentation through J05 62 the diffusion of aqueous iron (II) species and sulphur-bearing J05 63 species <}_><-|>form<+|>from<}/> opposite ends of a tube filled J05 64 with a porous medium. Pyrite did not form and this implies that J05 65 pyrite does not precipitate directly from solution. This conclusion J05 66 has been supported by Luther's (in prep) experimentation.

J05 67 Precursor Metastable Iron Sulphides (MIS)

J05 68 The nature of the precursor material is important for the J05 69 formation of the framboidal microarchitecture. Rickard (1990) J05 70 showed that the first formed phase was probably a<&|>sic! iron (II) J05 71 bisulphide cluster colloid (Silvester et al 1991) with a size J05 72 between the conventional 0.1<*_>mu<*/>m colloidal range and the 1nm J05 73 cluster range. This undergoes rapid re-reaction with the expulsion J05 74 of H2O and H2S to form amorphous Fe(II) J05 75 sulphide. A broad x-ray peak at around 0.5nm is observed after 2 J05 76 days of ageing at 25<*_>degree<*/>C coincident with the [001] J05 77 reflection of the tetragonal iron (II) sulphide mackinawite. This J05 78 appears to be evidence for the beginning of long-range ordering in J05 79 the Fe(II) sulphide. Further mackinawite reflections appear over J05 80 the next two years of ageing, indicating the increase of J05 81 short-range ordering in the material.

J05 82 The pathway for the formation of the cubic J05 83 Fe3S4, greigite, is not well documented. As far J05 84 as we are aware, greigite has not been synthesised directly from J05 85 solution, but is thought to be produced as an oxidation product of J05 86 iron (II) sulphide. Pyrrhotite-group iron (II) sulphides, with J05 87 basic hexagonal symmetries, have been produced J05 88 <}_><-|>form<+|>from<}/> aqueous solutions at low temperatures J05 89 (Rickard 1969; Sweeney and Kaplan 1974). However, the detailed J05 90 conditions for pyrrhotite development at low temperatures are J05 91 unknown.

J05 92 Pyrite nucleation

J05 93 The occurrence of precursor MIS before the more stable pyrite J05 94 has been cited as an example of Ostwald's Rule of Successive J05 95 Reactions. This is only partly true, since pyrite is more oxidised J05 96 than the precursor MIS and its formation requires oxidation J05 97 reactions as well as equilibration.

J05 98 For pyrite to nucleate, the solution must be supersaturated J05 99 with respect to pyrite. Rickard (1975) assumed that the apparent J05 100 reluctance of pyrite to nucleate directly from solution was purely J05 101 mechanistic in origin. The reaction between polysulphide and iron J05 102 (II) salts was investigated in detail by Schoonen (1989). His J05 103 results show that reactions like

J05 104 formula

J05 105 are faster than reactions like

J05 106 formula

J05 107 Once the reaction involves one or more precursor solid phases J05 108 it becomes potentially kinetically slower. On the other hand, this J05 109 also means that substantial supersaturations can build up and exist J05 110 metastably for some time. This would be consistent with the J05 111 framboidal property that the pyrite microcrysts within any given J05 112 framboid are of similar size: this property requires that the J05 113 pyrite microcrysts nucleated throughout the framboid. If this were J05 114 not so then the pyrite would tend to grow by simple crystal growth J05 115 mechanisms, which Schoonen has shown is relatively rapid for J05 116 pyrite, and form crystal aggregates with diverse sizes.

J05 117 However, it does not explain the organisation of microcrysts in J05 118 some framboids. Mann (1988) showed that, with respect to organic J05 119 matrices, two conditions must be satisfied: (1) molecular J05 120 preorganisation and (2) molecular complementarity between inorganic J05 121 ions and local matrix binding sites. Applying this to framboids, J05 122 microscopic and molecular organization within the precursor MIS J05 123 must be completed before pyrite nucleation starts. Otherwise, J05 124 nucleation will be at random sites throughout the iron sulphide J05 125 matrix. This is the situation observed in the experiments of J05 126 Rickard (1969) and Luther (in prep) where continuous reaction J05 127 between MIS and oxidised sulphur species gave precipitates of J05 128 random pyrite microcrysts: the MIS reactant was relatively young J05 129 and still disorganized.

J05 130 Some degree of order must be present in the precursor MIS J05 131 either as specific microdomains within the material or as a regular J05 132 aggregate of particles for the regular arrays of pyrite microcrysts J05 133 to be developed. Such order develops through ageing and/or J05 134 re-reaction of the MIS. In this interpretation the observed J05 135 variation in degree of ordering in pyrite framboids reflects the J05 136 preorganization in the precursor MIS matrix.

J05 137 The development of microdomains or regular particle aggregates J05 138 will produce active sites within the precursor iron sulphides, J05 139 which in themselves may catalyse the nucleation of pyrite. Indeed, J05 140 Taylor (1980) suggested that the S2 moiety may only be J05 141 produced at active sites on MIS surfaces or within MIS J05 142 aggregates.

J05 143 Both mackinawite and greigite are possible precursors to J05 144 pyrite, although only greigite has definitely been implicated in J05 145 framboid formation to date. There is thus at least a theoretical J05 146 possibility for different organizational geometries in the J05 147 precursor material, giving rise to the differently ordered arrays J05 148 of pyrite microcrysts observed in framboids.

J05 149 Microcryst form

J05 150 At high temperatures, Murowchik and Barnes (1987) were able to J05 151 show that <}_><-|>pryrite<+|>pyrite<}/> crystal morphology varied J05 152 in response to increasing saturation in the order J05 153 cube-octahedron-pyritohedron. The variation in microcryst form J05 154 (usually cubes or octahedrons) in framboids may reflect this. J05 155 Pyrite may nucleate under varying degrees of supersaturation J05 156 dependent on the local free energy of the surface. Interestingly we J05 157 have also observed microcrysts with holes in them, suggesting that J05 158 crystal growth in these cases proceeded from the outside inwards. J05 159 We have also observed regular framboid microcrysts consisting of J05 160 aggregates of sub-spherical pyrite particles in the cluster colloid J05 161 size range.

J05 162 Organic matter

J05 163 The origin <}_><-|>if<+|>of<}/> the organic matter in some J05 164 framboids has been of interest since Schneiderhohn (1923) J05 165 originally thought the textures to be fossilized bacteria and Love J05 166 (1957) isolated apparent microorganic remains from framboids. J05 167 Rickard (1970) suggested that the organic matter derived from a J05 168 coercevate that had been sulphidised, in process J05 169 three-dimensionally equivalent to the synthetic production of CdS J05 170 in a polymer film (Bianconi et al 1991). The role of organics is J05 171 interesting since as Mann (1988) pointed out, they could act to J05 172 reduce the surface free energy of the pyrite nucleating system, J05 173 thereby catalysing microcryst nucleation. At present, it is unknown J05 174 whether the organic matter within some framboids is a natural J05 175 abiologic equivalent of nucleation within a crystalline polymer or J05 176 merely organic substances absorbed onto pyrite during diagenesis J05 177 (Kribek 1975) or forced into the spheres during burial. (Elverhoi J05 178 1977).

J05 179 Conclusions: Hypothesis and test

J05 180 In this model, different degrees of ordering of pyrite J05 181 framboids reflect different degrees of organization of the J05 182 precursor MIS. The crystallinity of the precursor MIS is a function J05 183 of the structural state of the precursor material prior to J05 184 pyritisation. This in turn is dependent on the age and composition J05 185 of the MIS. The nucleation energy barrier for pyrite may be J05 186 overcome by a variety of factors, including supersaturation (which J05 187 may determine microcryst growth habit), active sites in the MIS and J05 188 organic catalysis.

J05 189 The model appears to be testable and we are at present J05 190 attempting to synthesise framboids under controlled conditions J05 191 using well-defined starting materials.

J05 192 J05 193 ASPECTS OF BIOLOGICAL SILICIFICATION; MODEL SILICA J05 194 PRECIPITATION STUDIES IN THE PRESENCE OF CARBOHYDRATE POLYMERS

J05 195 Carole C. Perry

J05 196 Silica is an important industrial chemical manufactured J05 197 worldwide with applications including glasses, catalyst supports, J05 198 adsorbates, pigments and fillers. The structure of the silica J05 199 determines the properties of the material and yet the basic J05 200 mechanisms of particle formation and aggregation in aqueous J05 201 solution remain imperfectly understood. Biology has mastered the J05 202 art of producing well controlled macroscopic silica structures eg. J05 203 sponges, diatoms, plant components etc. although all biogenic J05 204 silicas are complex materials comprising both organic and inorganic J05 205 phases. Biogenic silicas are built up from essentially J05 206 monodispersed particles ranging from 2-15nm in diameter depending J05 207 upon the structure. At a higher level, these particles are J05 208 organised into structural motifs (fibrils, tubules, sheets etc.) J05 209 which indicates a substantial degree of control over both J05 210 polymerisation and particle aggregation.

J05 211 The formation of all inorganic solids including silica from J05 212 aqueous solution is achieved by a combination of three J05 213 physico-chemical steps; supersaturation, nucleation and crystal J05 214 growth or maturation. A further essential element for controlled J05 215 biological mineralisation is spatial localisation, which may occur J05 216 either through the use of membrane bounded compartments or specific J05 217 cell wall regions and allows for the local regulation and control J05 218 over physico-chemical factors through selectivity in biochemical J05 219 processes such as ion and molecular transport.

J05 220 Studies of the formation of biological silica in relation to J05 221 changes in mechanical stresses, ionic composition and polymer J05 222 matrices have shown that all these factors may be relevant in the J05 223 production of specific aggregate structures, often with varying J05 224 surface chemistry. It is important to note that local ionic J05 225 environment effects can have little or no effect on the regulation J05 226 of morphological features at the micron level and other factors J05 227 must be important.

J05 228 Macromolecular assemblages are thought to be extremely J05 229 important in the regulation of both nucleation and growth processes J05 230 for crystalline and amorphous minerals. In many instances of J05 231 crystal deposition, for both in vivo and in J05 232 vitro systems, the nature of the interaction involves a J05 233 specific charge/ stereochemical matching but for amorphous J05 234 materials including silica the precise nature of the interaction J05 235 between the organic and inorganic phases is not known. At present J05 236 our biological studies are directed towards this end.

J05 237 A direct result of our biological studies has been the J05 238 development of model chemical systems for the investigation of J05 239 crystal or aggregate formation under carefully controlled J05 240 experimental conditions.

J05 241 This lecture will give information on (1) the structure and J05 242 form of biological silicas, (2) the role of polymers, the ionic J05 243 environment and physical factors in the control of J05 244 biosilicification, and (3) the precipitation of silicas from J05 245 silicon containing complexes in the presence of carbohydrate J05 246 polymers.

J05 247 J06 1 <#FLOB:J06\>Gas-phase reactions

J06 2 The existence of oscillations and the potential for chaotic J06 3 behaviour in gas-phase reactions have received a much less J06 4 widespread coverage than their solution-phase counterparts. An J06 5 introduction to this topic can be found in Gray and Scott (1985), J06 6 Griffiths (1985a), and in the companion to this book (Gray and J06 7 Scott 1990). Recent reviews have also been given by Griffiths J06 8 (1985b, 1986) and Griffiths and Scott (1987).

J06 9 Perhaps the simplest non-linear behaviour of interest here J06 10 shown by gas-phase systems is that of thermal explosion. In many J06 11 ways this is the equivalent of the solution-phase clock reaction. J06 12 The reactants are mixed and a relatively slow evolution begins, but J06 13 after some (usually short) time there is a rapid acceleration in J06 14 rate leading to virtually complete conversion. This process relies J06 15 on the exothermic nature of many chemical processes and their J06 16 sensitivity to temperature through the reaction rate constant. J06 17 Gas-phase systems are much more prone to significant departures J06 18 from isothermal operation, and temperature plays the role of J06 19 'autocatalyst'. There can also be genuine chemical autocatalysis in J06 20 gas-phase reactions, particularly oxidation processes involving J06 21 oxygen. Such branded-chain reactions can also provide spectacularly J06 22 non-linear responses. If self-heating and chemistry conspire J06 23 together, almost anything seems possible.

J06 24 9.1. Transient oscillations from self-heating: an J06 25 experimental realization of the Salnikov model

J06 26 Di-tertiary butyl peroxide (DTBP) is an organic species that J06 27 undergoes a simple first-order decomposition reaction

J06 28 formula

J06 29 with a significant rate at temperatures above about 400 K. The J06 30 change in number of moles gives an increasing pressure under J06 31 constant volume conditions and this allows the extent of reaction J06 32 to be monitored relatively easily as a function of time. At low J06 33 temperatures within this range and especially if the reactant is J06 34 heavily diluted with an inert gas such as N2, the J06 35 reaction is very well behaved and is widely used to demonstrate J06 36 first-order kinetics in undergraduate courses.

J06 37 graph&caption

J06 38 The detailed kinetic mechanism is not of great importance here, J06 39 but is believed to involve the unimolecular breaking of the O-O J06 40 bond, followed by elimination of methyl radicals which may then J06 41 combine:

J06 42 formulae

J06 43 leading to the stoichiometry (1). Step (2) is usually rate J06 44 determining, so the overall kinetics are insensitive to any changes J06 45 brought about in steps (3) and (4).

J06 46 The overall process (1) is exothermic: formula and the J06 47 overall reaction rate constant k has been found to have an J06 48 activation energy of 152 kJ mol-1. If O2 is J06 49 added to the reacting vapour, the effect is to increase the J06 50 exothermicity (by changing the final products through steps that J06 51 compete with (4) above) but this does not affect the kinetics which J06 52 remain first order with the same Arrhenius parameters.

J06 53 At sufficiently high reaction temperatures, the exothermic heat J06 54 release accompanying reaction can lead to self-heating. The primary J06 55 requirement for self-heating is that the reaction rate and, hence, J06 56 the heat release rate should at some stage become sufficiently high J06 57 so as to exceed the natural heat transfer rates (conduction, J06 58 Newtonian cooling, etc.). The evolution of self-heating in an J06 59 experiment can be followed using a coated fine-wire thermocouple J06 60 positioned at some point within the reacting gas (usually close to J06 61 the centre). With a reference junction on the outside wall of the J06 62 vessel, the thermocouple records directly the temperature J06 63 difference or extent of self-heating <*_>DELTA<*/>T. Figure 9.1(a) J06 64 shows a typical trace appropriate to a relatively low partial J06 65 pressure of the reactant or a low ambient temperature. The reacting J06 66 gas temperature (i.e. that inside the vessel) here increases above J06 67 that of the surroundings, but the temperature excess remains small J06 68 - the maximum <*_>DELTA<*/>T in this particular example is about 8 J06 69 K. The corresponding pressure and mass spectrometer records show a J06 70 steady consumption of the reactant. Because of the self-heating, J06 71 the reaction rate initially increases with increasing extent of J06 72 reaction - the indication of an acceleratory process, here through J06 73 thermal feedback.

J06 74 The maximum temperature excess observed increases if either the J06 75 ambient temperature of the initial partial pressure of the DTBP are J06 76 increased (both of these increase the rate of this first-order J06 77 process with a positive activation energy). For small changes, the J06 78 response changes smoothly, but if the increase in Ta J06 79 or pDTBP is to<&|>sic! large, there is a qualitative J06 80 change in the reaction evolution. Figure 9.1(b) shows a typical J06 81 response for a 'thermal explosion'. There is an induction period in J06 82 which the temperature excess increases to a value not much J06 83 different from that observed for the previous subcritical J06 84 behaviour, but this is now followed by a rapid acceleration during J06 85 which the gas temperature increases swiftly to a high transient J06 86 excess. There is virtually complete removal of DTBP during this J06 87 ignition pulse and the temperature then falls back to ambient.

J06 88 The conditions separating these two forms of response J06 89 (subcritical and supercritical) give rise to a sharp J06 90 p-Ta ignition boundary as shown in Fig. 9.2. The J06 91 exact location of this limit depends on many factors, such as the J06 92 vessel size and geometry and the presence or absence of J06 93 O2 and inert diluents (Egieban et al. 1982; J06 94 Griffiths and Singh 1982, Griffiths and Mullins 1984), but can be J06 95 relatively successfully predicted from classical thermal explosion J06 96 theory (Boddington et al. 1977, 1982a, b). The J06 97 length of the induction period can also be estimated, and expressed J06 98 in terms of the 'degree of supercriticality'. This latter can be J06 99 represented as the difference between the actual operating pressure J06 100 and the corresponding 'critical' pressure (i.e. the pressure on the J06 101 limit) for the operating ambient temperature. Then

J06 102 graphs&captions

J06 103 formula

J06 104 i.e. the period gets shorter as the system moves further into J06 105 the ignition region above the limit, but can become (arbitrarily) J06 106 long close to the limit.

J06 107 The application of such theories and interpretations implicitly J06 108 requires that the reactants should be admitted to the vessel in as J06 109 short a time as possible (preferably instantaneously) and the above J06 110 experiments come close to that.

J06 111 Of more interest in the present context, however, is an J06 112 experimental situation in which the admission process is not fast, J06 113 but becomes comparable or even slower than the reaction timescale. J06 114 Griffiths et al. (1988; Gray and Griffiths 1989) employed J06 115 a reactor into which the reactants were admitted via a capillary J06 116 tube. In this way it was possible to obtain a series of J06 117 "repetitive thermal explosions". Example time J06 118 series are shown in Fig. 9.3.

J06 119 A relatively simple explanation for the oscillatory waveforms J06 120 observed under these conditions of gradual admission can be given. J06 121 As the gas enters the vessel at some fixed ambient temperature, the J06 122 partial pressure of the reactant increases and so we move up a J06 123 vertical line in the pDTBP-Ta parameter plane J06 124 (the ignition diagram, Fig. 9.2). Once the pressure increases J06 125 beyond the critical pressure for that ambient temperature there J06 126 will be a thermal ignition (with perhaps some induction period). J06 127 This ignition decreases the partial pressure of the reactant below J06 128 the critical pressure (and probably close to zero), although the J06 129 total pressure will actually increase. Continued inflow of fresh J06 130 reactant on the slower timescale through the capillary increases J06 131 pDTBP again, and a second ignition can occur if the J06 132 critical pressure is reached again. This process may repeat a J06 133 number of times until the pressures inside the vessel and in the J06 134 reservoir (i.e. at each end of the capillary) become equal.

J06 135 This simple interpretation is of some value, but does not deal J06 136 too easily with all the experimental observations. For instance J06 137 there are both upper and lower limits to the temperature for which J06 138 multiple ignitions (oscillations) are found for any given final J06 139 pressure and inflow characteristics. Equivalently, there are upper J06 140 and lower final pressure limits for oscillations at any fixed J06 141 ambient temperature. The lower limits correspond simply to J06 142 conditions for which the system does not reach the ignition limit. J06 143 The upper limits are less intuitive, but relate to the matching of J06 144 inflow and the supercritical reaction rate, so the ignition becomes J06 145 sustained and steady rather than quenching due to reaction J06 146 consumption after each excursion.

J06 147 Fortunately there is a simple quantitative model for this J06 148 system - one we have already seen. The experiments described J06 149 constitute a realization of the Salnikov scheme of Section 3.5, J06 150 with some minor modification. In particular, the role of the first J06 151 step formula, which was taken to be a chemical reaction J06 152 with zero exothermicity and no activation energy, is now played by J06 153 the continuous slow inflow. The DTBP decomposition reaction then J06 154 provides the exothermic first-order step formula. The train J06 155 of oscillations must now be finite, as the inflow only continues J06 156 for a finite time and is decreasing in magnitude as the pressure J06 157 difference decreases. Numerically computed traces using the known J06 158 thermokinetic parameters are shown in Fig. 9.4 and successfully J06 159 reproduce the experiments.

J06 160 This design of transient oscillations can clearly be extended J06 161 to many other exothermic processes. Coppersthwaite and Griffiths J06 162 (Coppersthwaite 1990) have produced similar behaviour in the J06 163 H2 + Cl2 system. As in the above case, the J06 164 oscillatory excursions can lead to the attainment of dramatically J06 165 higher temperatures than those that occur in neighbouring J06 166 pseudo-steady-state responses and so there may be considerable J06 167 hazards associated with the onset of this mode of reaction.

J06 168 9.2. Clock reactions and the pic J06 169 d'arr<*_>e-circ<*/>t

J06 170 The idea of a thermal explosion as a fast-timescale clock J06 171 reaction may seem to push the unified view of solution-phase and J06 172 gas-phase non-linear behaviour to an extreme. There is, however, a J06 173 second situation in which this analogy is clearly justified to the J06 174 full. The oxidation of hydrocarbons will feature strongly in later J06 175 sections of this chapter, but here we can consider one aspect - the J06 176 evolution of reaction for relatively fuel-rich mixtures. Snee and J06 177 Griffiths (1989) studied the oxidation of cyclohexane under such J06 178 conditions, at atmospheric pressure in a wide range of vessel J06 179 sizes. Their investigation was primarily concerned with J06 180 investigating 'minimum ignition temperatures' - the lowest J06 181 temperature of a preheated flask for which ignition of a 1 J06 182 cm3 volume of cyclohexane would occur after injection of J06 183 the fuel. More standard kinetic studies showed that the reaction J06 184 could exhibit 'quadratic auto-catalysis' in the absence of J06 185 any self-heating, i.e. the dependence of the reaction rate on the J06 186 extent of reaction (measured in terms of the limited oxygen J06 187 concentration) has a simple parabolic form as shown in Fig. 9.5. In J06 188 the presence of self-heating which inevitably accompanied reaction J06 189 at higher temperatures, this parabolic dependence became skewed, as J06 190 is also shown in the figure.

J06 191 graph&caption

J06 192 Griffiths and Phillips (1989) found similar autocatalysis in J06 193 the oxidation of n-butane, again under fuel-rich conditions J06 194 (formula) at reduced pressure (500 Torr) over the range J06 195 589-600 K. The fuel-air mixture is admitted to a heated vessel and J06 196 the reaction can be monitored by a mass spectrometer (m/e = 32 for J06 197 O2), photomultiplier (chemiluminescence from J06 198 electronically excited formaldehyde CH2O*) and a J06 199 fine-wire thermocouple as described above. Figure 9.6 shows four J06 200 different O2 concentration time series for temperatures J06 201 within this range. With Ta = 589 K, no self-heating J06 202 is observed at any stage during the reaction. The time trace shows J06 203 a classic autocatalytic or clock reaction form, with a long J06 204 induction period followed by an accelerating rate and then slowing J06 205 due to reactant consumption. A plot of rate (i.e. slope of the time J06 206 trace) against the extent of conversion is again of simple J06 207 parabolic form appropriate to quadratic autocatalysis. At higher J06 208 temperatures, the induction periods shorten, and the increase in J06 209 magnitude of the rate at the end of this period becomes more J06 210 marked. Significant transient temperature rises (self-heating) J06 211 accompany the maximum reaction rate, up to 15 K for J06 212 Ta = 600 K. An additional feature here is that the J06 213 reaction produces a chemiluminescent intermediate CH2O* J06 214 at appreciable concentrations during the period of high rate and J06 215 this emission can be observed just before the reaction stops due to J06 216 (virtually ) complete O2 consumption. This phenomenon has J06 217 been termed the pic d'arr<*_>e-circ<*/>t J06 218 (D<*_>e-acute<*/>chaux and Lucquin 1968, 1971; J06 219 D<*_>e-acute<*/>chaux et al. 1968).

J06 220 graph&caption

J06 221 Griffiths and Phillips were able to match the form of their J06 222 experimental traces using a simple model based on quadratic J06 223 autocatalysis and self-heating (Kay et al. 1989). They J06 224 included a modification that allowed a surface-controlled J06 225 termination step, reflecting the sensitivity of their results to J06 226 pretreatment and ageing of their vessel. The model, then, has the J06 227 form

J06 228 formulae

J06 229 along with the appropriate heat-balance equation.

J06 230 J07 1 <#FLOB:J07\>Also present are the exclusively deep-sea J07 2 Limopsidae (Fig. 5.19), along with deep-sea species of the mytilid J07 3 genus Dacrydium (Fig. 10.2) and species of the families J07 4 Pectinidae and Arcidae. All are epifaunal, probably J07 5 suspension-feeding forms. One curious feature of deep-sea J07 6 suspension feeders seems to be that, although the soft parts may be J07 7 much reduced, the shell is proportionately much less reduced in J07 8 size (Fig. 5.20), possibly to enhance protection from predation at J07 9 minimal metabolic cost. As a result they include some of the J07 10 largest known, non-vent bivalves of the abyss (Knudsen, 1979).

J07 11 The superfamily Galeommatacea in shallow water are known to J07 12 include many small bivalves that live commensally with a variety of J07 13 larger infaunal invertebrates, particularly with echinoderms. At J07 14 least two unrelated species of this small-bodied group are known J07 15 from the deep sea: Montacuta (Axinodon) symmetros lives J07 16 with a byssal attachment to the spines of the echinoid J07 17 Pourtalesia (Bouchet & War<*_>e-acute<*/>n, 1979a) while J07 18 Galatheavalve holothuriae, collected from cavities in the J07 19 skin of the elasipod holothurian Psychropotes, has a J07 20 completely internal shell (Knudsen, 1970).

J07 21 There are few bivalves in the deep sea with truly world-wide J07 22 distributions; most are restricted to one or more adjacent basins J07 23 with morphological differences in the populations suggesting a J07 24 degree of ongoing speciation through isolation, whilst, even in the J07 25 cosmopolitan species Malletia cuneata (Fig. 5.21), subtle J07 26 changes in shell shape can be discerned (Sanders & Allen, 1985). J07 27 Although vertically segregated populations can be recognized J07 28 amongst the bivalves, overall, the degree of restriction in J07 29 vertical range, like the vertical range of prochaetodermatid J07 30 Aplacophora (see below), seems to be positively correlated to the J07 31 horizontal range displayed by species (Scheltema, 1985).

J07 32 figures&captions

J07 33 GASTROPODA J07 34 The class Gastropoda encompasses a much greater variety in form J07 35 and habit than bivalves. Most snail-like species with coiled shells J07 36 are thought to be motile carnivores, probably feeding on polychaete J07 37 worms and bivalves, or are ectoparasites, often on echinoderms J07 38 (Fig. 5.22). However, the limpet-like shape also occurs mainly as J07 39 opportunistic scavengers. As in shallow water, gastropods are J07 40 probably at least as diverse as any other molluscan group.

J07 41 From the Gay Head-Bermuda Transect, Rex (1976) identified 93 J07 42 species of shelled marine snails (subclass Prosobranchia) and 30 J07 43 species of the more slug-like subclass Opisthobranchia which have a J07 44 reduced shell, or lack it completely as in sea slugs (order J07 45 Nudibranchia); all with a wide range in feeding type. Deposit J07 46 feeders included 20 species amongst the more primitive prosobranch J07 47 families in the order Archaeogastropoda, 15 species of more J07 48 primitive families of the very diverse order Mesogastropoda (which J07 49 show particularly impressive radiation in the tropics) and five J07 50 species of the more primitive order of opisthobranchs, the J07 51 Cephalaspidea which retain a reduced thin shell and move J07 52 semi-burrowed through the sediment. At least one species, the J07 53 deposit-feeding archaeogastropod Bathybembix aeola is J07 54 known to be selective in terms of particle size, suggesting J07 55 adoption of an energetically more efficient strategy of selection J07 56 of finer particles with a relatively greater microbe-coated surface J07 57 area (Hickman, 1981). Bivalve and polychaete predators include one J07 58 whelk-like mesogastropod; 46 neogastropods, including a variety of J07 59 small poison-secreting cone shells and larger whelk-like J07 60 scavenger/predators with a typical roving tentaculate nostril; 17 J07 61 cephalaspids belonging to the more primitive families Acteonidae J07 62 (which eat polychaetes), Ringiculidae and Retusidae (both of which J07 63 seem to eat only foraminifers) and three nudibranchs which, at J07 64 least in shallow water, may be highly specific in diet on colonial J07 65 coelenterates. Ectoparasites of echinoderms and anthozoans include J07 66 11 and 8 mesogastropod species in the families Eulimidae and J07 67 Epitonidae, respectively, the former of which have lost the jaws J07 68 and radula and pump out the body fluids of their sedentary J07 69 invertebrate hosts through a proboscis inserted into the body J07 70 cavity (Fig. 5.22), while the latter rasp off tissue fragments with J07 71 their radula; and five pyramidellids which, although with J07 72 normal-looking small, spired shells, possess elaborate mouthparts J07 73 that are modified for piercing and sucking, with a diet of liquid J07 74 protein that is pumped out of their living mollusc or polychaete J07 75 prey. Although these two groups are conspicuous, no one group can J07 76 be said to predominate as amongst bivalves.

J07 77 figure&caption

J07 78 The neogastropod family, Turridae, are toxoglossan J07 79 (poison-toothed) cone shells that have emerged as the single most J07 80 diverse group of carnivorous gastropods in the deep sea, with a J07 81 species richness, peaking in the bathyal, rivalling that of the J07 82 cone shells of tropical sand flats (Bouchet & War<*_>e-acute<*/>n, J07 83 1980). Turrids are highly specialized predators on polychaetes, J07 84 killing their prey with poison from salivary glands that is J07 85 injected through modified radular teeth. They are thought to have J07 86 developed complex patterns of resource partitioning amongst J07 87 co-occurring species (Hickman, 1984). Eulimids are one of J07 88 the most diverse families of deep-sea gastropods, possibly even J07 89 richer in species than the Turridae (War<*_>e-acute<*/>n, 1984; J07 90 Bouchet & War<*_>e-acute<*/>n, 1986). It seems very likely that the J07 91 high degree of dietary specialization or host specificity amongst J07 92 the Turridae and Eulimidae has led to the spectacular radiation of J07 93 these families in the deep sea. Other neogastropods in the deep sea J07 94 include an assemblage of mostly large-sized species, some of which J07 95 have a wide-spread distribution (Bouchet & J07 96 War<*_>e-acute<*/>n, 1985).

J07 97 Although most of the different functional types amongst J07 98 gastropods seem to correspond to taxonomic groupings, the J07 99 cocculiniform limpets appear to have converged from at least 15 J07 100 families (Hickman, 1983). These are epifaunal scavengers on rather J07 101 unusual substrates such as wood, squid beaks (which can form highly J07 102 localized and dense accumulations, see Belyaev, 1966), empty J07 103 Hyalinoecia tubes, shark or skate egg cases, and fish and J07 104 whale bones. Most of these have been recorded from eutrophic J07 105 trenches where woody plant detritus is common (George & Higgins, J07 106 1979; Wolff, 1979). Although only some of these species, mostly J07 107 belonging to the family Cocculinidae, are believed to ingest wood, J07 108 the remainder graze the cover of microbes. Other cocculinid limpets J07 109 have been found on waterlogged wood from oligotrophic abyssal J07 110 bottoms far from land where their opportunistic, wood-ingesting J07 111 lifestyle parallels that of the deep-sea xylophaginous bivalves J07 112 (Wolff, 1979).

J07 113 Although diverse in the abyss, gastropod diversity seems to J07 114 peak at mid-slope depths (Rex, 1973, 1976; Bouchet & J07 115 War<*_>e-acute<*/>n, 1980). The distribution of gastropods also J07 116 appear to be much more sharply zoned with depth than amongst J07 117 bivalves (Grassle, Sanders & W. Smith, 1979), with a particularly J07 118 marked compression in vertical range amongst predatory species J07 119 compared to deposit-feeding species on the slope (Rex, 1977). J07 120 Despite this, certain species, such as the unusually open-coiled J07 121 epitonid Eccliseogyra nitida (Fig. 5.23), are known to be J07 122 widespread, with distributions throughout the Atlantic (Rex & Boss, J07 123 1973).

J07 124 Although it is becoming apparent that many deep-sea gastropods J07 125 have an early development in the plankton (see Chapter 13), many of J07 126 them, certainly most of those living at bathyal depths, lay egg J07 127 capsules of characteristic form attached to hard substrata (Fig. J07 128 5.24) from which the young hatch directly.

J07 129 SCAPHOPODS J07 130 The other major group of molluscs in the deep sea, the J07 131 Scaphopoda, are far less well known. This might be because they J07 132 constitute less than 1% of the mollusc fauna in shallow water. But J07 133 in the deep sea this proportion is much increased (Clarke, 1962), J07 134 and, although species richness may be less than that of other J07 135 molluscan groups, they achieve population densities next only to J07 136 those of bivalves. Modern species of this group have an ancient J07 137 lineage going back at least to the Devonian. They live in J07 138 cylindrical, often tapering, tube-like shells (Fig. 5.25(a)) J07 139 and in shallow water are known to be expert burrowers in soft J07 140 sediments using an elongated plug-like foot with a complex J07 141 musculature. This is thrust down into the sediment, while the body J07 142 and shell are pulled down to it by muscular contraction in a manner J07 143 analogous to the burrowing movement, using the foot, of protobranch J07 144 bivalves. The head hears curious bunches of ciliated tentacles J07 145 terminating in a secretory bulb, the captaculum, which reach out to J07 146 collect particles such as foraminifers, to which they show highly J07 147 developed selectivity (see Davies, 1987, for references). These J07 148 prey are crushed using a radula structure similar to that of J07 149 gastropods. They are best known from species large enough to be J07 150 collected by trawls: Dentalium megathyrus being known J07 151 from 0.45-4.1 km depth in the E. Pacific measures up to 70 mm in J07 152 length (Wolff, 1961). This, with other mainly larger species, is J07 153 put into the order Dentalioida. These large species, at least, may J07 154 be active burrowers in the deep sea leaving a conspicuous track as J07 155 they move through the superficial sediment (Fig. 5.25(b)). The J07 156 other subdivision, the order Siphonodentalioida may contain the J07 157 majority of, mainly small-sized, species. On the basis of shell and J07 158 radula morphology and the anatomy of soft parts, Scarabino (1979) J07 159 has separated 59 mostly small-sized species from epibenthic sled J07 160 samples taken from 0.3-6 km depth all over the Atlantic.

J07 161 figure&caption

J07 162 APLACOPHORA J07 163 Although still poorly known, the worm-like Aplacophora are J07 164 ubiquitous in the deep sea, occurring down to hadal depths. They J07 165 include two divisions, the Neomeniomorpha and Chaetodermomorpha. J07 166 Like the Scaphopoda, these forms are relatively abundant in the J07 167 deep-sea macro-fauna, yet are still very poorly known. J07 168 Species of the chaetodermomorph family Prochaetodermatidae are J07 169 found world-wide in the deep sea, with six species described from J07 170 the best-known area, the N. Atlantic (Scheltema, 1985a). They J07 171 are often numerous in deep-sea samples, being one of the J07 172 numerically dominant macrofaunal animals in quantitative samples in J07 173 the N.W. Atlantic slope and the Aleutian Trench (Jumars & Hessler, J07 174 1976; Scheltema, 1981; Maciolek et al., 1987a, J07 175 b), a density of 247 m-2 for the species J07 176 Prochaetoderma yongei (Fig. 5.26) being recorded at 3.64 J07 177 km depth off New England (Grassle & Morse-Proteous, 1987). Both J07 178 divisions have a skin covered with protective calcareous spicules J07 179 and possess simplified guts and a radula in the buccal cavity. The J07 180 Chaetodermomorpha probably make shallow burrows in the sediment J07 181 with the posterior end bearing the gills at the sediment surface J07 182 (Fig. 5.26). The Neomeniomorpha have the simplest gut, and most J07 183 species feed suctorially on the tissues of gorgonians and hydroids; J07 184 while most species of Chaetodermomorpha feed on small organisms J07 185 such as foraminifers and small macro-fauna. Scheltema J07 186 (1981) argued that large population densities of J07 187 Prochaetoderma may be attributable in part to their J07 188 gastropod-like rasping mouth-parts allowing a wide range of J07 189 particle sizes to be available to them.

J07 190 CHITONS AND MONOPLACOPHORANS

J07 191 Of the Polyplacophora, or chitons (coat-of-mail shells), only a J07 192 few species have been collected from the deep sea, these being J07 193 restricted to manganese nodules and pieces of wood, and belong to J07 194 the primitive genus Lepiodopleurus (Paul, 1976; Wolff, J07 195 1979).

J07 196 Discovery of living examples from 3.57 km depth in the Pacific J07 197 of an archaic molluscan group of Monoplacophora, which are related J07 198 to Cambrian and Silurian fossils and retain partial segmentation of J07 199 body organs (Fig. 5.27(a), (b)), was one of the great J07 200 discoveries of the 'Galathea' expedition (Lemche, 1957; Lemche & J07 201 Wingstrand, 1959). However, a related species dredged in 1869 has J07 202 only recently been recognized (War<*_>e-acute<*/>n, 1988). Although J07 203 having a limpet-like form, Neopilina has been photographed J07 204 ploughing through soft sediments, the elongated gut (Fig. J07 205 5.27(c)) suggesting that it subsists on the sparse and J07 206 refractory organic matter. Fragments of the xenophyophore J07 207 Stannophylum (see Fig. 5.37) have been found among the gut J07 208 contents. This, together with radula scraping marks on living J07 209 specimens of the xenophyophores, from the same locality, suggests J07 210 that N. galatheae preys on these giant protozoans J07 211 (Tendal, 1985b). Six genera of recent monoplacophorans have J07 212 now been described, the shell of the smallest, Micropilina J07 213 from about 0.9 km depth S. and W. of Iceland being only about 1 mm J07 214 in diameter (War<*_>e-acute<*/>n, 1989). The taxonomy, morphology J07 215 and relationships of the group as a whole have been recently J07 216 treated by Wingstrand (1985).

J07 217 figure&caption

J07 218 MEIOFAUNAL TAXA

J07 219 This category is much less well known than macrofauna. It J07 220 consists of both multicelled (metazoan) animals traditionally J07 221 regarded as 'meiofaunal' and larger single-celled protozoans, such J07 222 as Foraminifera.

J07 223 THE MULTICELLED TAXA: NEMATODA

J07 224 Although little is known of the taxonomic composition or mode J07 225 of life of the traditionally meiofaunal group, the Nematoda (thread J07 226 worms), they are probably by far the most numerous metazoans in all J07 227 marine soft bottoms. There are few data from the deep sea; but J07 228 there can be little doubt that this group is as important and J07 229 numerous in deep-ocean sediments as they are in shallow water (see J07 230 review by Heip, Vincx & Vranken, 1985). J07 231 J08 1 <#FLOB:J08\>The scintillation light was transmitted to J08 2 photomultipliers at the ends of the counters by wavelengths J08 3 shifting lightguides and a photon-counting system was used to J08 4 discriminate between neutrons and other events. A stack of eight of J08 5 these counters, around a central cavity, formed the multiplicity J08 6 detector.

J08 7 While the neutron counters were designed for a very specific J08 8 purpose, it is thought likely that high efficiency low background J08 9 counters of this type might find applications in other areas of J08 10 nuclear instrumentation. In particular, arrays of neutron counters J08 11 optimized for neutron multiplicity measurement have proved useful J08 12 in the non-destructive assay of fissile materials (Prosdocimi J08 13 1979).

J08 14 2. Design of the neutron multiplicity detector

J08 15 An experimental choice had to be made between designing a J08 16 detector that could detect a number of fast neutrons simultaneously J08 17 or designing a detector that first thermalized the neutrons in a J08 18 moderator and then detected the thermal neutrons over a period J08 19 determined by their diffusion time. Since the cross sections for J08 20 typical fast neutron induced reactions are orders of magnitude J08 21 lower than those of the corresponding thermal reactions, fast J08 22 neutron detectors have much lower detection efficiencies. While J08 23 they are more efficient, detectors based on thermalization can give J08 24 no information about the initial energy of the neutrons. For the J08 25 purposes of the current experiment, this was not considered a J08 26 disadvantage.

J08 27 The previously mentioned requirement to achieve high efficiency J08 28 and low background had also to be balanced against financial J08 29 considerations. Undoubtedly an array of high pressure 3He J08 30 proportional counters in a large volume of heavy water would prove J08 31 the most favourable arrangement, but the cost of such an array is J08 32 prohibitive. Using a hydrogenous moderator, rather than one based J08 33 on deuterium, enables the system to be much smaller but the J08 34 efficiency is then limited by the magnitude of the (n,p) capture J08 35 cross section. One design of neutron detector uses the 2.3 MeV J08 36 gamma ray from this reaction to register the neutron. Such an J08 37 arrangement, in which a liquid scintillator also acts as the J08 38 moderator, is simple to implement but the high background rate J08 39 prevents its use in many applications, including the present J08 40 one.

J08 41 A similar, though less severe, restriction also applies to J08 42 organic scintillators loaded with a small proportion of cadmium or J08 43 gadolinium. Tanks of gadolinium loaded liquid scintillator formed J08 44 the detectors used by Cheifetz et al (1972) and Becker J08 45 et al (1979). While extremely high neutron detection J08 46 efficiencies were achieved (65% and 75% respectively), both these J08 47 detectors were sensitive to environmental background gamma J08 48 radiation. To reduce this sensitivity, Becker et al J08 49 (1979) required the detection of fission fragments in coincidence J08 50 with the neutrons before an event was registered. Consequently J08 51 their detector could only be used with extremely thin samples from J08 52 which the fission fragments could escape.

J08 53 Because of the above considerations, other neutron multiplicity J08 54 detectors devised for superheavy element searches have used an J08 55 array of 3He proportional counters in a matrix of J08 56 hydrogenous moderator (Macklin et al 1972, Ter-Akop'yan J08 57 et al 1981) and this approach has also been adopted for J08 58 fissile material assay (Prosdocimi 1979). The neutron multiplicity J08 59 detector described here differs from all the previous detectors in J08 60 that the detection of the neutrons is achieved by the scintillation J08 61 of a lithium fluoride and zinc sulphide mixture and that the J08 62 hydrogenous moderator is a structural component of the individual J08 63 neutron counters.

J08 64 The neutron capture reaction in 6Li produces both a J08 65 triton and an alpha particle, with an energy release of 4.79 MeV. J08 66 Silver-activated zinc sulphide is well known as the most efficient J08 67 of all scintillators and consequently intimate mixtures of the two J08 68 materials coupled to photomultipliers are well established as cheap J08 69 and efficient slow neutron detectors. The main limitation of such J08 70 mixtures is that, as with other zinc sulphide scintillators, the J08 71 material is opaque to its own light and consequently can only be J08 72 used in thin layers. This renders it difficult to construct large J08 73 volume detectors. It was shown by Barton and Caines (1970) that J08 74 this limitation could be overcome by the use of wavelength shifting J08 75 lightguides. This technique is used again in the present detectors J08 76 with a slightly different geometrical arrangement.

J08 77 3. Construction of the neutron counters

J08 78 The neutron counters had an overall sandwich construction J08 79 consisting of thin layers of zinc sulphide scintillator attached to J08 80 the slabs of polypropylene moderator that provided the major part J08 81 of the moderating material. These slabs were interleaved with the J08 82 wavelength shifting lightguides, leaving a narrow air gap between J08 83 the neutron scintillator layers and the lightguides. The light J08 84 emitted by the scintillator traversed the lightguides causing J08 85 fluorescence and some of the re-emitted light was then carried J08 86 along the guides. Both ends of the lightguides were viewed by J08 87 photomultipliers, there being a second air gap between the J08 88 lightguides and the faceplate of each photomultiplier. The whole J08 89 counter was encased in a light-tight black Perspex tube of square J08 90 cross section. The photomultipliers were held in head units J08 91 constructed from diecast boxes bolted to the ends of the Perspex J08 92 casing. A longitudinal section of one of the counters is shown in J08 93 figure 1, while figure 2 shows a cross section.

J08 94 The zinc sulphide chosen had a larger than usual proportion of J08 95 nickel added to reduce the fraction of light in the tail of the J08 96 scintillation pulse (Barton and Ranby 1977). It was prepared by J08 97 Thorn Research and Engineering Laboratories as HS338A. The lithium J08 98 fluoride was material that had been enriched to 96% in J08 99 6Li by Harwell Stable Isotopes Group. The grains of both J08 100 powders had a most probable size of about 5 <*_>mu<*/>m but the J08 101 range of sizes was not closely controlled and included grains from J08 102 less than 1 to 10 <*_>mu<*/>m. The two powders were thoroughly J08 103 mixed together, then added to Dow Corning Sylgard 184. This J08 104 material is a thermosetting silicone resin which is rubbery and J08 105 optically transparent when set. The mass proportions were J08 106 ZnS:LiF:resin:hardener = 4:2:2.7:0.3. The resulting mixture was J08 107 used to coat sheets of highly reflective aluminized Melinex.

J08 108 Considerable effort was expended in finding a suitable J08 109 technique to manufacture large areas of the scintillator with a J08 110 uniform thickness. The technique finally adopted was to spread the J08 111 mixture which, before setting, had the consistency of thick paint. J08 112 A sheet of plate glass was mounted horizontally on the bed of a J08 113 milling machine to provide a flat working surface and the sheet of J08 114 Melinex to be coated was placed on top of this. The scintillator J08 115 mixture was loaded into a slotted hopper which was mounted on the J08 116 transverse of the machine and this hopper was pulled slowly across J08 117 the Melinex by moving the bed below, spreading the mixture as it J08 118 went. The intention was to provide layers of thickness J08 119 100<*_>unch<*/>10 <*_>mu<*/>m but in practice the variation, as J08 120 measured by a gamma-ray absorption technique, was slightly more J08 121 than this. The only other disadvantage of this technique was that J08 122 the wastage of material was about 25%.

J08 123 figures&captions

J08 124 The incorporation of such a high proportion of foreign chemical J08 125 material in the silicone resin increased its setting time J08 126 considerably and the surface of the layers remained permanently J08 127 sticky. When the first prototype counters were tested in an J08 128 underground laboratory, their background counting rates were J08 129 observed to increase slowly from their initial values. Tests showed J08 130 that this effect was due to radon decay products, either as atoms J08 131 or attached to dust particles, settling on the surfaces of the J08 132 scintillator layers. The problem was solved by applying a second J08 133 coating of pure silicone resin over the scintillator. This second J08 134 layer, which set perfectly, was of sufficient thickness to prevent J08 135 the alpha particles from any subsequently deposited atoms from J08 136 reaching the scintillator but did not interfere with the light J08 137 leaving the scintillator.

J08 138 The scintillator layers, on their Melinex backing sheets, were J08 139 attached to the polypropylene blocks by double sided adhesive tape. J08 140 The wavelength shifting lightguides were fabricated from J08 141 R<*_>o-umlaut<*/>hm and Haas Plexiglas GS2025, a J08 142 polymethyl-methacrylate glass containing the fluorescent J08 143 compound BBQ. All the surfaces were polished and the guides were J08 144 supported away from the sides of the Perspex casing of the counter J08 145 using conically tipped screws. The array of guides was viewed from J08 146 each end by a 130 mm photomultiplier with aluminium foil providing J08 147 a simple lightguide. The photomultipliers were EMI type 9791KA J08 148 tubes. They were individually selected to have a low afterpulsing J08 149 rate, specified as less than 400 ion-induced pulses, of magnitude J08 150 equivalent to more than 2.5 single photoelectrons, for every J08 151 106 photoelectrons.

J08 152 figure&caption

J08 153 The external dimensions of a complete counter, excluding the J08 154 photomultiplier head-unit were 900 mm x 144 mm x 144 mm. A stack of J08 155 eight of these counters could be arranged as shown in figure 3. The J08 156 head-units have been shown on one counter only and omitted on the J08 157 others for clarity. This arrangement of counters provided a central J08 158 cavity for samples with the same dimensions as a single counter J08 159 (volume 18.61), which could adequately contain several tens of J08 160 kilograms of sample material. The central detector in the top row, J08 161 immediately above the cavity, was kept in position by two strips of J08 162 aluminium sheet, each a few centimetres wide, below it. These J08 163 strips were placed close to the two ends of the stack across the J08 164 two detectors which comprised the second row and before the top row J08 165 was added.

J08 166 4. Output pulse processing and event recording

J08 167 The pulse-height distribution of neutron-induced scintillation J08 168 in lithium loaded zinc sulphide scintillator is such that simple J08 169 pulse-height discrimination between neutrons and noise signals is J08 170 not feasible. This is a consequence of the particular time J08 171 structure of the scintillation in zinc sulphide and of the wide J08 172 range of pulse heights resulting from both the heterogenic nature J08 173 of the photon source and variations in light collection and J08 174 conversion. Many types of pulse shape discrimination circuit have J08 175 been reported but none seemed suitable for the rather stringent J08 176 requirements of the present experiment. In addition to the desire J08 177 to maximize efficiency, there were also overriding requirements J08 178 that photomultiplier and electronic noise should not contribute to J08 179 the background counting rate and that the system should J08 180 discriminate against signals due to the passage of one or more J08 181 lightly ionizing particles.

J08 182 The method finally adopted was a digital pulse discrimination J08 183 system with two distinct stages. The first of these was a photon J08 184 counting discriminator, based on earlier work by Caines (1972) and J08 185 Davidson (1977). A block diagram of it is shown in figure 4; exact J08 186 details of the circuits can be found in McMillan (1990). Firstly, J08 187 the output of each photomultiplier was converted to a sequence of J08 188 uniform pulses by a simple monostable circuit based on a fast J08 189 voltage-comparator and having a pulse resolution of 40 ns. The J08 190 input level of this monostable was set so that it was triggered by J08 191 every photoelectron. Noise and most minimum ionizing events were J08 192 rejected by requiring that there was an initiating prompt J08 193 coincidence between signals from the photomultipliers at both ends J08 194 of the counter and that there was then a minimum of two other J08 195 pulses from each end within the next 20 <*_>mu<*/>s. After this J08 196 period there was a dead-time of a further 20 <*_>mu<*/>s to ensure J08 197 that single neutrons could not masquerade as doubles. If the above J08 198 criteria were met, the circuit produced a 'neutron output' signal. J08 199 Additionally, a pulse-train output was generated by combining the J08 200 outputs from the discriminators associated with the two J08 201 photomultipliers in an OR gate. The exact timing and selection J08 202 criteria chosen were inevitably a compromise between rejecting some J08 203 neutron events and accepting some non-neutron events. The J08 204 first stage, as described and with the values indicated, provided J08 205 neutron discrimination sufficient for most purposes. The efficiency J08 206 and neutron lifetime measurements described later were performed J08 207 using only this first stage. However, the fact that it failed to J08 208 reject certain minimum ionizing events, particularly small J08 209 electromagnetic showers, meant that for low background multiplicity J08 210 work an additional level of discrimination was necessary.

J08 211 figure&caption

J08 212 This second stage of pulse processing was not so much an J08 213 electronic discrimination system as a method of recording the J08 214 signals from the counters in sufficient detail that minimum J08 215 ionizing events could subsequently be recognised by eye in the data J08 216 set and rejected accordingly. J08 217 J09 1 <#FLOB:J09\>Proteolysis of factor VIII heavy chain J09 2 polypeptides in plasma and concentrates

J09 3 G. Kemball-Cook, S. J. Edwards and T. W. Barrowcliffe

J09 4 Summary. Factor VIII heavy chain (FVIII HC) polypeptides J09 5 have been studied in both normal plasma and FVIII concentrates on J09 6 exposure to three coagulation proteases.

J09 7 FVIII samples were incubated with labelled affinity-purified J09 8 anti-FVIII Fab' fragments, immunocomplexes formed were visualized J09 9 by autoradiography after sodium dodecyl sulphate polyacrylamide gel J09 10 electrophoresis (SDS-PAGE), and apparent relative molecular masses J09 11 (Mr) of each band assigned. FVIII HC polypeptides were J09 12 detected in all types of samples, including plasma, without further J09 13 purification.

J09 14 Normal plasma contained a range of polypeptides with the J09 15 largest dominant band at a net apparent Mr of 250-300 kD, J09 16 and the smallest at 80-90 kD; the bands visualized correspond to J09 17 the 90-120 kD HC species seen on conventional analysis of purified J09 18 FVIII. No bands were produced from samples of haemophilic J09 19 plasma.

J09 20 Treatment of plasma or FVIII concentrate with low J09 21 concentrations (1 IU/ml) of thrombin removed the 250-300 kD and J09 22 other intermediate bands, intensified then removed the 80-90 kD J09 23 polypeptide and produced a band at 40-50 kD. Thrombin-associated J09 24 rise and fall in FVIII clotting activity by one-stage assay J09 25 correlated with intensity of the 80-90 kD polypeptide. A J09 26 polypeptide of Mr 40-50 kD was also produced after J09 27 incubation with activated factor X: activated factor VII plus J09 28 thromboplastin had no effect on HC structure.

J09 29 FVIII polypeptides were visualized in prothrombin complex J09 30 concentrates, with a more degraded profile seen in a deliberately J09 31 'activated' product.

J09 32 In recent years the primary structure of factor VIII (FVIII) J09 33 has been elucidated by purification of the protein (Fulcher & J09 34 Zimmerman, 1982; Rotblat et al, 1985) and cloning of the J09 35 gene (Toole et al, 1984; Vehar et al, 1984). In J09 36 addition, recombinant FVIII has been produced using genetic J09 37 engineering techniques (Wood et al, 1984; Truett et J09 38 al, 1985). Despite these great advances, the exact nature of J09 39 the FVIII species active in coagulation is subject to J09 40 conjecture.

J09 41 Proteolysis of highly-purified FVIII by coagulation J09 42 pro-teases, especially thrombin, activated factor X (FXa) J09 43 and activated protein C (APC), closely define its biological J09 44 activity (Eaton et al, 1986). Specific cleavages made by J09 45 thrombin or FXa in the heavy chain (HC) and light chain (LC) are J09 46 thought to be essential to produce the active form which takes part J09 47 in the activation of FXa as a cofactor, while APC and FXa can make J09 48 inactivating cleavages in the heavy chain. Also, the cleaved J09 49 fragments of HC and LC must be in association [probably via J09 50 divalent cations] in order to retain activity (Fulcher et J09 51 al, 1985; Andersson et al, 1986; Fay et J09 52 al, 1986).

J09 53 The structure of thrombin-activated human FVIII has been J09 54 postulated to be a 51 kD/73 kD heterodimer derived from the HC and J09 55 LC respectively, with a conformational relaxation possibly J09 56 responsible for loss of activity on further incubation (Fay, 1987). J09 57 However, a stable form of porcine activated FVIII consisting of a J09 58 heterotrimer has also been reported (Lollar & Parker, 1989).

J09 59 All these studies on polypeptide cleavage of plasma-derived J09 60 FVIII have been performed using very highly-purified FVIII. J09 61 However, the FVIII molecule is very sensitive to proteolysis during J09 62 isolation, for example during the manufacture of FVIII concentrates J09 63 (Barrowcliffe et al, 1986; Kemball-Cook et al, J09 64 1990). Thus studies using purified FVIII may not represent the J09 65 proteolysis of the molecule in, for example, plasma or low purity J09 66 FVIII concentrates.

J09 67 In a recent paper (Kemball-Cook et al, 1990), we J09 68 demonstrated a modified sodium dodecyl sulphate polyacrylamide gel J09 69 electrophoresis (SDS-PAGE) method for visualization of factor VIII J09 70 heavy chain (FVIII HC) polypeptides. This approach, based on that J09 71 first described by Weinstein et al (1981) enables FVIII J09 72 structure to be studied in a wide range of samples including plasma J09 73 without further purification. We have therefore used this technique J09 74 to study the proteolytic breakdown of FVIII HC in plasma and J09 75 concentrates when exposed to a range of coagulation enzymes. In J09 76 addition a correlation between HC structure and FVIII activity J09 77 (FVIII:C) following thrombin treatment has been made.

J09 78 MATERIALS AND METHODS

J09 79 125I-labelled specific anti-FVIII Fab' fragments were J09 80 prepared from inhibitor plasma CC8000 as previously described J09 81 (Kemball-Cook et al, 1990).

J09 82 Fresh human plasmas were obtained by clean venepuncture and J09 83 immediately anticoagulated by mixing 9:1 with 0.109 M J09 84 trisodium citrate.

J09 85 Severe haemophilic plasma was obtained frozen from Dr P. B. A. J09 86 Kernoff at the Royal Free Hospital, London.

J09 87 FVIII samples were prepared for SDS-PAGE by diluting to between J09 88 0.5 and 5 IU/ml FVIII clotting activity (FVIII:C), then 10 J09 89 <*_>mu<*/>l were mixed with 10 <*_>mu<*/>l of 125I-Fab' J09 90 (3000-10 000 cpm) and 4 <*_>mu<*/>l of 9% (w/v) PEG 4000 (Sigma, J09 91 Poole, Dorset) in the presence of protease inhibitors (see below), J09 92 and incubated in a closed plastic tube at 37<*_>degree<*/>C for 2 h J09 93 to allow formation of immunocomplexes. 20 <*_>mu<*/>l of sample J09 94 buffer (62 mM Tris-HCl-10% glycerol-2% SDS-0.001% bromophenol J09 95 blue, pH 6.8) were added and incubation continued for a further 30 J09 96 min. after which samples (30-35 <*_>mu<*/>l) were loaded onto the J09 97 gel.

J09 98 Inhibitors added to the 125I-Fab' to terminate J09 99 reactions with proteases were recombinant desulphatohirudin (CGP J09 100 39393, kind gift of Dr R. B. Wallis, Ciba-Geigy, Horsham, Sussex), J09 101 100 <*_>mu<*/>g/ml; I-2581 (Kabi, Uxbridge, Middlesex), 25 J09 102 <*_>mu<*/>M; and PMSF (Sigma, Poole, Dorset), 10mM.

J09 103 Factor VIII concentrates were type 8Y (Bio Products Laboratory, J09 104 Elstree, Hertfordshire); Profilate HT solvent-treated J09 105 (Alpha Therapeutics, Thetford, Norfolk) and Hemofil M (Baxter J09 106 Travenol, Thetford, Norfolk). The BPL and Alpha concentrates are J09 107 low purity conventionally-produced materials (specific activity J09 108 approximately 5 IU FVIII:C/mg protein), while Hemofil M is made J09 109 using immobilized monoclonal antibodies to FVIII and is of high J09 110 specific activity (>1000 IU/mg) before addition of albumin J09 111 stabilizer.

J09 112 Prothrombin complex concentrates: non-activated J09 113 (Pro-thromplex) and activated (Feiba) materials were J09 114 supplied by Immuno (Sevenoaks, Kent).

J09 115 Purfified human thrombin was NIBSC reagent 82/570, specific J09 116 activity 700 IU/mg protein.

J09 117 Purified human factor X (FX) was prepared by the method of Modi J09 118 et al (1984) using factor IX concentrate (Bio Products J09 119 Laboratory, Elstree, Hertfordshire) as starting material. Sulphated J09 120 dextran was produced according to Miletich et al, J09 121 (1980).

J09 122 RVV-X, the FX-activating protein from Russell's Viper venom J09 123 (RVV), was purified from crude RVV (Diagen) by ion exchange J09 124 chromatography on QAE-Sephadex (Pharmacia).

J09 125 Human activated FX (FXa) was prepared in situ by J09 126 mixing purified FX with RVV-X protein at a ratio of 10:1 in the J09 127 presence of 5 mM CaCl2 for 15 min: under these J09 128 conditions FX was fully activated.

J09 129 Recombinant FVIIa was kindly supplied by Novo Nordisk, J09 130 Bagsvaerd, Denmark.

J09 131 Rabbit brain thromboplastin was a product of Manchester J09 132 Comparative Reagents Ltd, Stockport, Cheshire.

J09 133 Thrombin treatment of FVIII samples: samples were mixed with J09 134 thrombin and incubated at the temperature indicated. Subsamples J09 135 removed for SDS-PAGE were added immediately to J09 136 125I-Fab'/PEG in the presence of protease inhibitors.

J09 137 One-stage assays on subsamples from thrombin activation of J09 138 FVIII were carried out as follows: 100 <*_>mu<*/>l of severe J09 139 haemophilic plasma were mixed with 100 <*_>mu<*/>l kaolin (5 mg/ml) J09 140 and incubated for 9 min: 100 <*_>mu<*/>l of phospholipid (NIBSC J09 141 86/516 bovine brain extract, 10 <*_>mu<*/>g/ml) and 100 <*_>mu<*/>l J09 142 of FVIII sample were added, then at 10 min the mixture was J09 143 recalcified with 100 <*_>mu<*/>l 33 mM CaCl2 and the J09 144 clotting time recorded. The percentage activation was calculated by J09 145 reference to a standard curve constructed from the sample prior to J09 146 thrombin addition.

J09 147 FXa proteolysis of FVIII was studied by incubating very high J09 148 purity (VHP) FVIII concentrate (final concentration 5 IU/ml J09 149 FVIII:C) with a mixture of purified human FX (10<*_>mu<*/>g/ml), J09 150 RVV-X protein (1 <*_>mu<*/>g/ml) and CaCl2 (5 mM) J09 151 for 15 min at 37<*_>degree<*/>C before addition of labelled J09 152 antibody: the ratio of FX to RVV-X was selected to give full J09 153 activation of FX in 10-15 min. Recombinant hirudin (100 J09 154 <*_>mu<*/>g/ml) was present throughout to exclude the action of any J09 155 trace thrombin.

J09 156 Factor VIIa (FVIIa)/thromboplastin (TP): VHP FVIII concentrate J09 157 (2.5 IU/ml FVIII) was incubated with recombinant FVIIa (27 U/ml J09 158 FVII:C) in the presence or absence of 10 mM CaCl2 J09 159 and TP (Manchester, diluted 1 in 10), before incubation at J09 160 37<*_>degree<*/>C for 90 min. This FVIIa/TP/CaCl2 mixture J09 161 clotted FVII-deficient plasma in less than 1 min.

J09 162 Slab gel preparation, electrophoresis and autoradiography were J09 163 carried out as previously described (Kemball-Cook et al, J09 164 1990).

J09 165 Molecular weight markers for use in non-reduced Laemmli J09 166 SDS-PAGE were HMW kit (Pharmacia, Milton Keynes, Buckinghamshire) J09 167 and SDS-6H (Sigma, Poole, Dorset).

J09 168 Assignment of apparent relative molecular masses J09 169 (Mr): calculation of approximate Mr was by J09 170 measurement of polypeptide migration distance on the autoradiograph J09 171 and comparison with molecular weight markers on the dried Coomassie J09 172 blue-stained gels, with subtraction of 50 kD (=Mr of one J09 173 Fab' molecule) to give the apparent net mass of the polypeptide.

J09 174 Quantitative densitometry was carried out using an LKB J09 175 Ultroscan Laser Densitometer.

J09 176 RESULTS

J09 177 Visualization of FVIII in plasma and FVIII concentrates

J09 178 When 125I-labelled A/FVIII Fab' was incubated with J09 179 either fresh normal plasma or FVIII concentrate and J09 180 electrophoresed, autoradiography revealed a range of bands with J09 181 apparent Mr between 80-90 kD and 250-300 kD: no bands J09 182 were seen when severe haemophilic plasma was used (Kemball-Cook J09 183 et al, 1990).

J09 184 Effect of thrombin

J09 185 Fig 1 shows the effect of incubating either frozen normal J09 186 pooled plasma or, in a separate experiment, a sample of one type of J09 187 conventional low purity FVIII concentrate, with human thrombin J09 188 (final concentration of 1 IU/ml) for 20 min at 37<*_>degree<*/>C. J09 189 In the plasma sample, the band at Mr 250-300 kD J09 190 disappeared and a degraded band became visible below Mr J09 191 80-90 kD at Mr 40-50 kD. The concentrate showed a loss of J09 192 virtually all the original pattern with production of one major J09 193 degradation band at Mr approximately 40-50 kD. Diffuse J09 194 staining seen at Mr above 300 kD is due to limited J09 195 self-association of Fab' monomers. As previously reported J09 196 (Barrowcliffe et al, 1986; Kemball-Cook et al, J09 197 1990), a low purity FVIII concentrate from one manufacturer J09 198 consistently showed a polypeptide at 40-50 kD even in the absence J09 199 of added thrombin. In order to determine whether this polypeptide J09 200 could be a degradation product formed by trace thrombin present J09 201 during concentrate manufacture, samples of a batch of this product J09 202 (here coded B) and another low purity FVIII concentrate (A) were J09 203 incubated with thrombin (final concentration 1 IU/ml) for 20 min at J09 204 37<*_>degree<*/>C to ensure complete thrombin-induced degradation. J09 205 The SDS-PAGE results obtained before and after thrombin treatment J09 206 are seen in Fig 2. Although the FVIII profile in the two J09 207 concentrates was quite different before thrombin treatment, J09 208 afterwards very similar patterns were seen in both materials, with J09 209 essentially one band only at Mr 40-50 kD. This band J09 210 co-migrated with the lowest Mr band present in product B J09 211 before proteolysis, indicating the possibility of partial thrombin J09 212 proteolysis of FVIII in this product during manufacture.

J09 213 figure&caption

J09 214 Thrombin cleavage and biological activity

J09 215 In addition, the changes in polypeptide pattern in the two J09 216 concentrates were correlated with the thrombin-induced rise and J09 217 subsequent decay of one-stage FVIII:C. Fig 3 presents the J09 218 polypeptide patterns obtained from subsamples taken at various J09 219 times from thrombin-FVIII activation mixtures: the conditions were J09 220 chosen to produce a relatively prolonged cleavage of FVIII. Both J09 221 concentrates showed similar degradation patterns, with immediate J09 222 disappearance of the Mr 250-300 kD band (initially J09 223 visible only in product A), intensification of the Mr J09 224 80-90 kD band and then its reduction simultaneously with appearance J09 225 or intensification of the Mr 40-50 kD band. Similar J09 226 patterns of FVIII degradation over time were seen with thrombin J09 227 treatment of fresh plasma (not shown).

J09 228 figure&caption

J09 229 The strengths of the Mr 250-300 kD, 80-90 kD and J09 230 40-50 kD HC bands seen in product A's pattern were quantitated by J09 231 densitometry and are compared with the one-stage FVIII:C in Fig 4. J09 232 The rise (to 350% of pre-incubation activity) and fall (to below J09 233 10%) in FVIII:C were closely paralleled by the appearance and J09 234 disappearance of the Mr 250-300 kD band, whereas neither J09 235 the Mr 40-50 kD nor the Mr 250-300 kD band J09 236 correlated with clotting activity. A similar correlation was seen J09 237 with thrombin treatment of product B (not shown).

J09 238 figure&caption

J09 239 Effect of FXa

J09 240 VHP FVIII concentrate (final concentration 5 IU/ml) was J09 241 incubated with a mixture of FX (10 <*_>mu<*/>g/ml), RVV-X (1 J09 242 <*_>mu<*/>g/ml) and CaCl2 (5 mM) for 15 min at J09 243 37<*_>degree<*/>C in the presence of excess recombinant hirudin, J09 244 before addition of labelled antibody.

J09 245 Fig 5 presents the polypeptide patterns obtained. Firstly, it J09 246 is apparent that before exposure to thrombin this product was J09 247 largely deficient in the larger Mr HC species, with most J09 248 staining in the Mr 80-90 kD position: secondly, neither J09 249 RVV-X alone nor FX alone had any effect, but FXa generated in J09 250 situ caused generation of a band at Mr 40-50 kD.

J09 251 J10 1 <#FLOB:J10\>Figure 8 shows the scatter plot of the turbulent J10 2 deposition velocity determined from each probe for cases 1 and 2. J10 3 The regression line was formula. The standard error of the J10 4 coefficient was 0.218, while the correlation coefficient was J10 5 0.80.

J10 6 graphs&captions

J10 7 (b) Turbulence data from the sonic anemometer

J10 8 To compare deposition velocities from a variety of sites and J10 9 canopies with different wind speeds, vt was J10 10 normalised with vm as suggested by Slinn (1982). For the J10 11 calculations presented here we have used the in-situ J10 12 measurements of u* and wind speed from the sonic J10 13 anemometer. These were compared with u* values obtained J10 14 from the profile mast which allowed values of the zero-plane J10 15 displacement and roughness length to be determined. This was done J10 16 to provide an estimate of the reliability of point estimates of J10 17 fluxes of such canopies. The profile information was also used to J10 18 show whether the turbulence was in local equilibrium with respect J10 19 to the canopy. This is important to determine if such measurements J10 20 are to be compared with model results of cloud deposition to hills J10 21 in complex terrain as discussed by Gallagher et al. (1990). As J10 22 discussed later, good agreement was found between the data from the J10 23 sonic and profile mast, where a value of d of approximately J10 24 2h/3, where d is the zero-plane displacement and h is the J10 25 canopy height, was used for computing vm from the J10 26 profile data. The relationship between d and h is J10 27 consistent with a general consensus from a large number of studies, J10 28 for example Jarvis et al. (1976); however, during one J10 29 case study, when a kink developed in the wind profile, the value of J10 30 d was uncertain and agreement between sonic and profile mast J10 31 data required a larger value of d to be used.

J10 32 During case 1 the mean friction speed was 0.37<*_>unch<*/>0.05 J10 33 m s-1, with a mean wind speed at the measurement height J10 34 (9.03m) of 2.73<*_>unch<*/>0.36 m s-1. This yielded a J10 35 value for the canopy roughness length, z0, of J10 36 0.30<*_>unch<*/>0.09 m (z0/h=0.072). For case 2 the J10 37 observed values were u*=0.70<*_>unch<*/>0.19 m J10 38 s-1 with a wind speed of 6.12<*_>unch<*/>0.35 m J10 39 s-1. This yielded a somewhat smaller value for J10 40 z0 of 0.17<*_>unch<*/>0.04 m (z0/h=0.041). This J10 41 was probably due to a slight change in wind direction.

J10 42 (c) Wind profiles above the canopy and comparison with the J10 43 sonic anemometer data

J10 44 Wind profiles measured during case 1 (Fig. 9) showed the J10 45 presence of a large kink at 5 m above the zero-plane displacement J10 46 with a significant reduction in wind speed at this level. Since the J10 47 profile is not logarithmic, the measurements here must be J10 48 considered as point flux measurements whose representativeness is J10 49 open to debate. However, the profile data obtained during case 2 J10 50 (Fig. 10) showed much closer agreement with a logarithmic profile J10 51 indicating that the fetch during this period was more ideal, with J10 52 the turbulence in equilibrium with the underlying canopy (at least J10 53 up to the level of the USAT). The wind speeds during case 2 were J10 54 very much higher than during case 1, with no sign of a kink at the J10 55 5 m level above the zero-plane displacement. The profile J10 56 anemometers did not overlap the USAT and a detailed comparison J10 57 between flux profile and eddy-correlation-derived u* is J10 58 difficult for case 1 where it would appear that the USAT was J10 59 sampling a different turbulent boundary layer from most of the J10 60 profile anemometers.

J10 61 The shape of the profile from case 1 is characteristic of flow J10 62 from rough to smooth terrain, with a sharp transition between the J10 63 equilibrium boundary layer and the outer boundary layer. If a J10 64 similar process has caused the kink at 5 m above the zero-plane J10 65 displacement, then the air in the inner boundary layer would J10 66 accelerate slightly, resulting in horizontal divergence and J10 67 subsidence of the air above the inner layer. If one assumes that J10 68 the upper two points of the profile are representative of rougher J10 69 terrain upwind then a roughness length quite close to that measured J10 70 with the USAT can be calculated. In contrast the inner-layer data J10 71 produced a roughness length of 0.13 m, less than half that derived J10 72 from the USAT in the outer layer. This is consistent with the J10 73 decrease in canopy height up the hill. The fluxes derived at the J10 74 measurement site are thus more characteristic of the upwind J10 75 terrain.

J10 76 graphs&captions

J10 77 Since the canopy height changes significantly over the hill it J10 78 is difficult to derive a value for the zero-plane displacement for J10 79 the upwind terrain. The variables quoted above have all been J10 80 derived assuming a value of d=2.81 m which corresponds to the J10 81 local canopy height. The profile measurements indicate a local J10 82 zero-plane displacement between 3 and 3.5 m corresponding to a J10 83 d/h ratio of between 0.71 and 0.83.

J10 84 The profiles from case 2 showed close agreement with a J10 85 logarithmic wind profile, with a correlation coefficient of 0.98. J10 86 The zero-plane displacement was again estimated to be between 3.0 J10 87 and 3.5 m with a roughness length of 0.08 to 0.14 m. Applying the J10 88 profile-derived d values to the USAT measurements yielded a J10 89 roughness length of 0.15-0.17 m. Because of the uncertainty in J10 90 obtaining good estimates of d, absolute comparison between the J10 91 USAT and profile anemometers is limited. For case 2, using J10 92 d=3.5 m, good agreement between profile-derived and eddy J10 93 correlation u* was obtained with the ratio J10 94 formula with a correlation coefficient of 0.98.

J10 95 Data obtained during other periods of thin cloud during case 2 J10 96 produced values of z0 varying from 0.25<*_>unch<*/>0.10 J10 97 to 0.29<*_>unch<*/>0.10 m (z0/h=0.06 to 0.07). The J10 98 changes in roughness length probably resulted from small changes in J10 99 wind direction which could considerably alter the fetch at this J10 100 measuring site. The observed roughness lengths are generally J10 101 smaller than those quoted in the literature. Unsworth (1984) J10 102 presented data from a forest of Pinus sylvestris and J10 103 Pinus nigra var. maritima with a canopy height of J10 104 15.5 m; d/h was 0.76 and z0/h was 0.060 J10 105 (z0=0.93 m). Jaeger (1985) measured wind profiles above a J10 106 growing Scots Pine forest (Pinus sylvestris) over a J10 107 period of ten years and obtained linear regression estimates of J10 108 d and z0 as a function of stand height h. Using J10 109 his relationships we obtain d=2.42 m and z0=0.96 m. J10 110 Both the profile and sonic-derived zo values are significantly J10 111 smaller than those predicted by Jaeger, as is also the case with J10 112 the data from Unsworth. The scatter in Jaeger's data was, however, J10 113 considerable (r=0.44). Our z0/h values are within J10 114 the scatter of results reported elsewhere in the literature, for J10 115 example Brutsaert (1982).

J10 116 It is debatable whether the concept of zero-plane displacement J10 117 is at all applicable in the context of this work owing to the J10 118 uncertainties posed by use of the gradient technique when used over J10 119 non-ideal terrain. In particular, recent work over forest canopies. J10 120 Duyzer et al. (1990), tends to suggest that the normal J10 121 flux-gradient relationships for momentum and moisture vapour may J10 122 need to be modified close to the canopy, otherwise fluxes may be J10 123 significantly over-estimated by as much as 40%. For the J10 124 experimental site described here, the uncertainties in the J10 125 zero-plane displacement would lead to an over-estimate of fluxes, J10 126 in some wind directions by, at worst, 30%.

J10 127 (d) The turbulence coefficients and atmospheric J10 128 stability

J10 129 Typical values of turbulence coefficients A, B, formula J10 130 were 4.23<*_>unch<*/>1.06, 2.69<*_>unch<*/>1.05 and J10 131 1.22<*_>unch<*/>0.38 respectively. This shows that the vertical J10 132 component of the turbulence field was locally homogeneous and J10 133 compares well to flat idealized terrain results (see, for example, J10 134 Smith 1975), whereas the longitudinal component may have been J10 135 subjected to slope distortion and roughness change effects (see, J10 136 for example, Panofsky et al. 1982). These values are J10 137 consistent with turbulence measurements reported by Hogstrom J10 138 et al. (1984) over a coastal forest. This behaviour is to J10 139 be expected as the measurements are all made well within the inner J10 140 layer of the hill defined by Hunt et al. (1988), which is J10 141 about 30 m in this case.

J10 142 Sensible-heat-flux measurements made with the USAT enabled the J10 143 stability parameter z/L to be calculated, where L is the J10 144 Monin-Obukhov length. This parameter varied from +0.005 to +0.018. J10 145 The correction to the flux measurements was thus of the order of 5 J10 146 to 13% justifying the assumption of near-neutral stability. It is J10 147 not clear how reliable the measurements of heat flux from sonic J10 148 devices are in cloud. Schotanus et al. (1983) show that J10 149 under non-neutral conditions out-of-cloud contributions due to J10 150 vapour fluxes may be as large as 20%.

J10 151 (e) The normalized deposition velocities for liquid water J10 152 content

J10 153 For case 1 the mean ratio formula at the measurement J10 154 height was 0.22<*_>unch<*/>0.09 for the FSSP-100 and J10 155 0.26<*_>unch<*/>0.12 for the PVM-100. (These do not include periods J10 156 where the deposition velocity was negative). The deposition J10 157 velocity for momentum was 0.052<*_>unch<*/>0.012 m s-1 J10 158 averaged over the data set.

J10 159 For case 2 the normalized deposition velocities were slightly J10 160 higher, with the period indicated in Fig. 2 producing values of J10 161 formula of 0.28<*_>unch<*/>0.23 for the PVM-100 and J10 162 0.38<*_>unch<*/>0.29 for the FSSP-100. Most of the variation in J10 163 these results occurred during the first hour after the onset of J10 164 cloud. If these data are ignored the ratios for the rest of the J10 165 period were 0.24<*_>unch<*/>0.18 (PVM) and 0.37<*_>unch<*/>0.14 J10 166 (FSSP). The deposition velocity of rmomentum was J10 167 0.069<*_>unch<*/>0.028 m s-1 averaged over the data J10 168 set.

J10 169 The reasons for the deposition velocities for liquid water J10 170 being substantially less than for momentum in these two case J10 171 studies are discussed in detail below. Qualitatively, however, for J10 172 case 1 the wind speeds were very low, resulting in a very low J10 173 collection efficiency for the droplets. For case 2 the wind speed J10 174 was much higher, but the mean volume-weighted drop radius was only J10 175 4.4<*_>unch<*/>0.5 <*_>mu<*/>m, Fig. 3, and so the collection J10 176 efficiency was relatively small. These case studies serve to J10 177 underline the importance of a knowledge of the drop size J10 178 distribution undergoing deposition to a canopy and show that simple J10 179 momentum considerations employed in some models can lead to J10 180 over-estimates of cloud water fluxes.

J10 181 (f) Variation in deposition velocity with droplet J10 182 size

J10 183 Chamberlain (1967) measured, in a wind tunnel, the deposition J10 184 velocity of particles to grass as a function of size. Figure 11 J10 185 shows his results for different values of u*. Deposition J10 186 velocity increases with particle size below 10 <*_>mu<*/>m radius J10 187 and this is particularly marked for drops between 1 and 5 J10 188 <*_>mu<*/>m. Figure 11 also shows the results of Gallagher et J10 189 al. (1988) obtained over moorland. The behaviour is very J10 190 similar between 2 and 10 <*_>mu<*/>m although beyond 12 <*_>mu<*/>m J10 191 a decrease in deposition velocity was observed.

J10 192 Figures 12 and 13 show formula as a function of drop J10 193 radius for case 1 and case 2 with error bars calculated as J10 194 described earlier. The solid line shows formula for J10 195 droplets deposited at a rate equal to momentum. For case 1, as J10 196 expected, all but the very largest droplets fall below the momentum J10 197 value. Despite the appreciable liquid water content, turbulent J10 198 deposition rates in low wind speeds were small. Figure 13 shows J10 199 formula obtained for the first period of case 2 as cloud J10 200 formed over the canopy, again showing the error bars. The J10 201 deposition velocities are all much higher than for case 1, owing to J10 202 the higher wind speeds, and for larger droplets exceed that for J10 203 momentum.

J10 204 Figure 14 combines the results from the two case studies from J10 205 this work and compares these results with those of Gallagher J10 206 et al. (1988), normalized in the same way.

J10 207 The general conclusions from this figure are that for droplet J10 208 sizes of up to about 12 <*_>mu<*/>m radius the results from case J10 209 study 2 are very similar to those reported by Gallagher et J10 210 al. (1988), measured over moorland by a very different J10 211 technique. Above about 12 <*_>mu<*/>m radius the results diverge, J10 212 but the uncertainties were large in this range for both studies. J10 213 The light winds of case study 1 produced significantly different J10 214 results.

J10 215 diagram&caption

J10 216 5. MODEL COMPARISONS

J10 217 The assumption that cloud water deposition velocity J10 218 approximates to that for momentum is not valid under all J10 219 conditions. The variation of collection efficiency with drop size J10 220 must be taken into account, using a knowledge of the droplet size J10 221 spectrum. To provide descriptions for the deposition of cloud water J10 222 that are more accurate, two model studies were performed. J10 223 J11 1 <#FLOB:J11>3. SEDIMENTOLOGY

J11 2 (a) Unit 1

J11 3 This unit consists predominantly of gravelly lithofacies J11 4 represented by massive to crudely horizontally stratified, medium J11 5 to coarse sandy gravels (facies Gm [Miall, 1977]) and subsidiary J11 6 planar cross stratified sets of fine gravel (facies Gp). The sandy J11 7 matrix is grey in colour. The unit has a maximum observed thickness J11 8 of 2.8 m (Section 1, Fig. 4). The base of the unit was not seen.

J11 9 The gravels typically form laterally persistent sheets (0.2-0.7 J11 10 m in thickness) which lie on erosional planar or occasionally J11 11 concave-up surfaces. Internally, the gravels display well marked J11 12 fining-upwards sequences ranging upwards from coarse J11 13 clast-supported gravels at the base to finer matrix-supported J11 14 gravels (Section 1, Fig. 4). Individual sheets are separated by J11 15 lenses of planar cross-stratified sets (up to 0.4 m high) of medium J11 16 to fine sand (facies Sp). These sets often display reactivation J11 17 surfaces, marked occasionally by the presence of a pebbly foreset. J11 18 Also noted from within Unit 1 are several ice-wedge casts.

J11 19 In contrast to the grey sands within the basal part of Unit 1, J11 20 towards the top of the unit there are occasional beds of generally J11 21 structureless red-brown sands similar to those of Unit 2. Eighteen J11 22 palaeocurrent measurements taken from the planar sand foresets J11 23 within Unit 1 range from 32 <*_>degree<*/> to 220<*_>degree<*/>, J11 24 with a mode in the 60-75<*_>degree<*/> range.

J11 25 (i) Interpretation

J11 26 The basal sands and gravels of Unit 1 represent a high energy J11 27 fluvial environment. It is suggested that the gravels formed as J11 28 longitudinal/transverse bars at high stage in channels with high J11 29 width to depth ratios. The majority of the gravels (facies Gm) J11 30 probably moved at high stage as gravel sheets (Eynon and Walker, J11 31 1974). As high stage flows waned the coarse gravels at the base of J11 32 each sheet were deposited. These gravels then acted as a nucleus J11 33 for further deposition and progressively finer gravels were J11 34 deposited as flow strength declined. This process of fining-upwards J11 35 deposition may have been assisted by the vertical growth of the bar J11 36 itself, which led to decreasing flow depth (Hein and Walker, J11 37 1977).

J11 38 Occasionally these gravel sheets developed avalanche faces J11 39 (facies Gp) possibly due to the waning of the flood event resulting J11 40 in the development of separation eddies in the lee of the bar, or J11 41 due to the migration of the bar into deeper water areas or areas of J11 42 flow expansion (Hein and Walker, 1977).

J11 43 During reduced flows sedimentation would have been restricted J11 44 to lateral channels and channels cut along the upper surfaces of J11 45 the bars (facies Sp). The presence of reactivation surfaces within J11 46 the sands suggests that energy conditions were variable within the J11 47 lower stage flows.

J11 48 The predominantly grey matrix associated with the lower J11 49 cross-stratified sands of unit 1 suggests a source area for the J11 50 sediment predominantly from the Jurassic rocks to the south of the J11 51 site (Fig.2). This conclusion is supported by the palaeocurrent J11 52 measurements which suggest a flow from the south-west.

J11 53 The presence of intraformational ice-wedge casts suggests that J11 54 deposition was, at least in certain areas, discontinuous as the J11 55 development of ice wedge casts is not possible close to the active J11 56 river channel (French, 1976).

J11 57 Similar sequences to that described above have been recorded J11 58 from many terrace gravel sites within southern Britain (Bryant, J11 59 1983; Dawson, 1985) and are attributed to periglacial river systems J11 60 where flood generation is largely the result of spring snowmelt. It J11 61 seems probable that Unit 1 represents a similar environment.

J11 62 figure&caption

J11 63 (b) Unit 2

J11 64 This unit is predominantly a well-sorted red-brown medium to J11 65 fine sand with occasional gravels. The unit has a maximum observed J11 66 thickness of 3.2 m (Section 1, Fig. 4). The bounding surface J11 67 between Units 1 and 2 is erosional. The lower parts of Unit 2, Unit J11 68 2a, comprise variable sequences of planar cross-stratified (facies J11 69 Sp) sands, ripple laminated sands (facies Sr) and large scale J11 70 trough cross-stratified sands (facies St). The planar J11 71 cross-stratified facies usually comprise isolated sets up to 20 cm J11 72 thick and occasionally cosets up to 50 cm thick. Rolled clay balls J11 73 (1.5 cm diameter) of Mercia Mudstone lying upon foreset laminae of J11 74 the sands (facies Sp) are present in places. Seven palaeocurrent J11 75 measurements taken from the planar cross-stratified sands in J11 76 Sections 1 and 2 range from 80<*_>degree<*/> to 195<*_>degree<*/> J11 77 with a modal class range of 165-180<*_>degree<*/>.

J11 78 The trough cross-stratified sets are up to 1m thick, although J11 79 only thinner sets were recorded from the accessible sections J11 80 (Section 3, Fig. 4). These lithofacies tend to lie on concave-up J11 81 erosional bases cut, in places, into the underlying Unit 1.

J11 82 Cyclic sedimentary sequences occur within the upper parts of J11 83 Unit 2. Unit 2b (Section 4, Fig. 4) comprises basal J11 84 horizontally-bedded sands with alternating coarser and finer beds J11 85 (facies Sh) rising into ripple-drift dominated sands (facies Sr). J11 86 Within each cycle the angle of climb of climbing ripples steepens J11 87 up-sequence, commonly being near-vertical at the top with both J11 88 stoss and lee side preservation (transition from type A to type B J11 89 of Allen, 1982). Occasionally the cycle culminates with near J11 90 vertical type S climbing ripples. Finally, fine-grained laminae cap J11 91 each cycle, the uppermost of which marks the contact of Unit 2 with J11 92 Unit 3. This bounding surface appears to be conformable throughout J11 93 the whole exposure, preserving the upper surface of the ripples.

J11 94 (i) Interpretation

J11 95 Unit 2 represents a lower energy environment than that of Unit J11 96 1. The observed occurrence of the red-brown sands of Unit 2a within J11 97 the underlying Unit 1 suggests that, at least initially, deposition J11 98 of Unit 2 was pene-contemporaneous with deposition of the upper J11 99 parts of Unit 1.

J11 100 The sands of Unit 2a (Sections 2 and 3, Fig. 4) display J11 101 sedimentary structures consistent with fluvial deposition and given J11 102 their association with the underlying Unit 1 are interpreted as J11 103 representing a relatively lower energy fluvial environment.

J11 104 The large trough cross-stratified sets possibly represent dunes J11 105 occupying the deeper parts of the active channel, indicated by the J11 106 underlying concave-up base. The planar cross-sets possibly J11 107 developed by the migration of linguoid or transverse bars with J11 108 ripple cross-lamination developing in shallower areas of the active J11 109 channels (Miall, 1977).

J11 110 Palaeocurrent measurements from Unit 2a suggest a flow J11 111 dominantly from the north-west. This flow pattern is consistent J11 112 with a sediment source to the north and west within the Triassic J11 113 sands (Bromsgrove Sandstone and Kidderminster Formation) in the J11 114 area around Birmingham (Fig.2).

J11 115 The cyclic sediments of Unit 2b (Section 4, Fig. 3a) are J11 116 interpreted as turbidites deposited within a lake. Within each J11 117 cycle parallel-laminated sands pass rapidly upwards into ripple J11 118 drift sands, indicating a progressive decrease in flow velocity. J11 119 The succession from type A to type B climbing ripples is J11 120 interpreted as the result of increasing sedimentation rates in J11 121 response to the continuing trend of a decrease in flow strength J11 122 (Allen, 1972). Where present type S climbing ripples represent very J11 123 high sedimentation rates. The draped silt/clay laminations which J11 124 complete each cycle indicate further decline in flow strength, with J11 125 sediment settling out from suspension. This declining flow strength J11 126 cycle is typical of turbidity currents and is considered to record J11 127 the B-C-D transition of a typical Bouma sequence (Bouma, 1969).

J11 128 Comparable sedimentological sequences to that within Unit 2b at J11 129 Snitterfield recorded from Lake Malaspina (Gustavson, 1975) and J11 130 glacial Lake Hitchcock (Gustavson et al., 1975) were J11 131 considered to represent sedimentation on or near to the pro-delta J11 132 slope. Thus a similar depositional environment is envisaged for J11 133 Snitterfield with the cyclic nature of sedimentation being J11 134 controlled by turbidity currents down the delta slope.

J11 135 The lack of a full Gilbert-type delta sequence, in particular J11 136 the lack of delta foresets, suggests that either there was not J11 137 enough time for a complete delta sequence to develop, or that not J11 138 enough sediment was arriving at the site to build a substantial J11 139 delta.

J11 140 (c) Unit 3

J11 141 The lower part of Unit 3 consists of rhythmitic coarse-fine J11 142 sands and silts. The nature of the rhythmites changes upwards J11 143 through the profile. At the base of the sequence each consists of a J11 144 sand layer, grading upwards into finer silts which occasionally J11 145 grade upwards into finer-grained, graded clay layers. Towards the J11 146 top of the rhythmite section the coarse sand layer becomes less J11 147 prominent and the clay layers thicken. Within the upper parts of J11 148 Unit 3 the rhythmites give way to a more massive clay. The J11 149 rhythmites described below were recorded from Section 4 (Fig. 4) J11 150 and are shown in detail in Fig. 5.

J11 151 The exact thickness of each of the sand layers varies J11 152 throughout the deposit, but generally their thickness decreases J11 153 upwards from about 1 cm in the lowest rhythmites to less than 1 mm J11 154 half way up the recorded section, before disappearing completely J11 155 towards the top of the recorded section.

J11 156 figure&caption

J11 157 The sand layers within the lower rhythmites display ripple J11 158 cross-stratification, with individual foreset laminate often being J11 159 less than 1 mm in thickness, although still exhibiting normal J11 160 internal grading. Ripples reach a maximum length of 2.5 cm with J11 161 recorded amplitudes of up to 6 mm (Fig. 5). The ripple forms are J11 162 generally well-preserved due to the draping of the overlying silts. J11 163 Climbing ripple sequences are also observed within the sands.

J11 164 The sands frequently display parting lineations, which show J11 165 sorting of coarser and finer grains into ridges and troughs. These J11 166 are interpreted as current lineations and, as such, are consistent J11 167 with the palaeocurrent directions recorded from Unit 2. The base of J11 168 the sand layers frequently shows scour of the underlying bed to J11 169 depths of 1-2 mm.

J11 170 Multiple graded beds of silt occur between the sand and clay J11 171 layers, silt always appearing to form the major part of the J11 172 rhythmite. The silt layers progressively decrease in thickness J11 173 relative to that of the clay layers up through the sequence. The J11 174 silt layers consist of normally graded beds with occasional sand J11 175 partings, especially in the lower part of the sequence. The clay J11 176 layers, where present at the base of Unit 3, are thin (3 mm or J11 177 less) but gradually become thicker towards the top of the J11 178 <}_><-|>rhthmite<+|>rhythmite<}/> section with recorded thicknesses J11 179 of up to 10 mm.

J11 180 Scattered clasts of quartz and quartzite occur throughout Unit J11 181 3. The largest recorded was 11 mm diameter. Typically, the clasts J11 182 are less than 4 mm in maximum observed dimension.

J11 183 Structures including faulting, folding and dewatering J11 184 structures, are present throughout the whole rhythmite section J11 185 (lower Unit 3). Faulting is predominantly normal (extensional) and J11 186 near vertical, with movement on the scale of millimetres, although J11 187 reverse (thrust) faulting is also present. The deformation does not J11 188 appear to extend upwards into the more massive clays (upper Unit 3) J11 189 and thus it is considered that the faulting is syndepositional. J11 190 Additionally dewatering fissures are present with associated J11 191 disruption of the surrounding sediment.

J11 192 Numerous trace fossils are present within Unit 3, mainly in the J11 193 form of short narrow burrows (4 mm long, 2 mm wide) which often J11 194 combine to form dendritic networks. The burrows are both J11 195 horizontal, along bedding planes, and vertical, across multiple J11 196 graded units and appear to have no preference as to grade of J11 197 sediment within the rhythmite.

J11 198 (i) Interpretation

J11 199 Unit 3 displays the rhythmitic dark/pale sequence recorded by J11 200 Shotton (1953). It is suggested that it is unlikely that the J11 201 rhythmites represent annual glacial varves as interpreted by J11 202 Shotton (1953) and it is considered more probable that they J11 203 represent small scale turbidites. Douglas (1980<}_><-|>)<+|><}/>, J11 204 p.282) proposed a similar interpretation for rhythmites within the J11 205 Bosworth Clays and Silts of western Leicestershire based upon the J11 206 presence of multiple graded silt beds and small-scale cross-bedding J11 207 within the silt bands.

J11 208 During deposition of the lower part of Unit 3 the higher J11 209 velocity flows were sufficient to transport sands, develop J11 210 migratory bedforms and cause scour of the underlying beds. The J11 211 presence of current lineations suggests flow velocities J11 212 sufficiently great to form upper phase plane beds (Leeder, 1982). J11 213 Such features have been recorded from within Pleistocene lacustrine J11 214 sequences in Ireland (Cohen, 1979). The nature of sedimentation at J11 215 the beginning of each rhythmite indicates an initial rapid increase J11 216 in flow strength, followed later by a more gradual decline. This J11 217 flow pattern is consistent with that found in turbidity currents J11 218 and is <}_><-|>simlar<+|>similar<}/> to that responsible for the J11 219 sands of Unit 2b. Thus these sediments are considered to be distal J11 220 equivalents to the sands of Unit 2b.

J11 221 J11 222 J12 1 <#FLOB:J12\>MATERIALS AND METHODS

J12 2 (Ca2+-Mg2+)-ATPase was purified from rabbit J12 3 skeletal muscle SR as described previously (Colyer et J12 4 al., 1989) and gave a preparation which, on polyacrylamide J12 5 gels stained with Coomassie Blue, was >97% pure (Gould et J12 6 al., 1987). mAbs were prepared as described in Colyer et J12 7 al. (1989) and were purified from ascites fluid by J12 8 precipitation with 40%-satd. formula followed by dialysis J12 9 against 2x1 litre of phosphate-buffered saline (PBS; 137 J12 10 mM-NaCl, 2.7 mM-KCl, 8.1 mM-formula, 1.5 J12 11 mM-formula, pH 7.2), giving a preparation containing J12 12 approx. 50% mAb by weight at a protein concentration of 1-5 J12 13 mg/ml.

J12 14 Synthesis and use of hexapeptides

J12 15 Hexapeptides were synthesized on blocks of polyethylene pins J12 16 designed to fit into 96-well microtitre plates, using kits supplied J12 17 by Cambridge Research Biochemicals, as described by Geysen et J12 18 al. (1987) by stepwise chain elongation using Fmoc J12 19 (fluorenylmethoxycarbonylamine) chemistry (Van Regenmortel et J12 20 al. 1988). NN'-Dimethylformamide (DMF; Romil Chemicals Ltd.) J12 21 was stored over molecular sieve 4A (BDH) until assay for free amino J12 22 groups using fluorodinitrobenzene showed it to be free of J12 23 dimethylamine. All other reagents were AnalaR grade, and water was J12 24 distilled. The pre-derivatized pins were supplied with Fmoc-alanine J12 25 linked via hexamethylenediamine to the pins. The Fmoc N-terminal J12 26 protecting group was removed from the pins and from subsequent J12 27 coupled Fmoc-amino acids by washing with 20 % (v/v) piperidine in J12 28 DMF for 30 min at room temperature with shaking. The pins were then J12 29 washed for 5 min with DMF, four times with methanol for 2 min, J12 30 air-dried for 10 min and washed for a further 5 min with DMF. J12 31 Fmoc-amino acid active esters (Fields & Noble, 1990) were dissolved J12 32 in DMF containing 1-hydroxybenzotriazole and the mixtures were J12 33 dispensed into the appropriate wells of a polypropylene e.l.i.s.a. J12 34 plate according to a computer-generated synthesis schedule. The J12 35 pins were inserted into the wells and coupling was allowed to J12 36 proceed at room temperature for at least 18 h. The blocks were then J12 37 removed and washed as above. After synthesis of the hexapeptides J12 38 and de-protection of the N-terminal amino acid, the N-terminal J12 39 amino acid was acetylated by placing the block of pins into a J12 40 mixture of DMF, acetic anhydride and J12 41 di-iso-propylethylamine (50:5:1, by vol.) and incubating J12 42 for 90 min at room temperature. The blocks were then washed as J12 43 above, Side-chain-protecting groups were removed by placing J12 44 the block of pins in a mixture containing trifluoroacetic acid, J12 45 phenol and ethanedithiol (38:1:1, v/w/v) for 4 h at room J12 46 temperature. The pins were then washed twice with dichloromethane J12 47 for 2 min, twice with 5% di-isopropylethylamine in dichloromethane J12 48 for 5 min, and once with dichloromethane for 5 min before J12 49 air-drying and finally washing in water for 2 min and J12 50 methanol for 18 h before drying in vacuo.

J12 51 Enzyme immunoassays were carried out by first precoating the J12 52 pins for 1 h by insertion into wells of e.l.i.s.a. plates J12 53 containing buffer A (1% ovalbumin, 1% BSA, 0.1% Tween 20 and 0.05% J12 54 NaN3 in PBS). Pins were then incubated overnight at 4 J12 55 <*_>degree<*/>C in e.l.i.s.a. plates containing mAb at a 1:500 J12 56 dilution with the buffer mixture described above. The pins were J12 57 then washed three times for 10 min with PBS containing 0.05% Tween J12 58 20, with shaking. Bound antibody was detected by incubation for 1 h J12 59 at 25 <*_>degree<*/>C in e.l.i.s.a. plates containing second J12 60 antibody conjugated to horseradish peroxidase (HRP; Sera-Tec, J12 61 diluted 1:1000 in buffer A but in the absence NaN3). The J12 62 pins were washed as before and then incubated in substrate solution J12 63 [50 mg of azino-di-3-ethylbenzthiazodinosulphate, 35 <*_>mu<*/>l of J12 64 100 volume H2O2 in 100 ml of citrate/phosphate J12 65 buffer (80mM), pH 4.0] in the wells of an e.l.i.s.a. plate. J12 66 The colour was allowed to develop in the dark for 30 min and then J12 67 the plates were read at 410 nm on a Dynatech MR588 Microelisa Auto J12 68 Reader. To allow re-use, antibodies were removed from the pins by J12 69 cleaning in a sonication bath with 1% SDS/0.1% J12 70 2-mercaptoethanol/0.1 M-formula, pH 7.2 at 60 J12 71 <*_>degree<*/>C for 30 min, followed by washing in hot distilled J12 72 water (55-60 <*_>degree<*/>C). Pins were then immersed in boiling J12 73 methanol for 2 min and dried in air.

J12 74 Preparation of polyclonal antibodies

J12 75 Polyclonal antibodies to the purified ATPase were raised in J12 76 sheep by intramuscular injections of 0.5 mg of ATPase in Freund's J12 77 complete adjuvant followed by boosts in Freund's incomplete J12 78 adjuvant. Antibodies were purified from antisera by precipitation J12 79 with 40% formula followed by dialysis against 2x1 litre of J12 80 PBS. Antibodies were immunoprecipitated as follows. Samples (0.2 J12 81 ml, containing 0.2 mg of antibody) were diluted to 10 ml with PBS. J12 82 Purified ATPase (1 mg in 100 <*_>mu<*/>l) was then added to give a J12 83 ratio of ATPase/antibody of 5:1 (w/w), and the sample was incubated J12 84 for 1 h at room temperature with shaking to allow precipitation of J12 85 the antibody-antigen complex. Samples were then spun at 37000 J12 86 g for 30 min at 20 <*_>degree<*/>C, and the supernatant was J12 87 removed and assayed as described above for the mAbs.

J12 88 Preparation of peptides and anti-peptide antibodies

J12 89 Peptides are named according to the residues to which they J12 90 correspond in the ATPase. Peptides were synthesized by the method J12 91 of Merrifield (1986) and checked for purity by h.p.l.c. on a J12 92 reversed-phase C8 column. Peptides were coupled to the carrier J12 93 protein keyhole limpet haemocyanin (KLH) before immunization J12 94 according to the method of Green et al. (1982). Primary J12 95 immunizations were carried out with approx. 0.5 mg of peptide-KLH J12 96 in Freund's complete adjuvant which was injected by the J12 97 intramuscular route in New Zealand white rabbits on day 1. A J12 98 booster injection of 0.25 mg of peptide-KLH was given in Freund's J12 99 incomplete adjuvant by the same route after 28 days, and blood was J12 100 taken for the production of antisera 7-14 days later. Antisera were J12 101 stored at -70 <*_>degree<*/>C. Antibody was purified from antisera J12 102 by precipitation with 40% formula, as described above. J12 103 Further purification was carried out using an ATPase affinity J12 104 column made by reacting SR vesicles with CNBr-activated Sepharose. J12 105 Antiserum was passed through the column at pH 8.0 and bound J12 106 antibody was eluted with glycine/HCl (0.1 M, pH 2.5). The J12 107 eluent was immediately neutralized with a few drops of Tris/HCl (2 J12 108 M, pH 7.4) and dialysed against 2x1 litre of PBS. To study J12 109 effects of the affinity-purified anti-peptide antibody on the J12 110 activity of the ATPase, the enzyme was incubated for 30 min with J12 111 antibody at a 3:1 molar ratio of antibody to ATPase. ATPase J12 112 activity was determined by a coupled enzyme assay (Colyer et J12 113 al. 1989). All anti-peptide antibodies were mapped against the J12 114 sets of pins as described above for the mAbs, and all showed J12 115 recognition of the appropriate hexameric peptides, with no J12 116 significant binding to other hexameric peptides.

J12 117 figure&caption

J12 118 Competitive e.l.i.s.a.

J12 119 A modified e.l.i.s.a. technique was used to demonstrate that J12 120 the monoclonal and polyclonal antibodies bound to native ATPase J12 121 (Colyer et al., 1989). Antibody (0.7 <*_>mu<*/>g) was J12 122 incubated overnight at 4 <*_>degree<*/>C in PBS/Tween in uncoated J12 123 e.l.i.s.a. plates with the ATPase (2-160 <*_>mu<*/>g/ml) either in J12 124 the native form or after heat denaturation (100 <*_>degree<*/>C for J12 125 3 min). Samples were then transferred to e.l.i.s.a. plates coated J12 126 with the ATPase (1 <*_>mu<*/>g/well) and e.l.i.s.a. was performed J12 127 using second antibody conjugated to HRP (Colyer et al., J12 128 1989).

J12 129 RESULTS J12 130 Competitive e.l.i.s.a. was used to show that mAbs bound to the J12 131 native ATPase (Fig. 1). Native and heat-denatured ATPases were J12 132 incubated overnight with mAb, and the incubations were transferred J12 133 to e.l.i.s.a. plates coated with native ATPase to determine the J12 134 amount of free mAb remaining in the incubations. Fig. 1 shows that J12 135 although mAbs Y/1F4 and Y/2E9 bound more strongly to denatured than J12 136 to native ATPase, the difference was rather small. For example, for J12 137 Y/1F4, twice as much of the native ATPase than of the denatured J12 138 ATPase was required to lower the e.l.i.s.a. reading by 50 %. J12 139 Similar results were obtained with the other mAbs (results not J12 140 shown), confirming that the epitopes for these mAbs are J12 141 surface-exposed in the native ATPase.

J12 142 To define the exact epitopes for the mAbs, a set of overlapping J12 143 hexameric peptides was synthesized on polyethylene pins arranged to J12 144 fit the wells of standard microtitre plates (Geysen et J12 145 al., 1987). On the basis of our preliminary mapping of the J12 146 epitopes (Colyer et al., 1989), we chose to synthesize J12 147 hexapeptides corresponding to amino acid regions 1-208, 277-381 and J12 148 486-751.

J12 149 Our previous studies suggested that the epitope for mAb Y/1F4 J12 150 lay between amino acids 500 and 615 (Colyer et al.., J12 151 1989; Tunwell et al., 1991). The results of the binding J12 152 analysis for mAb Y/1F4 with the set of hexapeptides from amino J12 153 acids 486-751 are shown in Fig. 2(a). mAb Y/1F4 showed strong J12 154 recognition of hexapeptide NKMFVK, corresponding to residues J12 155 510-515, indicating that this is the epitope for the antibody. The J12 156 adjacent peptides VGNKMF, KMFVKG and MFVGAP gave lower signals, J12 157 suggesting that these peptides contain only part of the epitope, J12 158 with the sequence MF (Met-Phe) being the most critical part of the J12 159 epitope. As shown in Fig. 2(d), mAb Y/2E9 gave a strong signal J12 160 for hexapeptides LPLAEQ and PLAEQR, suggesting that the epitope was J12 161 PLAEQ, corresponding to amino acids 662-666; the very weak binding J12 162 to the neighbouring hexapeptides suggest that in this case the J12 163 sequence PL (Pro-Leu) is critical for binding. mAb A/4H3 showed J12 164 binding very similar to that of mAb Y/2E9, and mAb B/4H3 showed J12 165 strong binding to peptides PLAEQR and LAEQRE, suggesting that their J12 166 epitopes are the same as that of mAb Y/2E9 (Table 1). As shown in J12 167 Fig. 2(c), mAb 1/2H7 bound strongly to hexapeptides DDSSRF, J12 168 DSSRFM, SSRFME and SRFMEY, so that the major component of its J12 169 epitope is probably SRF (Ser-Arg-Phe), corresponding to residues J12 170 583-585. Also, as shown in Fig. 2(b), the epitope for mAb J12 171 1/3D2 is SSRFME, corresponding to amino acids 583-588.

J12 172 From our previous epitope mapping experiments (Colyer et J12 173 al., 1989; Tunwell et al., 1991), epitopes from mAbs J12 174 B/3D6, Y/1G4, Y/1G6, Y/2D8, Y/2D11 and Y/3H2 were all thought to J12 175 lie between amino acid residues 486 and 750, but for none of these J12 176 mAbs was any clear pattern of binding to the hexameric peptides J12 177 observed (results not shown), so that the epitopes for these mAbs J12 178 cannot be defined using this approach.

J12 179 Epitopes for mAbs Y/3H5, Y/1H12, Y/2A2 and Y/4D6 have all been J12 180 suggested to lie between residues 320 and 376 (Colyer et J12 181 al., 1989, Tunwell et al., 1991). Binding of these J12 182 mAbs was therefore tested with a set of hexapeptides corresponding J12 183 to amino acids 277-381. The results for mAb Y/4D6 are shown in Fig. J12 184 3 and are considerably less clear cut than those described above. J12 185 Binding to hexapeptides. AAIPEG, AIPEGL and IPEGLP was strong, J12 186 suggesting an epitope whose main binding IPEGLP was strong, J12 187 suggesting an epitope whose main binding determinant is IPE J12 188 (Ile-Pro-Glu), corresponding to amino acids 307-309. However, J12 189 equally strong binding was seen to hexapeptide VRSLPS (residues J12 190 333-338) which could suggest that the epitope for the mAb is an J12 191 assembled one, containing both of these regions. Slightly weaker J12 192 binding is also observed to hexapeptides PVHGGS (residues 282-287) J12 193 and VHGGSW (residues 283-288); since hydropathy plots suggest that J12 194 either residues 288-307 (MacLennan et al., 1985) or J12 195 residues 295-315 (Green, 1989) make up a transmembrane J12 196 <*_>alpha<*/>-helix, these could not be part of the same epitope as J12 197 residues 307-309 and 333-338, and binding to these two hexapeptides J12 198 would then have to be artifactual. Alternatively, one set of J12 199 reactive peptides might be mimotopes of the true epitope. The J12 200 pattern of binding observed with mAb Y/3H5 was very similar to that J12 201 shown in Fig. 3 for mAb Y/4D6. mAbs Y/1H12 and Y/2A2 only gave weak J12 202 binding to any of the hexapeptides, and so could not be clearly J12 203 mapped.

J12 204 graphs&captions

J12 205 To check the epitope assignments made above, the peptides shown J12 206 in Table 1 were synthesized using the method of Merrifield (1986), J12 207 with all peptides containing an N-terminal Cys. E.l.i.s.a. was J12 208 performed using peptide coated on to e.l.i.s.a. plates, and, as J12 209 shown in Table 1, Y/1F4, Y/2E9 and 1/2H7 all bound to their J12 210 respective peptides. No binding of mAb Y/3H5 was observed in J12 211 e.l.i.s.a. to peptides AVAAIPEGLPAV (residues 303-314) or J12 212 RRMAKKNAIVRSLPSV (324-339), which contain the two possible J12 213 contributory epitopes for this mAb (see Table 1); a negative result J12 214 in this experiment does not, however, prove that the epitope for J12 215 the mAb is not contained in the peptides and could, for example, J12 216 simply indicate that the peptides are coated on to the plastic in J12 217 conformations unsuitable for binding.

J12 218 J13 1 <#FLOB:J13\>Preparation and counselling

J13 2 The main centres involved in presymptomatic testing for HD have J13 3 all taken a cautious and responsible approach to what has J13 4 universally been recognised to be a difficult situation. The J13 5 testing has been performed initially as a research evaluation, with J13 6 detailed pre-test counselling sessions and careful arrangements for J13 7 post-test support and follow up. Possibly as a result of this, J13 8 there have been no serious adverse short term effects in those J13 9 given a high risk result, though the longer term outcome remains to J13 10 be seen.

J13 11 Should such a thorough procedure, involving many hours of J13 12 individual counselling, be regarded as the accepted standard for J13 13 this type of genetic prediction? If so, who will provide the J13 14 necessary resources once the initial research phase is over? If J13 15 not, how is the minimal acceptable standard to be defined? Already J13 16 one major American HD testing centre has had to stop its activities J13 17 because funding was not renewed; applicants, in some instances, J13 18 have had to move their residence to a different part of America J13 19 where service funding for a testing programme was available.

J13 20 As testing becomes more widespread, what restrictions should be J13 21 placed on those clinicians, laboratories (and possibly commercial J13 22 organisations) offering predictive testing for HD to ensure quality J13 23 of both counselling and laboratory procedures? A disturbing example J13 24 of such practice (not from the UK) was reported at a recent meeting J13 25 on the subject, where the applicant was informed of a high risk J13 26 result for HD by telephone with no adequate prior counselling and J13 27 no subsequent support. Should such practice be regarded as J13 28 'negligent' and hence subject to legal pressures, or should there J13 29 be a voluntary code of practice?

J13 30 Looking to other disorders, it is already clear that the J13 31 arrangements for counselling often fall far short of those regarded J13 32 as necessary for HD. Perhaps the most obvious example is for the J13 33 transmissible but non-genetic disorder, acquired immunodeficiency J13 34 syndrome (AIDS), where counselling and follow up in relation to J13 35 human immunodeficiency virus (HIV) testing is often minimal despite J13 36 the profound consequences. In genetics, a recent report published J13 37 by the Royal College of Physicians (1989) has highlighted important J13 38 shortcomings in the provision of genetic counselling.

J13 39 Genetic testing in childhood

J13 40 Our predictive testing programme for HD has, like others so J13 41 far, only accepted adults, a policy in line with international J13 42 guidelines drawn up by the International Huntington's Association J13 43 and World Federation of Neurology (1990). We have been surprised, J13 44 however, how frequently requests for the testing of children have J13 45 arisen. In our series there were 28 such requests, 20 from parents. J13 46 While most parents have accepted the reasons for postponing testing J13 47 until the child could make its own decision, in one case the mother J13 48 of a two year old child, whose father was at 50 per cent risk, was J13 49 insistent on testing and sought a second opinion elsewhere when J13 50 testing was refused.

J13 51 Although it seems logical - for a disorder such as HD where J13 52 childhood onset is exceptional - to confine testing to those who J13 53 can consent for themselves, the situation is far from being clear J13 54 in all circumstances. Should DNA testing be done where behaviour or J13 55 other clinical abnormalities raise a suspicion of the rare juvenile J13 56 form? Should medical opinion override the wishes of a family which J13 57 has lived with the disease and is fully aware of its significance? J13 58 What is the position about testing other late onset disorders whose J13 59 age range is much more variable? The X-linked muscular dystrophies J13 60 provide a relevant example here: while in the classical Duchenne J13 61 type the onset is relatively constant in early childhood and severe J13 62 disability inevitable by late childhood, the sitation is quite J13 63 different for the later onset 'Becker' type, determined by J13 64 mutations at the same locus. Here onset may vary from late J13 65 childhood to middle age, while disability varies from minimal to J13 66 severe. An even greater range of disability applies to the J13 67 autosomal dominant myotonic dystrophy: what factors should J13 68 determine whether molecular genetic testing in childhood is J13 69 appropriate for such disorders?

J13 70 It is important to look critically at the reasons for J13 71 requesting testing in childhood. The concerns of parents fall into J13 72 two main categories; the future health of their child, and the J13 73 genetic implications. Regarding the latter it is surprising, though J13 74 understandable, how often parents seek to influence their J13 75 offspring's future reproductive decisions; this could be seen as J13 76 analogous socially to parental influence in the choice of marriage J13 77 partner. In discussing this subject with parents we have found it J13 78 helpful to stress the increasing choice that is likely to be J13 79 available to their children when adult, in particular that early J13 80 prenatal (possibly even preimplantation) diagnosis may allow them J13 81 to achieve a healthy family regardless of their own genetic status. J13 82 When the principal reason for requesting testing relates to the J13 83 health of the child, present or future, the situation may be very J13 84 much less clear-cut and careful consideration of the individual J13 85 circumstances is needed.

J13 86 One approach that is sometimes underestimated is the value of a J13 87 careful history and physical examinaton. If negative this will J13 88 allow the question 'is the disorder present now?' to be answered, J13 89 even if the future remains unresolved. If suspicious or positive J13 90 clinical features are indeed present, this may make further J13 91 investigations easier. Immediate health problems are not the only J13 92 reason why parents may request genetic testing. Early knowledge J13 93 that a disorder will develop may be important for educational and J13 94 future career decisions with such disorders as Becker muscular J13 95 dystrophy, which might well preclude a physically demanding career, J13 96 or in retinitis pigmentosa, where future blindness may result.

J13 97 Regarding these issues, there is no essential difference J13 98 between DNA testing for the gene and phenotypic presymptomatic J13 99 tests such as creatine kinase analysis in the muscular dystrophies, J13 100 electroretinography in retinitis pigmentosa, and ultrasound J13 101 examination for adult polycystic kidney disease. Indeed it is J13 102 likely that these more 'clinical' approaches are at present being J13 103 widely used in an uncritical and at times inappropriate manner by J13 104 paediatricians and others, though no clear evidence on this exists. J13 105 However the power of molecular testing to make a prediction in J13 106 disorders where no previous tests were available, together with its J13 107 independence of age, makes it especially important to apply it only J13 108 after careful thought and in the context of ethical guidelines. J13 109 Perhaps a widespread debate is needed on this neglected topic J13 110 (Harper and Clarke, 1990)?

J13 111 Preadoption testing

J13 112 In our HD prediction study we were surprised to receive five J13 113 requests for predictive tests to be done on children at risk for HD J13 114 who were being placed for adoption, and we are aware of two others. J13 115 This represents the extreme situation of genetic testing in J13 116 childhood discussed above, and brings up additional issues such as J13 117 the rights of the child in relation to those of the potential J13 118 adoptive parents. We have not undertaken such testing for reasons J13 119 set out briefly elsewhere (Morris et al., 1988), which J13 120 include the removal from the child of future choice regarding J13 121 testing, the possible (indeed likely) stigmatisation from the J13 122 knowledge that HD would develop in later life, together with the J13 123 general uncertainty of the long-term effects of prediction even in J13 124 adults, and the current lack of effective treatment.

J13 125 Our experience has however raised serious questions over the J13 126 legal situation, and over what tests for genetic disorders in J13 127 general are appropriate both now and in the future when adoption J13 128 placement is being considered. Currently some tests are already J13 129 mandatory (for example phenylketonuria), while others would be J13 130 generally advised (such as Duchenne muscular dystrophy in a child J13 131 at risk), but the question of testing for disorders of adult life J13 132 has not been addressed. Studies have shown that there is J13 133 considerable variation in the standard of the pre-placement J13 134 examinations and investigations.

J13 135 'INADVERTENT' GENETIC TESTING

J13 136 In most medical situations it is an individual who seeks advice J13 137 and is tested; in genetic disorders it is often the entire family, J13 138 nuclear or extended, that is under consideration. Testing on one J13 139 member may prove to have major implications for others, and these J13 140 should be considered before the testing is done, rather than J13 141 afterwards.

J13 142 While the discovery of a genetic disorder in an individual may J13 143 put relatives at risk, it rarely gives conclusive evidence as to J13 144 whether or not they will be affected. An exceptional situation is J13 145 provided by identical twins, where a diagnosis (or predictive test) J13 146 in one twin for a disorder such as HD would give certainty that the J13 147 co-twin would also become affected at some time. We have not so far J13 148 encountered this problem.

J13 149 A more frequent difficulty, and one that is easy to overlook, J13 150 is the inadvertent prediction that may result from samples being J13 151 tested of relatives who themselves do not wish for prediction. J13 152 Molecular genetic testing for most disorders currently remains J13 153 dependent on family testing for polymorphisms, whether J13 154 gene-specific or linked. Pedigree structures are frequently far J13 155 from perfect in disorders such as HD, and testing of sibs may be J13 156 the only way by which parental genotypes can be inferred to allow J13 157 prediction. Figure 10.1 shows an example of this situation.

J13 158 Unless great care is taken a laboratory may find itself testing J13 159 samples from a relative where the results may have major J13 160 implications for that relative, as well as for the person who has J13 161 sought prediction. The only way to avoid such a problem is to J13 162 prevent it occurring in the first place. To begin with, no sample J13 163 should be taken without careful review of the pedigree to be sure J13 164 that it is really necessary. There is often a temptation to take a J13 165 blood sample from relatives attending a clinic or on a home visit J13 166 'just in case' it might be needed in the future. Such samples J13 167 should not be taken, or if they are, the samples should not be J13 168 tested. When such a sample is really needed it must be made J13 169 completely clear to the donor (preferably in writing) that the J13 170 sample is being taken for the benefit of the relative and that no J13 171 result is to be expected. To avoid the laboratory being placed in J13 172 the invidious position of knowing information that is not known to J13 173 the individual, we insist on such samples being made anonymous J13 174 before testing.

J13 175 figure&caption

J13 176 How far is this a general policy outside the rigorous situation J13 177 of HD testing? Our own experience suggests that it is not, and that J13 178 laboratories commonly test all samples sent to them without J13 179 clarifying who has requested prediction and who not. We have found J13 180 ourselves in this sitation for such disorders as myotonic J13 181 dystrophy, and it is likely that a careful audit of requests for J13 182 genetic testing would uncover an appreciable number of comparable J13 183 situations. While the consequences may be less extreme than for HD, J13 184 the principle remains the same.

J13 185 THE USE OF RESEARCH RESULTS AND SAMPLES

J13 186 Most major advances in human molecular genetics have come from J13 187 the study of DNA from affected families who have donated blood J13 188 samples in the hope that this will help research. Not unnaturally, J13 189 when the promised breakthrough occurs they will wish to know what J13 190 implications it will have for them personally; they may enquire J13 191 about the results that a research worker has obtained, or about the J13 192 samples that remain stored for future use.

J13 193 Despite the recent tightening of ethical constraints for J13 194 research projects, this is probably the area of genetic prediction J13 195 that has been least carefully thought out and is most open to J13 196 abuse. Part of the problem results from the very simplicity of the J13 197 material needed. Whereas ethical committees may query the value of J13 198 a brain biopsy or a potentially hazardous neuropharmacological test J13 199 in a disorder like HD, few will cavil at the taking of a single J13 200 blood sample. The stability of DNA also compounds the problem, as J13 201 does the responsible and far-sighted behaviour of most geneticists J13 202 who insist (rightly) on banking DNA for its possible future service J13 203 use. Such a policy has already had direct benefits in allowing J13 204 testing for HD and other disorders such as Duchenne dystrophy after J13 205 the death of the affected family member. The same policy however J13 206 creates difficulty for decisions on the use of what was initially J13 207 collected as research material.

J13 208 J14 1 <#FLOB:J14\>This peak incidence is expressed as a fraction of the J14 2 total population. An exact expression for the peak incidence of J14 3 infection in a homogeneous population (which is equal to J14 4 vymax) is given by eqn (6.20), which can be seen to J14 5 reduce to eqn (11.95) when R0 is significantly in excess J14 6 of unity. The presence of carriers and of heterogeneity in sexual J14 7 habits produce complications, but eqn (11.95) remains a good J14 8 back-of-an-envelope estimate (so long as R0 is J14 9 appreciably greater than unity), and is consistent with broad J14 10 trends seen in Fig. 11.23. The result (11.95) arises essentially J14 11 because a fraction f of those infected move into the class J14 12 with full-blown AIDS at a rate v1, following the initial J14 13 stages of the epidemic in which (for R0 large) most of J14 14 the population is infected. Thus observations of the incidence of J14 15 clinical AIDS as the epidemic begins to peak give a very rough J14 16 estimate of the combination f/D, where D=1/v1 is the J14 17 average interval between infection and diagnosed AIDS. For example, J14 18 if a peak incidence of clinical AIDS of 2-3 per cent of the at-risk J14 19 population were found each year, we could roughly estimate J14 20 f<*_>approximate-sign<*/>10 per cent if D is around 3-4 J14 21 years - or f<*_>approximate-sign<*/>50 per cent if D is J14 22 around 15-20 years - and so on.

J14 23 The long-term public health issues, however, depend more on the J14 24 total fraction of the at-risk population who are likely to develop J14 25 clinical AIDS, that is (for R0 large) upon f. As J14 26 discussed above, the observed levels of seropositivity suggest J14 27 R0 is unlikely to be less than around three, and it may J14 28 well have a value of five or more. With formula, this means J14 29 the factor inside the square brackets two paragraphs above is J14 30 likely to have a value of 5 years or more. But this value could be J14 31 attained in two ways, which have different implications for J14 32 estimates of f. On the one hand, if formula, we have J14 33 the estimates that D is around 5 years or more, whence J14 34 observations about peak incidence of AIDS lead to a crude estimate J14 35 of f. On the other hand, if 1/v2>>1/v1, it J14 36 could be that the value of R0/<*_>beta<*/>c is determined J14 37 essentially by 1/v2, whence the above argument gives J14 38 virtually no information about D, and thus is of no help in J14 39 putting bounds on f.

J14 40 These arguments can be refined into more quantitative J14 41 estimates, with the help of numerical studies of the kind reported J14 42 in Figs. 11.22 and 11.23(a,b). An essential ambiguity remains, J14 43 however, in whether diagnosed AIDS-developers and carriers do or do J14 44 not have similar durations of infectiousness; that is, whether J14 45 v1<*_>approximate-sign<*/>v2 or not. The J14 46 analysis presented above suggests some ways of gathering J14 47 information to resolve this issue.

J14 48 Consider first the incidence of infection (as revealed by J14 49 seropositivity) as a function of time, among classes of individuals J14 50 distinguished according to levels of sexual activity. As shown in J14 51 Fig. 11.31(a), if v1=v2 there is a tendency for J14 52 the more sexually active individuals to acquire infection earlier, J14 53 and less active individuals later, so that the ratio of more J14 54 sexually active to less sexually active individuals among those J14 55 infected shifts as the epidemic progresses. As seen in Fig. J14 56 11.31(b), this tendency is much more pronounced if v1 is J14 57 substantially greater than v2 (carriers on average remain J14 58 infectious significantly longer than diagnosed AIDS-developers): J14 59 the more active people tend to acquire infection early on, whereas J14 60 infection among the less active fraction of the population slowly J14 61 builds up over time, depending primarily on the long-lived J14 62 infectiousness of carriers. Systematically compiled serological J14 63 data that discriminate among the levels of sexual activity of J14 64 recently infected individuals, over time, should in this way be J14 65 capable of shedding light on the ratio between v1 and J14 66 v2.

J14 67 graphs&captions

J14 68 The same general considerations apply to incidence of newly J14 69 diagnosed cases of full-blown AIDS, distinguished according to J14 70 levels of sexual activity in the patients, as a function of time. J14 71 Here, however, the patterns are more blurred, by virtue of the J14 72 statistical distribution of times between acquiring infection and J14 73 manifesting full-blown AIDS. Figures 11.32(a) and (b) correspond to J14 74 Figs. 11.31(a) and (b), illustrating the incidence of cases of J14 75 diagnosed AIDS disease, among more sexually active individuals and J14 76 among less active ones, as a function of time. Figure 11.32(a) is J14 77 for v1=v2 and Fig. 11.32(b) is for J14 78 v1=0.2 yr-1 and v2=0.05 J14 79 yr-1. The trends seen in Figs. 11.31(a) and (b) are again J14 80 evident, although somewhat less pronounced. Again, systematic J14 81 investigation of data of this kind could help determine the ratio J14 82 of v1 to v2.

J14 83 On another tack, it would be helpful to look at the incidence J14 84 of diagnosed AIDS disease, as a function of time, among those J14 85 individuals who were most sexually active. Most of these people are J14 86 likely to have acquired infection early on, so that their J14 87 distribution of incidence of disease over time gives a reasonable J14 88 indication of the actual distribution of incubation times (in a way J14 89 that is not possible if all data, from more active along with less J14 90 active people, are combined). Such analyses could help provide an J14 91 independent estimate of whether the distribution of incubation J14 92 times found among those infected with HIV by blood transfusions are J14 93 indeed representative of broader categories of HIV infections.

J14 94 Whether AIDS will establish itself as an endemic disease in the J14 95 USA and Europe depends on whether sexual habits change sufficiently J14 96 to push R0 below unity. This requires the distribution in J14 97 the number of sexual partners per unit time among homosexuals to J14 98 decrease to a level such that

J14 99 c'0. (11.96)

J14 100 Here R0 is the initial value of the basic J14 101 reproductive rate for AIDS, estimates of which we have just J14 102 discussed, and c and c' are the initial and later values, J14 103 respectively, of the appropriate average over the distribution in J14 104 sexual habits (formula). Downward changes in the mean J14 105 clearly help. But downward changes at the high end of the J14 106 distribution help disproportionately, because they have more effect J14 107 on 2>. Alternatively, R0 can be lowered by J14 108 changes in the intrinsic transmission parameter <*_>beta<*/> ('safe J14 109 sex'). In either event, it is clear that the magnitude of the J14 110 changes needed to drive R0 below unity depend on how J14 111 large R0 originally was; as we have seen, this number is J14 112 uncertain, but is unlikely to be less than five or so.

J14 113 All the above discussion has treated AIDS as a disease of J14 114 homosexuals. Needle-sharing among intravenous drug users involves J14 115 the same ideas, with different values for the transmission factors J14 116 <*_>beta<*/> and c, but the same parameters characterizing the J14 117 course of infection. For such transmission, <*_>beta<*/>c can be J14 118 estimated from initial doubling times, and the above discussion J14 119 repeated, mutatis mutandis.

J14 120 The same is true for heterosexual transmission, which is J14 121 currently low in developed countries but high in many sub-Saharan J14 122 African countries. As we have seen, simple models for purely J14 123 heterosexual transmission suggest that in the earlier stages of the J14 124 epidemic - before any subgroup becomes saturated with infection - J14 125 the incidence of HIV infection and of AIDS cases among males and J14 126 among females grows roughly exponentially at the same rate in both J14 127 populations. The doubling rate (which previously was <*_>beta<*/>c J14 128 - 1/D) is now approximately formula, where D is J14 129 the average duration of infectiousness. In developing countries, we J14 130 would expect to find the female-to-male transmission probability, J14 131 <*_>beta<*/>', and probably also the male-to-female transmission J14 132 probability, <*_>beta<*/>, less than the transmission probability J14 133 among homosexuals. We would also expect the appropriately averaged J14 134 rate of acquiring new sexual partners by women and by men, c' J14 135 and c, to be less than the corresponding quantity for J14 136 homosexuals in large cities in the developed world in the early J14 137 1980s. As was discussed in the section on epidemiological J14 138 parameters, data from recent studies in the United Kingdom suggest J14 139 a factor 10 to 20 difference in the average rate at which male J14 140 homosexuals and male heterosexuals acquire new partners (see Fig. J14 141 11.16). Overall, therefore, we would expect the doubling rate to be J14 142 smaller - and the doubling time to be longer - for HIV infections J14 143 transmitted along hetero-sexual chains than along J14 144 homosexual ones in developed countries.

J14 145 Whether or not a self-sustaining epidemic can be generated by J14 146 purely heterosexual transmission in developed countries depends on J14 147 the magnitude of the basic reproductive rate for such transmission; J14 148 this quantity is given approximately by formula. Such an J14 149 epidemic is likely to take off if R0 exceeds unity, and J14 150 not otherwise. At present, uncertainty surrounds each one of the J14 151 five quantities entering into this expression for R0 J14 152 (especially the female-to-male transmission probability, J14 153 <*_>beta<*/>'), nor are there enough data to attempt the kind of J14 154 indirect inference of R0 from population-level J14 155 considerations, as we did for homosexual transmission in developed J14 156 countries. We believe that reasonable estimates span a range of J14 157 R0 values from below unity to above unity.

J14 158 Bearing all this in mind, our guess is that, in developed J14 159 countries, HIV could well spread by purely heterosexual J14 160 transmission within relatively promiscuous subgroups. The long-term J14 161 possibility of a much more generally disseminated epidemic seems to J14 162 us to depend crucially on whether a significant fraction of HIV J14 163 infectees remain asymptomatic carriers (never developing AIDS), J14 164 transmitting infection with some low probability essentially for J14 165 the remainder of their sexually active lives. If this is so, then J14 166 quite modest levels of promiscuity could result in HIV infections J14 167 spreading slowly, over many decades, among a population who would J14 168 not think of themselves as promiscuous. In our earlier language, J14 169 such a long duration of infectiousness among a significant fraction J14 170 of infectees corresponds to the basic reproductive rate, J14 171 R0, for heterosexually transmitted HIV being appreciably J14 172 greater than unity.

J14 173 This brings us to a final and important point to do with J14 174 predicting AIDS cases. It might at first seem that the larger the J14 175 fraction of HIV infectees who go on to develop AIDS and die, the J14 176 larger the eventual number of deaths from the epidemic. This is not J14 177 necessarily so. Contrast the following two possibilites, neither of J14 178 which can be completely ruled out by currently available data: in J14 179 Case A, 30 per cent of those infected develop AIDS, remaining J14 180 infectious typically for 5 years and then dying, while the J14 181 remaining 70 per cent of those infected remain asymptomatic yet J14 182 infectious for 30 years; in Case B, 100 per cent of those infected J14 183 develop AIDS, remaining infectious typically for 8 years and then J14 184 dying. Notice immediately that the characteristic duration of J14 185 infectiousness is larger in Case A (0.3x5+0.7x30=22.5 years) than J14 186 in Case B (8 years); this means that the basic reproductive rate J14 187 for HIV infection is almost three times greater under assumption A J14 188 than under B, if the transmission probability is assumed to be the J14 189 same in both cases. The total number of deaths caused by AIDS in J14 190 either case depends on the fraction ever infected with HIV and on J14 191 the fraction of those infected who develop AIDS. For Case A, with J14 192 its substantially larger basic reproductive rate, a larger fraction J14 193 of the population will acquire HIV infection, but a smaller J14 194 proportion (30 per cent) of these will die. In contrast, in Case B J14 195 a smaller fraction will be infected, but all who are will die. We J14 196 simply cannot say whether Case A or Case B will result in a larger J14 197 total number of deaths, until we have made a detailed analysis of J14 198 the non-linear way in which the differences in the basic J14 199 reproductive rates are likely to affect the numbers ever infected. J14 200 The results of such a calculation are sensitive to the amount of J14 201 variability in degrees of sexual activity; the greater this J14 202 variability, the more likely that Case A will lead to more deaths J14 203 than Case B, in defiance of simple intuition.

J14 204 For homosexually transmitted HIV infection in developed J14 205 countries, our preliminary calculations suggest that the total J14 206 number of AIDS deaths increases as f (the fraction of HIV J14 207 infectees going on to develop full-blown AIDS) increases up to J14 208 around 50 per cent or so, but that total deaths remain roughly J14 209 constant, or can even decrease, as f increases toward 100 per J14 210 cent. As emphasized at the outset, current evidence cannot rule out J14 211 f values as low as 20-30 per cent nor as high as 100 per cent. J14 212 J15 1 <#FLOB:J15\>Ultrasound in the investigation of the right J15 2 iliac fossa mass

J15 3 By F. C. Millard, FRCS, M. C. Collins, FRCR and R. J. Peck, J15 4 FRCR

J15 5 Abstract. Patients presenting with a right iliac fossa J15 6 (RIF) mass are a diagnostic problem. The objective of this study J15 7 was to assess the role of ultrasound (US) in their investigation. A J15 8 prospective series of 50 patients presenting with a clinically J15 9 suspected RIF mass was examined by US and the findings correlated J15 10 with the final diagnosis. There was a positive finding in 34 J15 11 patients (68%). Ultrasound correctly identified the organ of origin J15 12 in 33 (97%) and was able to guide the patients' further management. J15 13 In 12 cases no abnormality was found in the RIF, of which 11 had no J15 14 positive findings at follow-up and one was shown to have an J15 15 unrelated abnormality at laparotomy. In four cases the findings J15 16 were due to normal variants. Ultrasound is the imaging modality of J15 17 first choice in patients presenting with a RIF mass.

J15 18 Patients presenting with a right iliac fossa (RIF) mass are a J15 19 clinical diagnostic puzzle. Frequently, these patients have a J15 20 series of investigations including barium enema and intravenous J15 21 urography, which are often inconclusive.

J15 22 When an RIF mass is palpable it should be demonstrable by J15 23 ultrasound (US) and therefore it should be possible to assess its J15 24 organ of origin, and give some idea of its nature. Conversely, a J15 25 normal US may exclude a lesion and prevent further unnecessary J15 26 imaging, particularly if involving ionizing radiation.

J15 27 It was decided to perform a prospective study to assess the J15 28 role of US in the investigation of these patients, and to determine J15 29 any US features of RIF masses which may help establish their J15 30 nature.

J15 31 Method J15 32 A prospective study of 50 consecutive patients, presenting with J15 33 the clinical diagnosis of an RIF mass and referred for US at the J15 34 Royal Hallamshire Hospital, was carried out between January and J15 35 August 1989. Their ages ranged from 18 to 94 years (mean 61 years). J15 36 Thirty-six were female and 14 male.

J15 37 The US was performed by a consultant radiologist using an ATL J15 38 Ultramark 4 scanner with 3 MHz and 5 MHz sector transducers. A full J15 39 abdominal scan was also performed routinely.

J15 40 The findings were confirmed by further imaging, laparotomy or J15 41 in a few cases clinical follow-up.

J15 42 Results J15 43 Thirty-four patients (68%) were shown to have a lesion J15 44 corresponding to the palpable RIF mass. In four patients the mass J15 45 was due to a normal variant (two - low lying kidney, two - J15 46 Riedel's lobe), and in 12 no mass was visualized in the RIF.

J15 47 Of the 34 masses, US correctly identified the organ of origin J15 48 in 33. Eighteen of these were gastro-intestinal and the pathology J15 49 was correctly suggested in 14 (Table I). Seven masses were arising J15 50 from the genito-urinary tract, and US was correct in all seven J15 51 (Table II). In a further nine patients, the mass arose from another J15 52 organ system, and US made the correct diagnosis in seven (Table J15 53 III).

J15 54 tables&captions

J15 55 Of all the cases studied, the site of origin was wrongly J15 56 suggested by US in only one. This was a 71-year-old female patient J15 57 on anti-coagulant therapy, who presented<&|>sic! with lower J15 58 abdominal pain, constipation, and a tender RIF mass. Clinical J15 59 assessment suggested a complicated ovarian cyst. Ultrasound showed J15 60 a complex mixed-echo mass anterior to the bladder. The appearances J15 61 were difficult to interpret, and bowel malignancy was suggested. At J15 62 laparotomy, however, an extensive rectus sheath haematoma was J15 63 found.

J15 64 In the 12 patients in whom no mass was found at US, six J15 65 underwent no further imaging. Of the remainder, five had a barium J15 66 enema, of whom two also had a small bowl enema. No further J15 67 pathology was demonstrated. One patient had the clinical picture of J15 68 large bowel obstruction, and at laparotomy was found to have a J15 69 carcinoma of the transverse colon and no actual mass in the RIF.

J15 70 Of the four patients in whom US found an anatomically normal J15 71 variant of kidney or liver, two had no further imaging and two had J15 72 a barium enema.

J15 73 table&caption

J15 74 As far as can be determined, therefore, of the 12 patients with J15 75 'negative' US there were no false negatives (excluding the one J15 76 patient who later was found to have a carcinoma of the transverse J15 77 colon).

J15 78 Discussion J15 79 In any patient with a palpable RIF mass, we felt US should be J15 80 the primary investigation. This suggestion was also made in a J15 81 recently published retrospective study of US in the palpable J15 82 abdominal mass (Barker & Lindsell, 1990).

J15 83 A normal US may exclude a lesion and prevent further J15 84 unnecessary imaging. Our study confirmed this, US being highly J15 85 specific (100%) in excluding the presence of an RIF mass (Table J15 86 IV). Equally, our study found US to be highly sensitive (100%) in J15 87 correctly identifying patients with an RIF mass (Table IV). We have J15 88 found US to be highly accurate in identifying the organ of origin J15 89 of the mass (97%) as compared with clinical assessment (76%).

J15 90 Ultrasound has always been thought to be of limited value in J15 91 evaluating bowel problems, but gradually it is being shown to have J15 92 much value in this area (Gholkar & Khan, 1989). The US features of J15 93 appendicitis (Puylaert, 1986; Jeffrey et al, 1988), Crohn's disease J15 94 (Sonnenburg et al. 1983) and Campylobacter enteritis (Puylaert et J15 95 al, 1988) have all recently been described. Some ultrasonologists J15 96 feel that a general review of the abdomen during an ultrasound J15 97 examination can provide significant useful information, J15 98 particularly in the detection of occult neoplasms (Price & J15 99 Metreweli, 1988).

J15 100 table&caption

J15 101 The correct pathological diagnosis was made by US in 82% of J15 102 cases. Although we were able to determine bowel as the cause of the J15 103 mass in all the 19 cases in which this proved so, it was more J15 104 difficult to differentiate neoplastic from inflammatory lesions.

J15 105 In conclusion, this prospective study found US, in experienced J15 106 hands, to be accurate in assessing RIF masses, with the advantage J15 107 that it is non-invasive and does not involve the use of ionizing J15 108 radiation. It should be the first investigation of choice in these J15 109 patients.

J15 110 J15 111 A new phantom for mammography

J15 112 By J. Law, PhD

J15 113 Abstract. A new mammography phantom from Du Pont is J15 114 described. It has a wider range of types of detail than in previous J15 115 phantoms, including some which closely simulate breast tissues. J15 116 Experience of its use, and comparisons with an earlier ('Barts') J15 117 phantom, are reported. The Du Pont phantom appears to have greater J15 118 sensitivity to changes in tube kV and focal spot size, and better J15 119 discrimination between different film-screen combinations and J15 120 between films from different breast screening centres.

J15 121 The importance of a suitable phantom, or test object, in J15 122 monitoring and maintaining image quality in mammography is well J15 123 known (see for example Pritchard, 1988). Such a phantom should J15 124 contain details which mimic those which a radiologist looks for in J15 125 clinical mammograms, and may also contain means for assessing J15 126 limiting contrast and resolution. These two approaches are quite J15 127 distinct, since many clinical details depend on a combination of J15 128 both contrast discrimination and resolution. If the latter are J15 129 assessed separately, it is not always a simple matter to combine J15 130 the results to give a single measurement with an obvious J15 131 relationship to the radiologist's needs.

J15 132 It is even more important that films of a phantom show clear J15 133 differences if tube kV, focal spot size, or film/screen/processing J15 134 characteristics change, especially since these may drift gradually. J15 135 In screening mammography in particular, a useful phantom should be J15 136 able to demonstrate such drifts before they become apparent in J15 137 clinical mammograms. (Even if a phantom does not do this, it may J15 138 still have a role in identifying the origins of such drifts.)

J15 139 One phantom which adopts the first approach is that described J15 140 by White and Tucker (1980) and sometimes called the 'Barts' J15 141 phantom. This has been extensively used in a limited number of UK J15 142 centres and found very valuable in at least some. However, it was J15 143 difficult to construct, was never available commercially, and is no J15 144 longer obtainable. It is also only moderately sensitive to changes J15 145 in kV (Law et al, 1989) and, at least for contact films, shows J15 146 little sensitivity to changes in focal spot size. A more sensitive J15 147 replacement has been needed for some time, but until very recently J15 148 only one other phantom has shown any clearly demonstrable J15 149 improvement, and some have failed to match it. That exception was J15 150 the Wisconsin randomized phantom, which is too time consuming for J15 151 routine use. This paper describes results obtained from a new J15 152 phantom with a variety of types of detail, marketed by Du Pont. It J15 153 appears to have most of the features outlined above.

J15 154 Description of phantom details

J15 155 The details are contained within a 'black box' measuring J15 156 approximately 10cm x 8cm x 2cm. This can be placed in various ways J15 157 within an assembly of different sized pieces of Perspex, which all J15 158 together simulate a compressed breast of average size.

J15 159 The details are shown in Fig. 1. They include various objects J15 160 with a range of contrasts, which mimic clinical mammographic J15 161 features and which lend themselves to quantitative scoring, plus J15 162 seven line bar patterns in two directions at right angles, from 8 J15 163 to 20 lp/mm at low contrast, in steps of 2 lp/mm, which can also be J15 164 counted to give a resolution score. In addition, there is a range J15 165 of simulated tissues and structures, much more like breast tissue J15 166 than any of the other features, but which are difficult to score J15 167 except qualitatively or on a scale of (say) 1 to 3. Lastly, over J15 168 small areas the line bar patterns at right angles to each other J15 169 overlap to give an indication of point resolution.

J15 170 The objects having a range of contrasts are of four kinds, and J15 171 each is provided in five decreasing sizes and contrasts. These J15 172 are:

J15 173 (a) Al2O3 specks from 320 to 83 <*_>mu<*/>m J15 174 which mimic microcalcifications;

J15 175 (b) plastic wires from 0.8 to 0.3 mm which mimic fibrous J15 176 structures;

J15 177 (c) carbohydrate or amino-acid spheres from 1.2 to 0.35 mm;

J15 178 (d) discs containing 0.8 mm diameter holes, arranged to form J15 179 the single digits 7, 5, 4, 3 and 2, where each digit indicates the J15 180 depth of the holes.

J15 181 The simulated breast tissues are provided in three clusters of J15 182 four each, which have very low, low, and normal absorption. Many J15 183 also contain random micro-calcifications. There is also a J15 184 single sample about 2cm in diameter representing the appearance of J15 185 'dense breast', and some small fine structures said to represent J15 186 lymph nodes.

J15 187 Methods of use of phantom

J15 188 The arrangement recommended by the manufacturers has been used J15 189 to obtain the results which follow (Fig. 2). One piece of Perspex, J15 190 12cm x 12cm and 3cm thick, is placed directly on the grid cover at J15 191 the patient side, and with the 10cm x 8cm 'black box' on top of J15 192 this. The 4cm space to one side of the box is partly filled by J15 193 another Perspex piece, 12cm x 3cm x 1cm, and a further Perspex J15 194 piece 12cm x 12cm x 1cm is placed above this and the black box. J15 195 Thus the total attenuation is about 5cm Perspex. Behind this J15 196 assembly, i.e. away from the patient, a Perspex step wedge can J15 197 be placed, transverse to the anode cathode direction now normal in J15 198 mammographic tubes. This wedge has eight steps from 5 to 40mm, and J15 199 may be placed on top of another Perspex piece, 12cm x 3cm x 1cm, if J15 200 desired. Lastly, the compression plate is lowered onto the whole J15 201 assembly.

J15 202 figure&caption

J15 203 To 'score' the films, i.e. to obtain a numerical figure of J15 204 merit with some relationship to imagine quality, the manufacturers J15 205 make no recommendations and numerous methods might be derived. To J15 206 obtain the results reported here, the following method was adopted. J15 207 For the four types of feature (a) to (d) above, each of the five J15 208 sizes was given one point if it could be seen quite clearly, with J15 209 edges appearing sharp to the eye, and half a point if some features J15 210 of the type concerned were unambiguously visible but less clear. J15 211 Doubtful appearances were given no score. This scoring was helped J15 212 initially by a radiograph of the 'black box' with no Perspex or J15 213 other scattering material adjacent. Radiologists' binocular viewers J15 214 were always used to cut out surrounding glare.

J15 215 J16 1 <#FLOB:J16\>Moreover, it could be argued that by not ascribing a J16 2 sufficient commercial value to human products this leads to an J16 3 inefficient use of such a resource. It is noteworthy that the Nazi J16 4 concept of lebensunwertes leben ('life unworthy J16 5 of life') led to a pragmatic cost-benefit approach seen in such J16 6 Nazi mathematical texts as 'Mathematics in the Service of National J16 7 Political Education'. While moral aversion to pricing the human J16 8 body may argue against biotechnology as a whole, so long as J16 9 commercial exploitation of the body continues, the argument should J16 10 not deny profit participation to the patient alone.

J16 11 It would be <}_><-|>anomolous<+|>anomalous<}/> not to treat J16 12 body parts as involving property interests by considering them as, J16 13 for example, res nullius. In a sophisticated J16 14 legal system where personal property rights are highly developed so J16 15 as to include such commodities as electricity and gas the list of J16 16 items consisting res nullius is short. J16 17 Furthermore, recognition on res nullius of J16 18 material like limbs severed by assault or accident may be J16 19 inconsistent with the person's interest in its preparation for J16 20 prompt surgical rejoining to his body. Moreover, the law already J16 21 recognises a property interest in body waste as in R v Welsh where J16 22 an accused man who poured his urine sample down the sink was found J16 23 guilty of stealing it from the police department.

J16 24 The recognition of proprietary rights would provide economic J16 25 incentives for patients to consent and, in principle, where a J16 26 donor's non-vital and naturally replaceable body material can save J16 27 the life of a recipient in desperate need, any ethical J16 28 fastidiousness demanding that donation be only gratuitous could J16 29 condemn the sick. One major objection to this is based on the work J16 30 of Titmuss in his book The Gift Relationship (1970) where J16 31 he pointed out that blood donors in the U.S.A. who sell their J16 32 blood, more often than not are poor and often unhealthy. Subsequent J16 33 work has challenged this assumption and, in any case, quality J16 34 control can be solved with adequate screening and independent J16 35 checks on quality. Moreover the problem of economic coercion as a J16 36 motivating factor for donation can apply in very many areas of J16 37 human activity so that Dukeminier has pointed out:

J16 38 "Yet if a charitable motive is so important in judging J16 39 conduct in situations involving a risking of one's life, how can we J16 40 permit men to risk their lives in driving racing cars, in entering J16 41 boxing contests, and in pursuing all kinds of paid risky J16 42 occupations and still object to the paid kidney donor?"

J16 43 Clearly, a decision to sell nonregenerative human body parts J16 44 which will have lifelong implications is not to be taken lightly J16 45 and the law could insist upon elaborate procedures to be satisfied J16 46 as seen currently with the marriage contract or the 'cooling off' J16 47 period seen in consumer credit transactions. Such safeguards are J16 48 necessary in any case even where there is donation as anticipated J16 49 under the Human Organ Transplants Act 1989 since great J16 50 psychological and emotional pressure may be brought to bear where J16 51 there is a needy relative. Notwithstanding that the pressure is J16 52 different from financial inducement, it is nonetheless pressure J16 53 which is hardly less likely to lead to exploitation of those J16 54 vulnerable to it.

J16 55 The need to protect the proprietary interest in human material J16 56 will increase as genetic engineering techniques improve. This J16 57 phenomenon can be illustrated by the facts in U.S. v J16 58 Garber. Here the defendant learned that her blood contained an J16 59 extremely rare antibody. A manufacturer of diagnostic supplies J16 60 induced Garber to undergo frequent withdrawals of her blood to J16 61 isolate the antibody which the company then marketed. Soon J16 62 afterwards other companies approached Garber with offers of even J16 63 more money and eventually Garber profited handsomely, giving rise J16 64 to a federal tax dispute. The point is that once the commercial J16 65 value of Garber's unique antibody became apparent, she was still J16 66 able to bargain for the highest bid. However, when genetic J16 67 engineering improves it may have permitted the original company to J16 68 withdraw enough material in the first place so, in the absence of a J16 69 proprietary right, prejudicing the donor. More recently the Court J16 70 of Appeal of California in John Moore v The Regents of the J16 71 University of California et al concluded that the plaintiff J16 72 did enjoy a property right in his own "blood and bodily J16 73 substances". It is interesting how the influence of J16 74 contract has permeated this area. Before the commercial potential J16 75 of genetic engineering became evident, scientists freely J16 76 transferred cell lines and cell products. As the commercial value J16 77 of the cell lines developed, originators of cell lines found J16 78 written agreements increasingly necessary to protect economic J16 79 rights in their creations. This is reflected in the words of Leon J16 80 Rosenberg, Dean of the Yale School of Medicine, when he said:

J16 81 "The biotechnology revolution has moved us, literally J16 82 or figuratively, from the class room to the board room and from the J16 83 New England Journal to the Wall Street Journal."

J16 84 Given the scope of biotechnology, the question is not whether J16 85 or not to recognise the human body as property but rather what the J16 86 proper scope of the rights and liabilities in the body should J16 87 be.

J16 88 The Application of Property Law Doctrine to Property Rights J16 89 in the Human Body

J16 90 (a) Product Liability

J16 91 (i) Tort

J16 92 The Consumer Protection Act 1987 envisages strict liability for J16 93 any loss caused by using a defective product within the chain of J16 94 distribution. Since the Act deals with liability for a 'product', J16 95 the question is whether an organ or body tissue can constitute J16 96 this? In the U.S.A. blood has been considered a 'product' and in J16 97 principle this should be the case where there is commerce in body J16 98 parts. However, the state of the art defence is product liability J16 99 will be relevant here as long as the defect was not detectable by a J16 100 test. Thus, s.4(1)(e) of the Act provides

J16 101 "... that the state of scientific and technical J16 102 knowledge at the relevant time was not such that a producer of J16 103 products of the same description as the product in question might J16 104 be expected to have discovered the defect if it had existed in his J16 105 products while they were under his control."

J16 106 (ii) Contractual liability

J16 107 The question of whether the commercial supply of body material J16 108 involves a supply of goods or services is highly relevant. In J16 109 Perlmutter v Beth David Hospital there is U.S. authority that J16 110 although units of blood constitute personal property, in therapy, J16 111 when service predominates, what we have is a supply of services J16 112 rather than goods. The Supply of Goods and Services Act 1982 places J16 113 upon the supplier the requirement that he will carry out the J16 114 service with "reasonable care and skill". On the J16 115 other hand the Sale of Goods Act 1979 has implied warranties of J16 116 fitness for purpose which is a more extensive liability. In J16 117 principle, blood or body products should be 'goods' within the J16 118 meaning of the 1979 Act whenever they are dealt with commercially. J16 119 However it may be possible to draw a distinction between blood J16 120 transfusions and skin and bone grafting and fitting cardiac J16 121 pacemakers. In the later cases the transfer of the material may be J16 122 minor compared to the service of installation which will be J16 123 important for the purposes of product liability.

J16 124 (b) Property as defining the limits of medicobiology

J16 125 An adequate legal order should define the limits of J16 126 medicobiology and in this respect, it may be appropriate to J16 127 distinguish between donated and discarded materials. If the doctor J16 128 has a professional interest in the discarded material which affords J16 129 the patient no therapeutic or diagnostic benefit, the patient's J16 130 express consent should be sought beforehand lest abandonment of his J16 131 property may appear to have been effected by the exercise of undue J16 132 influence. This was the view of the Warnock Committee which J16 133 required the "informed consent" of the couple if J16 134 research was to be carried out on "spare embryos", J16 135 an approach which recognises the couple's proprietary interest in J16 136 their embryos.

J16 137 It was argued in John Moore v The Regents of the J16 138 University of California et al that recognition of the donor's J16 139 property rights would prevent unjust enrichment which was J16 140 significant because it was held that there was inequality of J16 141 bargaining power between the doctor and patient. In effect a J16 142 fiduciary obligation was anticipated here which imposed a duty of J16 143 full disclosure, viz a physician should not commercially J16 144 profit by being in breach of trust. The legal inference from this J16 145 is that passivity or silence by the source of body material will J16 146 not, where this is commercially exploited, lead to the conclusion J16 147 that the separated body material is yielded to the hospital or J16 148 researcher. Here we have echoes of the Peel Report which noted:

J16 149 "Where the separation of the fetus from the mother J16 150 leads to the termination of its life there is no statutory J16 151 requirement to obtain the parent's consent for research, but J16 152 equally there is no statutory power to ignore the parent's wishes J16 153 .... We believe the patient must be offered the opportunity to J16 154 declare any special directions about the disposal of the J16 155 fetus."

J16 156 If human body material is treated as the personal property of J16 157 the patient, then it is possible to argue that the researcher J16 158 converts the body material when he exploits its commercial value J16 159 without the express consent of the patient. There is a conceptual J16 160 problem here in the sense that the remedy for conversion sounds in J16 161 damages based upon market value. Since the Human Organ Transplants J16 162 Act 1989 criminalises the active promotion of a sale as being J16 163 contrary to public policy then this type of legislation prima J16 164 facie would seem to prejudice this remedy. However, conversion J16 165 does recognise a special damages heading which the law does not J16 166 consider too remote such as the lost opportunity for receiving J16 167 satisfaction from giving benefit to a specified recipient or cause. J16 168 On the other hand, if full commerciality in body parts were J16 169 recognised this could lead to complex tracing issues as in the J16 170 Moore case, where the physicians researchers<&|>sic! produced from J16 171 the patient's bodily substances without his consent a 'cell line' J16 172 which was predicted to be worth $3 billion by 1990.

J16 173 The essence of the tracing doctrine is that it involves an J16 174 in specie claims to property to which the claimant J16 175 has no property interest other than that which the right provides J16 176 and the property claimed is identifiably linked to the J16 177 misappropriated property. The concept is restitutionary in the J16 178 sense that it is premised on the principle of unjust enrichment, J16 179 the 'unjust' element often being satisfied by the defendant's J16 180 violation of his fiduciary duties. Attempts have been made to J16 181 restrict the reaches of the tracing doctrine as in the case of J16 182 'innocent' converters through the Torts (Interference with Goods) J16 183 Act 1977 and by reference to considerations of 'fairness' as is J16 184 obvious in the dissenting judgments in Boardman v Phipps. J16 185 The arguments are often predicated upon the thesis that the J16 186 defrauded person's recovery of the increased value of the trust J16 187 res is a windfall. Interestingly, tracing analysis is still J16 188 applied but this is restricted to the loss incurred. J16 189 Particular difficulties arise in measuring the loss and the courts, J16 190 in applying tracing rules, do not generally effect a sophisticated J16 191 analysis concerning the question of 'benefit' to the wrongdoer.

J16 192 By concentrating on physical events, tracing conclusively J16 193 presumes that the wrongdoer would not have acquired the 'product' J16 194 but for the conversion. If an inquiry were made into the causal J16 195 events then this could illuminate the problem of quantifying the J16 196 wrongdoer's benefit which under tracing rules has to be disgorged: J16 197 for example, why should it be assumed that the wrongdoer could only J16 198 have obtained the 'product' by conversion? The donor may very well J16 199 have been prepared to sell the material and if this is the case, J16 200 calculation of benefit should instead consist of the value of the J16 201 misappropriation of the material at the time of its occurrence plus J16 202 the interest on that value. If the wrongdoer augments the value of J16 203 the bodily substance by, for example, developing a cell line then J16 204 there are two aspects to the causal approach; firstly, the J16 205 wrongdoer would not have had the asset 'but for' the J16 206 misappropriation; secondly, the traceable product that the owner J16 207 claims may not have been so valuable, 'but for' the wrongdoer's J16 208 efforts. J16 209 J17 1 <#FLOB:J17\>Thalidomide treatment for chronic J17 2 graft-versus-host disease

J17 3 D. HENEY, D. R. NORFOLK, J. WHEELDON, C. C. BAILEY , I. J. J17 4 LEWIS AND D. L. BARNARD

J17 5 Departments of Paediatrics and Haematology, St James's J17 6 University Hospital, Leeds, Department of Haematology, Leeds J17 7 General Infirmary, Leeds

J17 8 Received 29 August 1990; accepted for publication 7 January J17 9 1991

J17 10 Summary. The treatment of chronic graft-versus-host J17 11 disease (GVHD) may present a difficult therapeutic problem. We used J17 12 thalidomide to treat six patients with severe chronic GVHD who J17 13 failed to respond to standard immunosuppressive agents. Four of the J17 14 six patients showed a clear response to thalidomide, with the fifth J17 15 patient showing a partial response. The best results were seen in J17 16 patients with chronic cutaneous GVHD. Two of the patients developed J17 17 neurophysiological evidence of a peripheral neuropathy, associated J17 18 with clinical signs in one patient. Measurement of thalidomide J17 19 plasma levels and pharmacokinetic curves showed a significant J17 20 inter-patient variation. Peak plasma levels varied from 0.47 to J17 21 1.46 <*_>mu<*/>g/ml. Thalidomide has a role to play in the J17 22 management of chronic GVHD and further studies are needed.

J17 23 The immunosuppressive properties of thalidomide have been known J17 24 for over 20 years (Hellman, 1966). Although withdrawn from the J17 25 market in 1961 it has remained available for research and J17 26 restricted clinical use. In the decade that followed its J17 27 withdrawal, thalidomide was shown to be an effective J17 28 immunosuppressive agent for renal allotransplantation in a variety J17 29 of animal models (DeKlerk et al, 1969; Murphy et J17 30 al, 1970). This information was, however, not followed up and J17 31 interest in its use in transplantation declined. Nevertheless J17 32 thalidomide continued to be used in a variety of immunologically J17 33 mediated diseases and has found a place in the management of a J17 34 number of primarily mucocutaneous disorders including lepromatous J17 35 leprosy, apthous stomatitis and discoid lupus erythematosus J17 36 (Barnhill & McDougall, 1982).

J17 37 In 1986 Vogelsang reported the use of thalidomide in the J17 38 treatment of acute graft-versus-host disease (GVHD) in a rat bone J17 39 marrow transplant model. It was shown to be effective in the J17 40 treatment and prevention of acute GVHD, and subsequent data J17 41 confirmed its value in chronic GVHD (Vogelsang et al, J17 42 1986, 1989). These experiments generated a renewed interest in J17 43 thalidomide and were followed by reports of the beneficial effect J17 44 of thalidomide in patients with acute and chronic GVHD (Heney J17 45 et al, 1990; Saurat et al, 1988; McCarthy J17 46 et al, 1988). We report on the use of thalidomide in the J17 47 treatment of chronic GVHD in six patients. We have also modified J17 48 previously described methods for the measurement of plasma levels J17 49 of thalidomide and report pharmacokinetic data on our patients.

J17 50 PATIENTS AND METHODS

J17 51 We treated six patients with thalidomide. Their background J17 52 details are shown in Table I. All patients received HLA matched, J17 53 mixed lymphocyte culture (MLC) compatible bone marrow transplants J17 54 from either a parent or a sibling. They all had severe progressive J17 55 chronic GVHD which had proved unresponsive to standard J17 56 immunosuppressive therapy.

J17 57 Pharmacokinetics. Thalidomide plasma levels were measured J17 58 in four patients. The patients were studied when they were J17 59 established on thalidomide therapy. All patients followed the same J17 60 protocol and were studied in the morning after an overnight fast. J17 61 No solid food was allowed until 2 h after ingestion of a 100 mg J17 62 thalidomide tablet. This was the standard thalidomide dose for each J17 63 of the four patients. Heparinized plasma samples were obtained J17 64 prior to thalidomide ingestion and hourly thereafter for a minimum J17 65 of 6 h and longer where possible. Patient 6 declined to participate J17 66 in the study, and in patient number 3 treatment was withdrawn J17 67 before the study was initiated.

J17 68 Neurophysiological testing. All patients were J17 69 examined at regular intervals for clinical evidence of a J17 70 neuropathy. Four patients underwent neurophysiological testing J17 71 which included nerve conduction studies and electromyography. J17 72 Patient 6 declined the study and patient 4 died before being J17 73 tested.

J17 74 Thalidomide assay. Thalidomide was extracted from J17 75 plasma by solid phase extraction, using a 3 ml C18 Bond Elut J17 76 cartridge (Analytichem International). Thalidomide was measured by J17 77 high performance liquid chromatography (HPLC) (Waters Associates) J17 78 utilizing a LiChrospher C18 5u 25 cm x 4 mm column. The mobile J17 79 phase consisted of 20% Acetonitrile, 80% water plus 750 <*_>mu<*/>l J17 80 of 30% HCL per litre. The flow rate was 1 ml/min and detection was J17 81 at 230 nm. Acetophenetidin was used as an internal standard.

J17 82 table&caption

J17 83 Standards were prepared by adding purified thalidomide J17 84 (Grunenthal GmbH) to drug-free plasma. Thalidomide concentrations J17 85 were calculated by comparing the peak height ratio of thalidomide J17 86 and internal standard in study plasma samples with those from the J17 87 calibration curve. The analytical recovery of thalidomide from J17 88 plasma was 92% at 1 <*_>mu<*/>g/ml. Using the above chromatographic J17 89 conditions, thalidomide and internal standard were well separated J17 90 with approximate retention times of 12 and 17 min respectively (see J17 91 Fig 1). The limit of detectibility was 0.1 <*_>mu<*/>g/ml. All J17 92 solvents were supplied by Rathburn Chemicals, Walkerburn, Scotland. J17 93 The thalidomide standard was kindly donated by Grunenthal GmbH, J17 94 Germany. The thalidomide tablets were supplied by Champion J17 95 Farmaceutica Ltd, S<*_>a-tilde<*/>o Paulo, Brazil. All other J17 96 reagents came from Sigma Chemicals, Poole, U.K.

J17 97 RESULTS

J17 98 Clinical response

J17 99 The dosage of thalidomide and the clinical repsonse is shown in J17 100 Table II.

J17 101 Patient 1 developed disabling chronic GVHD with extensive J17 102 scleroderma which necessitated his attending a school for the J17 103 physically handicapped. Thalidomide 200 mg daily was introduced but J17 104 immediately caused significant somnolence. The dose was reduced to J17 105 100 mg daily, with no further somnolence, and a rapid clinical J17 106 response over several weeks. The improvement has been sustained J17 107 over a 24 month period allowing him to partake in normal J17 108 activities, with the discontinuation of his azathioprine and a J17 109 reduction in his prednisolone from 15 mg to 5 mg alternate days.

J17 110 Patient 2 also developed severe sclerodermatous chronic GVHD J17 111 with marked pain and joint contractures. His mobility was greatly J17 112 impaired and he had difficulty attending school. A clear J17 113 improvement of both skin texture and mobility was noted within 6 J17 114 weeks of starting thalidomide. There was a continued improvement J17 115 over a 24-month period allowing him to complete his schooling and J17 116 obtain full-time employment. His prednisolone was discontinued. At J17 117 24 months, neurophysiological testing revealed a moderately severe J17 118 sensorimotor neuropathy, although he had no clinical signs or J17 119 symptoms. The thalidomide was tailed off with no deterioration of J17 120 his chronic GVHD.

J17 121 figure&caption

J17 122 Patient 3 had severe multi-system chronic GVHD. She experienced J17 123 moderate drowsiness when started on thalidomide 100 mg b.d., and J17 124 the dose was reduced to 100 mg nightly. A marked improvement in her J17 125 cutaneous GVHD was evident within 8 weeks, and this continued over J17 126 a 12-month period. Her azathioprine was discontinued and J17 127 prednisolone reduced from 7.5 mg to 5 mg daily. She continued to J17 128 have intercurrent respiratory infections but these were easily J17 129 controlled. After 12 months of therapy she presented with numbness J17 130 of both feet with clinical and neurophysiological evidence of a J17 131 peripheral neuropathy. Thalidomide was discontinued and was J17 132 followed immediately by a relapse of her cutaneous GVHD, with J17 133 thickening and tightening of her skin. Her parasthesiae continued J17 134 to worsen over a 3-month period but subsequently settled and 12 J17 135 months later she has no clinical evidence of peripheral neuropathy. J17 136 Her GVHD also stabilized.

J17 137 table&caption

J17 138 Patient 4 developed severe mucocutaneous chronic GVHD with J17 139 erythema and desquamation of her skin, dystrophic nails and oral J17 140 ulceration and lichen planus. Thalidomide therapy was associated J17 141 with an improvement in both her skin and oral mucosa. Her nais fell J17 142 out shorty after starting thalidomide but thereafter started to J17 143 grow again. Thalidomide therapy was continued for 6 months. During J17 144 this time her azathioprine was discontinued and prednisolone J17 145 reduced from 10 mg to 5 mg daily. Cyclosporine was maintained at J17 146 100 mg b.d. She then presented with a varicella zoster pneumonitis J17 147 which proved rapidly fatal despite appropriate therapy.

J17 148 Patient 5 developed multisystem chronic GVHD with recurrent J17 149 respiratory infections, sclerodermatous skin changes, vaginal J17 150 stenosis, sicca syndrome and hepatic GVHD. She was started on J17 151 thalidomide 200 mg nightly but this was associated with undue J17 152 drowsiness and the dose was reduced to 100 mg nightly. There was a J17 153 gradual improvement in her oral mucosa and also her skin which J17 154 remained dry but otherwise stable. Problems with vaginal stenosis J17 155 and respiratory infections persisted. Her azathioprine was J17 156 discontinued although prednisolone 10 mg daily and cyclosporine 50 J17 157 mg b.d. was maintained. Thalidomide therapy was continued for 18 J17 158 months. While on thalidomide she developed pre-malignant vaginal J17 159 changes necessitating a hysterectomy and vaginectomy, and also J17 160 pre-malignant changes of her peri-anal skin. She then presented J17 161 with lymphadenopathy of her neck and a diagnosis of stage IA J17 162 nodular sclerosing Hodgkins disease was made for which she is J17 163 receiving local radiotherapy.

J17 164 Patient 6 developed chronic respiratory GVHD with both J17 165 restrictive and obstructive lung disease. He also had sicca J17 166 syndrome and oral lichen planus. There was no response to J17 167 thalidomide therapy over a 15-month period and the thalidomide has J17 168 been discontinued. No side effects were noted.

J17 169 Neurophysiological testing

J17 170 Patient 2 had ill-formed, low amplitude, sensory action J17 171 potentials of both upper and lower limbs. Motor conduction J17 172 responses were also of low amplitude with conductive delay. The J17 173 electromyogram (EMG) showed mild denervative changes. These J17 174 findings confirm a moderately severe axonal neuropathy with sensory J17 175 and motor symptoms. Patient 3 had low amplitude sensory and motor J17 176 action potentials, consistent with a mild generalized sensorimotor J17 177 neuropathy. Patients 1 and 5 had no evidence of a neuropathy.

J17 178 Thalidomide pharmacokinetics

J17 179 Thalidomide plasma levels of four patients are shown in Fig 2. J17 180 Base line levels varied between 0.26 and 0.50 <*_>mu<*/>g/ml, with J17 181 Cmax (peak levels) varying from 0.47 to 1.46 <*_>mu<*/>g/ml. The J17 182 rate, extent and duration of the rise of thalidomide levels J17 183 following ingestion also varied. Peak to trough ratios were 2.9, J17 184 2.1, 2.0 and 1.4 for patients 1, 2, 4 and 5 respectively. Thus J17 185 patient 5 had only a minimal rise while patients 1 and 4 had a more J17 186 pronounced rise. In general there was a gradual return to base line J17 187 levels at approximately 6-8 h.

J17 188 figure&caption

J17 189 DISCUSSION J17 190 Four of the six patients showed a clear response to J17 191 thalidomide. The best responses were seen in those patients with J17 192 chronic cutaneous GVHD. Patients 1, 2 and 3 all showed a marked J17 193 improvement in skin texture and mobility. Oral mucocutaneous J17 194 symptoms also improved, most noticeably in patients 3 and 4. The J17 195 poorest reponses were seen in patients with multi-system disease, J17 196 in particular where there was a combination of active respiratory J17 197 and mucocutaneous disease. However, thalidomide may still be of J17 198 value in this situation, as the cutaneous GVHD appeared to J17 199 stabilize in patient 5, and there are reports in the literature of J17 200 respiratory disease responding to thalidomide (McCarthy et J17 201 al, 1988).

J17 202 In three patients an improvement was noted within 6-8 weeks of J17 203 starting therapy. The improvement usually continued over a period J17 204 of months. In all cases the improvement was sustained and there was J17 205 no evidence of recurrence while on thalidomide. In patient 3 J17 206 thalidomide was discontinued abruptly because of peripheral J17 207 neuropathy and a recurrence of symptoms was noted immediately. In J17 208 patient 2 thalidomide was discontiuned gradually with no relapse of J17 209 symptoms.

J17 210 Thalidomide has a number of early side effects which are seldom J17 211 severe and are reversible when treatment is interrupted (Grosshans J17 212 & Illy, 1984). They include drowsiness, dizziness, headaches, J17 213 constipation, nausea, dryness of mouth and skin, erythematous skin J17 214 eruptions and oedema. Less common side effects include allergic J17 215 vasculitis and thrombocytopenic purpura. Three of our patients J17 216 experienced drowsiness at a dose of 200 mg daily and necessitated a J17 217 reduction of the dose. No other early side effects were noted.

J17 218 Apart from the well-known teratogenic effects, the most J17 219 important side effect is the occurrence of polyneuropathies. J17 220 Attention was first drawn to the neurotoxic effects of thalidomide J17 221 by Florence (1960). These patients had used thalidomide as a J17 222 sedative in varying dosages from 50 to 600 mg at night and usually J17 223 for prolonged periods from 6 months to years. In affected subjects J17 224 sensory symptoms predominated with parasthesia of hands and feet, J17 225 pallor and coldness of fingers and toes and muscular cramps of the J17 226 extremities. The overall incidence of neuropathy remains difficult J17 227 to assess. The manufacturers originally suggested an incidence of J17 228 0.5% in patients who had taken the drug for 2 months or more. J17 229 Recent reports, however, suggest the incidence may be much higher, J17 230 in the order of 25% or more, especially if neurophysiological tests J17 231 are performed (Clemmensen et al, 1984; Hess et J17 232 al, 1986). J17 233 J18 1 <#FLOB:J18\>Axiom of Reducibility

J18 2 For every concept Fi(x) of objects of type J18 3 n, for i>n, there exists an extensionally equivalent J18 4 concept formula of type n+1, i.e. a concept J18 5 formula such that formula.

J18 6 This axiom asserts that increasingly complex definitions do not J18 7 give any new powers of classification - all classes of individuals J18 8 can be marked out with basic concepts of type 1, and all classes of J18 9 objects of type n can be marked out by concepts of type J18 10 n+1. This means that when using a version of Frege's J18 11 definition of 'natural number' but with its quantifier limited to J18 12 ranging only over concepts whose type is next above that of the J18 13 numbers being defined one will have achieved almost the same effect J18 14 as quantifying over all concepts, of whatever type, since one will J18 15 have quantified over all the classes that they could delimit.

J18 16 This is again a principle which has to be admitted not to be a J18 17 purely logical principle. It is not something which can be proved, J18 18 or even, strictly speaking, formulated as a logically legitimate J18 19 statement, for it involves generalizing over concepts of more than J18 20 one type and is thus not itself logically correct. Russell is by J18 21 this means able to avoid the inconsistencies of Frege's system and J18 22 retain many of the essentials of his account of arithmetic but only J18 23 at a price and only in a way which must lead one to ask whether he J18 24 has not demonstrated that and why arithmetic is not reducible to J18 25 logic, rather than that it is. To evaluate the position further it J18 26 is necessary to inquire into Russell's justification for adopting J18 27 the Vicious Circle Principle.

J18 28 EMPIRICISM, LOGICAL POSITIVISM AND THE STERILITY OF J18 29 REASON

J18 30 If one were to adopt a strongly Platonist position, saying that J18 31 numbers, classes, concepts and functions have an existence which is J18 32 independent of us and our mathematical activities, then the Vicious J18 33 Circle Principle could not be justified as a general logical J18 34 principle. A definition such as

J18 35 Ben Cullin <*_>unch<*/> the highest mountain in the Hebrides

J18 36 which specifies the reference of a name by quantifying over a J18 37 class in order to select an individual member from that same class, J18 38 violates the Vicious Circle Principle. Nonetheless, provided that J18 39 we are confident that there must be a highest mountain in the J18 40 Hebrides (there are mountains there, all mountains are comparable J18 41 with respect to height and it is extremely unlikely that there are J18 42 two of exactly the same height), the definition would normally be J18 43 regarded as legitimate, even though it does not necessarily provide J18 44 sufficient information to enable someone to identify the mountain J18 45 in question. This definition would be regarded as legitimate J18 46 because the totality by reference to which the name is defined J18 47 exists as a determinate collection of objects prior to and J18 48 independently of the definition. The definition does not introduce J18 49 a new object to add to the totality in question, but names an J18 50 already existing member. For the Platonist all definitions of J18 51 (words referring to) numbers, classes, concepts and functions will J18 52 be of this kind; they are definitions whose function is to link J18 53 linguistic expressions with pre-existing entities. So for the J18 54 Platonist, violation of the Vicious Circle Principle cannot, in and J18 55 of itself, invalidate a definition. This means that he must look J18 56 elsewhere for a solution to Russell's paradox. One route available J18 57 to him is to treat Russell's paradox as proving two truths about J18 58 classes: (i) no class belongs to itself, and (ii) there is no class J18 59 of all classes. Alternatively he could treat it as indication of a J18 60 need to distinghish between classes and sets; mathematicians deal J18 61 with sets and the nature of the set-theoretic universe is captured J18 62 by the axioms of set theory whose truth is recognized by J18 63 set-theoretic intuition. These routes are available, but they J18 64 represent a departure from the logicist programme. Once appeal to J18 65 set theoretic intuition is allowed it is not clear why intuition of J18 66 numbers and geometric intuition should not also be allowed.

J18 67 Russell's advocacy of logicism and his reasons for thinking the J18 68 Vicious Circle Principle to be a general logical principle are, on J18 69 the other hand, both grounded in his empiricism. His early, more J18 70 rationalist, philosophy gave way to empiricism at the same time J18 71 that his account of mathematics became more closely tied to formal J18 72 logic. For empiricists the only reality is the empirical world, the J18 73 world with which we are acquainted through sense experience. J18 74 Abstract objects, such as numbers or classes, have no independent J18 75 existence but must be a product of our linguistic or mental J18 76 constructions. Since there is no realm of mathematical reality of J18 77 which to have mathematical intuitions, true mathematical J18 78 statements, if they can be known to be true independently of J18 79 experience, must be analytic truths, having their origin in the way J18 80 in which the abstract objects are constructed. As Russell concurred J18 81 with Frege's desire to separate sharply between the logical and the J18 82 psychological, he too shied away from any appeal to mental J18 83 constructions. The objectivity of mathematics requires that the J18 84 meaning of mathematical language cannot be given by reference to J18 85 ideas or mental constructions. Rather the meaning of all J18 86 expressions which apparently refer to abstract objects must be J18 87 shown, by the provision of suitable definitions, to be logical J18 88 constructs (fictions) built up from constituents of the empirical J18 89 world. The assignment of meaning to names can then take one of two J18 90 forms. Either the name is simply a label for an empirically given J18 91 object, in which case Russell called it a logically proper name, J18 92 one which has a reference but no sense, or it is a descriptive J18 93 expression (a definite description), which identifies an object via J18 94 its relations to other given objects, or via its mode of J18 95 construction out of given objects.

J18 96 The only kind of definition which can be provided for a J18 97 logically proper name is an ostensive one - the label is attached J18 98 by pointing to the object and uttering the name. Thus only J18 99 empirically given entities can have logically proper names. Here J18 100 Russell disregards Frege's injunction to treat a name as having J18 101 meaning only in the context of a sentence; logically proper names J18 102 can function in isolation. Definite descriptions, on the other J18 103 hand, are always defined in a sentential context. Such expressions J18 104 are required either for picking out empirical individuals via their J18 105 relation to other empirical individuals, or for showing how new J18 106 entities can be constructed out of those which are empirically J18 107 given or have previously been defined. In the former case one will J18 108 be using a description to pick out an object from a given class, J18 109 but in the latter one will not be picking out an already existing J18 110 entity since one will be introducing a logical fiction. The two J18 111 cases need to be treated differently if the Vicious Circle J18 112 Principle is not to seem unduly and unjustifiably restrictive.

J18 113 Definite descriptions are treated as having meaning only in the J18 114 context of a sentence. According to Russell's theory of definite J18 115 descriptions

J18 116 Ben Cullin is the highest mountain in the Hebrides J18 117 <*_>unch<*/>

J18 118 There is a highest mountain in the Hebrides, there is at most J18 119 one and it is Ben Cullin.

J18 120 Here there is quantification over the domain to which the J18 121 descriptively identified object belongs, but the definition makes J18 122 it quite clear that the function of the descriptive phrase is to J18 123 pick out an object from that domain and that no object is being J18 124 added to it. Moreover, the descriptive phrase does not have to have J18 125 a reference for the sentence containing it to make sense. The J18 126 sentence is simply false if the descriptive identification J18 127 fails.

J18 128 Logical fictions are constructed by collecting already defined J18 129 objects into classes which are themselves regarded as mere J18 130 fictions. To this end Russell gives contextual definitions showing J18 131 how apparent reference to classes can be eliminated. For example, J18 132 the simplest eliminations would be

J18 133 1 a belongs to the class of Fs <*_>unch<*/> F(a).

J18 134 2 the class of Fs = the class of Gs <*_>unch<*/> J18 135 formula.

J18 136 In (2) the Vicious Circle Principle requires that the domain of J18 137 quantification on the right hand side not include the class of Fs J18 138 or of Gs. Moreover, it requires that the concepts F(x) and G(x) not J18 139 themselves contain any quantification over a domain which includes J18 140 themselves or a domain which includes entities of higher type than J18 141 their arguments. This means that the actual contextual definitions J18 142 used in Principia have to be somewhat more complicated. The J18 143 basic idea behind the remaining clauses, which cover the cases J18 144 where one wants to say things about classes or to form classes into J18 145 further classes, is that every statement about a class is really a J18 146 statement about its members (since by (1) and (2) classes are J18 147 identical when they have the same members). If this is so, then an J18 148 apparently simple statement about a class can, in principle, be J18 149 written as a logically complex statement concerning its members. J18 150 So, for example,

J18 151 'n = the number of Fs' becomes

J18 152 'the class of Fs belongs to the class of all n-membered J18 153 classes'

J18 154 which in turn reduces to

J18 155 formula.

J18 156 It is this view of classes as in principle eliminable logical J18 157 constructs which justifies the Vicious Circle Principle. An entity J18 158 (whether object or concept) cannot be constructed out of itself, J18 159 but only out of entities previously given or constructed. Similarly J18 160 a verbal expression cannot be defined in terms of itself, but only J18 161 from expressions which are given as primitive or have previously J18 162 been defined.

J18 163 But if this is the justification for the Vicious Circle J18 164 Principle, then the introduction of the Axiom of Reducibility looks J18 165 even more embarrassing than it seemd at first sight. Not only is it J18 166 a non-logical assumption needed for the derivation of arithmetic, J18 167 and hence an admission of failure in the logicist programme, but, J18 168 being an existential axiom, it suggests a return to some form of J18 169 Platonism. This would undercut the Vicious Circle Principle which J18 170 formed the foundation of Russell's theories of logical types which J18 171 in turn represented his solution to the paradoxes. This is J18 172 essentially the argument presented by G<*_>o-umlaut<*/>del (1944). J18 173 If we are entitled to assume the existence of concepts J18 174 independently of their definition, as the Axiom of Reducibility J18 175 seems to assert, then, as outlined above, there is no good reason J18 176 for thinking the Vicious Circle Principle to be a general logical J18 177 principle. However, it is also possible to read the Axiom of J18 178 Reducibility as an assertion of faith in the logicist programme in J18 179 mathematics, and more generally of the philosophical position of J18 180 the logical positivists. This does not wholly exonerate Russell. J18 181 There is a curious circularity in having the success of a position J18 182 rest on an assertion to the effect that it can be successfully J18 183 carried out. But, in the first place, it is not clear that any J18 184 philosophy can ever avoid this kind of circularity and, in the J18 185 second, it will at least defend Russell against the charge of J18 186 having introduced an element which renders his position J18 187 philosophically incoherent.

J18 188 The position of the logical positivists, unlike that of some J18 189 earlier empiricists, such as Berkeley and Hume, is not J18 190 phenomenalist. The logical positivists had, as Russell (1919 p.170) J18 191 said, "a robust sense of reality", i.e. a belief J18 192 that the empirical world exists independently of us and our J18 193 experiences - we do not live only in a world of ideas. There is a J18 194 sharp distinction to be drawn between Hamlet and Napoleon, between J18 195 unicorns and lions. The former are fictions, existing only as J18 196 ideas, whereas the latter are empirically real and have an J18 197 existence beyond our ideas. In this respect their position has more J18 198 in common with Locke's empiricism than with Hume's. Sense J18 199 experience affords genuine knowledge of the empirical world and is J18 200 our only route to knowledge of it. Since this is the only reality J18 201 of which we have an experience, empirical reality is the only J18 202 reality which we have any basis for supposing to exist J18 203 independently of us. (For further discussion and detailed arguments J18 204 concerning the grounds for belief in the existence of an external J18 205 world see, for example, Russell (1912).) The only objective J18 206 knowledge there can be is factual knowledge of this independently J18 207 existing reality.

J18 208 J19 1 <#FLOB:J19\>However in the absence of knowing what these are we J19 2 run the same experiment as in case a.

J19 3 This may be an impossible generalization anyway. Problems of J19 4 differences between the selection of experimental and real groups J19 5 may invalidate it. The unknown numbers in each category in the J19 6 target population make it impossible to estimate the main effects J19 7 when an interaction is present. If the effect of one of the J19 8 variables depends on the level of the other then the overall effect J19 9 of the first variable could be any weighted average of its effects J19 10 at each level on the second variable. Indeed it should be the J19 11 weighted average of its effects over all naturally occurring levels J19 12 of the second variable including those not used in the experiment. J19 13 Since this design is not OSCAR, the numbers of observations in the J19 14 experimental conditions do not correspond to population J19 15 frequencies. Thus an EWO analysis is inappropriate. For the main J19 16 effects we shall have to confine ourselves to a simple effects J19 17 analysis. The alternative is to give up attempting to generalize to J19 18 existing groups and analyse the data as for case a.

J19 19 Case c. The group study

J19 20 <*_>black-square<*/>We are interested in the effects of age and J19 21 sex on memory loss in head injured patients. Samples of male and J19 22 female head injured patients in each of a number of age groups are J19 23 obtained.

J19 24 Here the design could be regarded as OSCAR. If, because of J19 25 sampling bias, we were not prepared to make the assumption that the J19 26 design was OSCAR, it would be difficult to justify the assumption J19 27 that the design was ROWT. It is hard to imagine that there is a J19 28 bias in the sample which affects only the age and sex of the J19 29 sampled subjects and not any other related characteristics. The J19 30 numbers in the different groups reflect differences in that real J19 31 population. The EWM analysis of main effects depends arbitrarily on J19 32 the other factors in the study. We wish to test whether the mean of J19 33 male patients differs from the female mean and whether the means in J19 34 the different age groups differ. Here and EWO analysis is J19 35 appropriate for both main effects. An EWM analysis would only be of J19 36 interest if one wished to generalize to the hypothetical results of J19 37 a study on this population which used a balanced design.

J19 38 Case d. The first hybrid case

J19 39 <*_>black-square<*/>We are interested in differences in reading J19 40 between a dyslexic group (assumed to be a random sample from a J19 41 population of dyslexics) and a control group, together with the J19 42 effects of social class on the two groups. The control group is J19 43 assumed to be a random sample of children from the same age J19 44 cohorts.

J19 45 Here we assume we have a design which is OSCAR with respect to J19 46 the social class factor but is ROWT with respect to the other J19 47 factor. The unequal numbers in each social class would therefore J19 48 reflect the true population differences. In this case the EWM main J19 49 effect of dyslexics versus controls would be arbitrarily related to J19 50 the categories of social class we had defined. Only an EWO analysis J19 51 of the differences between the dyslexics and normals generalizes to J19 52 the population of dyslexics. An EWM analysis refers to a population J19 53 of dyslexics equally distributed across the social classes. As the J19 54 relative sizes of the dyslexic and control groups are arbitrary, a J19 55 simple effects analysis of social class effects in dyslexics and J19 56 normals seems appropriate, or perhaps an EWM analysis of the J19 57 differences between classes generalizing to the results of balanced J19 58 design replications of this study.

J19 59 Case e. A hybrid study with a matched J19 60 design

J19 61 <*_>black-square<*/>We are again interested in differences in J19 62 reading between a dyslexic group (assumed to be a random sample J19 63 from a population of dyslexics) and a control group, together with J19 64 the effects of social class on the two groups. Here the control J19 65 group is a sample of children from the same age cohorts matched for J19 66 social class.

J19 67 Here the design is OSCAR for the dyslexic group with respect to J19 68 social class but in other respects it is ROWT. An EWO analysis of J19 69 dyslexics versus controls tells us whether dyslexics differ from a J19 70 sample of matched normals, whereas an EWM analysis tells us whether J19 71 we could expect this difference in a balanced replication. Again a J19 72 simple effects analysis or an EWM analysis should be used to J19 73 investigate social class.

J19 74 To sum up, in two factor designs EWO tests of main effects are J19 75 recommended for OSCAR designs generalizing to natural populations. J19 76 In two factor balanced designs or where the interaction is assumed J19 77 to be negligible, EWM tests of main effects can be used. Here one J19 78 is generalizing to hypothetical results from balanced replications J19 79 of the experiment. In all other cases no overall test of the main J19 80 effects is possible and one should resort to simple effects J19 81 analyses.

J19 82 A WORKED EXAMPLE

J19 83 Suppose we obtained a random sample of students at a particular J19 84 university and classified them by sex and whether they were J19 85 studying predominantly arts or science subjects. The attitude of J19 86 these students to proposed changes at their university was then J19 87 measured on a twenty point scale; the results are in Table 3.

J19 88 tables&captions

J19 89 This is an OSCAR design and an EWO analysis of main effects and J19 90 an EWM analysis of the interaction is appropriate. The results of J19 91 the analysis appear in Table 4.

J19 92 It can be seen that both the main effects are significant; sex J19 93 at the 0.01 level and faculty at the 0.05 level. This means that we J19 94 can conclude that in this university population males have greater J19 95 attitude scores than females and science students have greater J19 96 attitude scores than arts students. Since the imbalance in the J19 97 study implies that the hypotheses being tested are related one J19 98 could ask the question: "Suppose we performed a balanced J19 99 replication could we still expect the differences to be J19 100 present?" This question could be answered by applying an J19 101 EWM analysis to the main effects. The results of the analysis J19 102 appear in Table 5. It now appears that the main effect of the J19 103 subject area is no longer significant. The F (1, 32) testing J19 104 for an effect of faculty has changed from 7.34 in the EWO analysis J19 105 to 1.22 in the EWM analysis, from an effect almost significant at J19 106 the 0.01 level to barely a trace of an effect. The conclusions of J19 107 the analyses are that sex differences have been established in both J19 108 OSCAR designs and ROWT designs balanced with respect to sex and J19 109 subject area; whereas the subject area difference has only been J19 110 established with respect to OSCAR designs. It has not been J19 111 demonstrated that we could expect to find a subject area difference J19 112 in samples balanced with respect to sex.

J19 113 Now suppose the same data had been obtained but this time the J19 114 arts and science groups were random samples of arbitrary size from J19 115 the two faculties. This time the design is OSCAR with respect to J19 116 sex but ROWT with respect to subject area. Using the ROWT analysis J19 117 in Table 5 to generalize to the results of balanced designs is J19 118 still legitimate. If we want to generalize to the population of J19 119 arts and science students the tests applied to the subject area J19 120 effect and the interaction in Table 4 still stand. However the main J19 121 effect of sex cannot be tested without assuming the interaction is J19 122 zero. Instead, separate simple effects analyses of sex in the arts J19 123 and in the science groups should be performed. These appear in J19 124 Table 6. From the analysis we can conclude that there is a J19 125 difference in sexes in both the science and the arts groups at the J19 126 0.01 level. From this we can be sure that there will be an overall J19 127 effect of sex regardless of the distribution of arts and science J19 128 students. If the sex effect had not been significant in both the J19 129 groups we could not have concluded that there would be an overall J19 130 effect of sex unless we assumed that the interaction between sex J19 131 and faculty was zero.

J19 132 table&caption

J19 133 The above analyses were performed using BMDP4V which has J19 134 facilities for all the tests discussed so far. SPSSX also J19 135 performs such analyses. If you do not have a big statistical J19 136 package to hand, the EWO analysis of main effects can be performed J19 137 by treating the analysis as one way and ignoring the other factors. J19 138 Several methods exist for approximating to the EWM analysis when J19 139 the design is unbalanced. For example, Bartlett (1937) used a J19 140 procedure involving a multiple covariance analysis, and Winer J19 141 (1971) gives a harmonic means analysis. These procedures may be J19 142 useful where one does not have access to a modern computing J19 143 environment but the EWM analysis is available on SPSSX, J19 144 BMDP and SAS and is to be performed.

J19 145 FILLING IN FOR MISSING DATA

J19 146 So far we have assumed that the only results we have are the J19 147 observed data. Sometimes we know in addition that it was intended J19 148 to make more observations in a particular condition but because of J19 149 such factors as experimenter error, equipment failure, subject J19 150 refusal, or failure to understand instructions, a known amount of J19 151 data was lost. Missing data can only be handled by knowing the J19 152 nature of the mechanism giving rise to it. The simplest case is J19 153 where the data are Missing Completely At Random (MCAR). Here each J19 154 observed value is equally likely to go missing. A more complex case J19 155 occurs when some conditions are more likely to produce missing data J19 156 than others but within each condition each observation is equally J19 157 likely to be missing. This we have called Missing At Random Only J19 158 Within Treatment (MROWT). Yet more complex cases arise when the J19 159 missing data depend on the observed score (for example, high values J19 160 are more likely to go missing). This changes the nature of the J19 161 distribution of scores and one must either use nonparametric J19 162 analyses or specify the distribution of the missing data. The case J19 163 where missing data depend on the combination of the treatment and J19 164 the value of the score makes the concept of average differences, in J19 165 both observed and missing data, between conditions, exceedingly J19 166 hypothetical. The more complex the assumptions necessary for the J19 167 analysis the better it is to restrict oneself to inferences about J19 168 observable data. Little & Rubin (1987) describe some such models J19 169 which make complex assumptions about missing data. These will be J19 170 discussed below.

J19 171 All the analyses we have considered assume that the data are J19 172 MCAR or MROWT. This being so the missing data are ignorable, so why J19 173 not just analyse the data using the procedures suggested earlier? J19 174 There are three answers. The analyses of unbalanced designs are J19 175 less robust against violations in the underlying assumptions J19 176 (Milligan et al., 1987) and imputing new data to restore J19 177 the balance would also restore the robustness. Secondly, in the J19 178 analysis of OSCAR designs, we may have tried to obtain data from a J19 179 known number of subjects in each condition but have been J19 180 unsuccessful in certain cases. Here it may be preferable to use the J19 181 known number of subjects in each condition rather than the number J19 182 of observations for which data exist, since this is a more accurate J19 183 estimate of the relative sizes of the conditions in the target J19 184 population. Finally, in appropriate circumstances, knowledge of a J19 185 covariate can tell us something about the likely values of missing J19 186 data. This information can be used to increase the precision of the J19 187 analysis.

J19 188 The simplest way of handling missing data is to replace them J19 189 with something else. Perhaps the most obvious way is to collect J19 190 more data. Less desirable ad hoc methods include J19 191 selecting data at random from previous studies or randomly sampling J19 192 the current data and duplicating the selected points. An J19 193 alternative procedure is to substitute the mean of the condition J19 194 where the missing data occurred. This method reduces the error J19 195 variance but this can be corrected using the actual numbers of J19 196 observed data points. When this is done the analysis is equivalent J19 197 to weighting the sample means with the frequencies of missing and J19 198 observed rather than just the frequency of the observed data.

J19 199 J20 1 <#FLOB:J20\>HEXAGONS, CONICS, A5 AND J20 2 PSL2(K)

J20 3 R. H. DYE

J20 4 1. Introduction

J20 5 1.1 When a group occurs geometrically there is a natural J20 6 expectation that the geometry should account for its subgroups. J20 7 Perhaps this is particularly so for the groups J20 8 PSL2(K) and PSL3(K) that arise from J20 9 those simplest of geometric objects, the projective line PG (1, J20 10 K) and the projective plane PG(2, K) over a field K. J20 11 Yet a simple graphic geometrical raison d'<*_>e-circ<*/>tre for the J20 12 existence, when it occurs, of the alternating group A5 as J20 13 a subgroup of PSL2(K) does not seem to be known. The J20 14 initial aim of this paper is to discover a picturesque object that J20 15 accounts for this occurrence of A5. We shall show that if J20 16 PG (1, K) is represented as a conic C in PG(2, K), J20 17 then the pertinent object is a certain type of hexagon, J20 18 distinguished by special concurrencies of its edges. For us, a J20 19 hexagon is a set of six points, no three collinear, in PG(2, J20 20 K): we call the six points the vertices of the hexagon, and J20 21 their 15 joins in pairs its edges. And there is a natural J20 22 concurrent aim: obtain the action of PSL2(K) on J20 23 these hexagons, and determine their geometry in relation to C. J20 24 There is, since A5 is a subgroup of A6, a J20 25 consequent aim: discover a configuration of these hexagons that J20 26 accounts, when it occurs, for A6as a subgroup of J20 27 PSL3(K). Such a configuration has been found, but to J20 28 keep this paper within reasonable bounds the details will form a J20 29 sequel. However, to paint the whole picture we give a brief J20 30 description in Section 1.6.

J20 31 1.2 It is convenient to pause and recall some facts about J20 32 C. Its group PO3(K) is a copy of J20 33 PGL2(K), and the commutator subgroup formula J20 34 is, except when K is the field of two elements, J20 35 PSL2(K). Now P<*_>OMEGA<*/>3(K) is J20 36 transitive on the points of C and on its chords. If K J20 37 does not have characteristic 2 then C has an associated J20 38 non-degenerate polarity. In that case the poles of the chords are J20 39 external points, each lying on two tangents. There may also be J20 40 internal points lying on no tangents; these internal points may not J20 41 be in a single orbit under PGL2(K), that is, may not J20 42 all be of the same type. The polars of internal points are J20 43 non-secant lines of C. When K is GF(q), the finite J20 44 field with q (odd) elements, C has q + 1 points. So J20 45 there are q(q + 1)/2 chords and external points. As PG(2, J20 46 q) has formula points there are formula J20 47 internal points, in this case all of the same type and in a single J20 48 orbit under formula.

J20 49 1.3 The subgroups of PSL2(q) have been known J20 50 since the turn of the century [14; 12; 3, pp. 260 J20 51 et seq]: Wagner [13, p. 397] gives a convenient J20 52 non-overlapping list with information about PG(1, q). This J20 53 translates readily to C. Some of the subgroups are immediately J20 54 transparent: dihedral groups, cyclic groups, or semidirect products J20 55 of elementary abelian groups by cyclic groups, that fix chords, J20 56 non-secant lines or tangents; the groups of the sub-conics J20 57 Cr (isomorphic to PG(1, r)) that C J20 58 contains whenever GF(r) is a subfield of GF(q). Other J20 59 subgroups occur only when q is odd. Then A4 appears J20 60 when GF(q) has characteristic greater than 3, and J20 61 <*_>SIGMA<*/>4 when, in addition, formula. Their J20 62 existence is not immediate from the geometry of PG(1, q); J20 63 their orbit sizes depend on further congruence properties of q J20 64 [13, p. 397]. They are immediately apparent in the richer J20 65 geometry of C: they are the stabilizers of self-polar J20 66 triangles [6, pp. 364, 369, 370]. The only other subgroup J20 67 remaining in Wagner's list is A5. The desire to complete J20 68 the story and have C account for A5 is thus J20 69 pressing. Altogether A5 occurs in three places in the J20 70 list if q is odd:

J20 71 (i) if formula and GF(q) does not have J20 72 characteristic 3 it occurs and has orbits of lengths 12, 20 or 30 J20 73 according to whether formulae respectively;

J20 74 (ii) if GF(9) <*_>unch<*/> GF(q) it occurs acting J20 75 transitively on the 10 points of a sub-conic J20 76 C9;

J20 77 (iii) if GF(5) <*_>unch<*/> GF(q) then A5 occurs J20 78 as the PSL2(5) of the sub-conic C5, and if J20 79 formula then formula is also a subgroup of J20 80 PSL2(q).

J20 81 The hexagons that we shall display account for all these three, J20 82 apparently dissimilar, appearances of A5 as a subgroup of J20 83 PSL2(q). And our hexagons are not restricted to the J20 84 finite case: they exist and account for the occurrence of J20 85 A5 in certain PO3(K) for any field K J20 86 of characteristic not 2 in which 5 is a square. Notice that if 5 is J20 87 not a square in K then PGL2(K) contains no J20 88 element of order 5 so we cannot have A5 as a subgroup of J20 89 PSL2(K): a swift direct matrix proof is given in J20 90 Section 2.7. That A5 is a subgroup of J20 91 PSL2(C) has been known at least since Klein [9] J20 92 found all the finite subgroups of PSL2(C). So we J20 93 complete the story here too; all the other finite subgroups of J20 94 PSL2(C) are analogues of, and have corresponding J20 95 interpretations to, other subgroups of PSL2(q). If J20 96 GF(4) is a subfield of K then A5 occurs in J20 97 PSL2(K) as the group PSL2(4) = J20 98 PGL2(4) fixing the subconic C4. Although, J20 99 because of the degeneracy of the polarity of C, this case is J20 100 not captured by our attack, we show in Section 1.7 that there is a J20 101 parallel, though distinct, hexagonal explanation.

J20 102 1.4 The main clue to the discovery of the hexagons relevant to J20 103 A5 was some recent geometry for C in the case when J20 104 K is GF(9). It was shown in [5] that PO3(9) J20 105 acts on a set of 12 hexagons defined by the requirement: each J20 106 vertex is an internal point and three edges pass through each of J20 107 the 10 points on C [5, p. 440]. The geometry shows that J20 108 under P<*_>OMEGA<*/>3(9) these hexagons form two orbits J20 109 of size 6, and that P<*_>OMEGA<*/>3(9) is A6 J20 110 acting, inequivalently, on each orbit [5, p. 443]. One then J20 111 deduces that the stabilizer of such a hexagon in both J20 112 PO3(9) and P<*_>OMEGA<*/>3(9) is A5: J20 113 it acts on five self-polar triangles. There are two reinforcing J20 114 clues. When K is GF(5) then P<*_>OMEGA<*/>3(5) is J20 115 A5, and C has six points which form a hexagon. J20 116 Through each of the 10 internal points passes three chords; these J20 117 are just the edges of the hexagon. If K is GF(11) then Edge J20 118 [6, p. 380] has presented a hexagon whose vertices are J20 119 external points, and for which there are 10 internal points at J20 120 which three edges concur; its stabilizer in J20 121 P<*_>OMEGA<*/>3(11) is A5.

J20 122 A non-vertex point through which pass three edges of a hexagon J20 123 has been called a Brianchon point [8, pp. 393 J20 124 et seq.]. Since Clebsch encountered [2, p. 336] such J20 125 a hexagon in PG(2, R) when considering the plane J20 126 representation of the diagonal cubic surface, we shall call a J20 127 hexagon with (exactly) 10 Brianchon points a Clebsch J20 128 hexagon. The simplest example is a regular pentagon and its J20 129 centre considered as in PG(2, R); the clues provide three J20 130 other examples. We show first (Theorem 1) that PG(2, K) has J20 131 Clebsch hexagons if and only if K is not of characteristic 2 J20 132 and 5 is a square in K. Suppose that we have such a K. J20 133 Which Clebsch hexagons are significant for C? It is not enough J20 134 to demand that all the vertices are the same type of point, and J20 135 that so are all the Brianchon points: a Clebsch hexagon in PG(2, J20 136 C) has all its vertices and Brianchon points external to any J20 137 conic not through them; if C is any conic not through the J20 138 vertices or Brianchon points of a Clebsch hexagon in PG(2, q) J20 139 then, considered as in PG(2, q2), these 16 points J20 140 are all external to the conic (containing) C. For a Clebsch J20 141 hexagon H in PG(2, K) there are (Theorem 2) five J20 142 triangles each with the property that its three sides contain the J20 143 six vertices, two on each side, of H. We call these the J20 144 triangles of H. Motivated by the clues we say that H is a J20 145 Clebsch hexagon of a conic C if its five triangles are J20 146 self-polar for C. We show (Theorem 2) that there is a unique J20 147 orthogonal polarity P for which the triangles of a Clebsch J20 148 hexagon H are self-polar, and that P corresponds to a J20 149 conic; that is, has self-polar points over K, unless K J20 150 has characteristic 0 and the quadratic form formula is J20 151 anisotropic. It follows (Theorem 4) that, apart from the J20 152 exceptional case, a conic has Clebsch hexagons. The stabilizer of J20 153 H in both PGL3(K) and the group of P is J20 154 A5, except when K has characteristic 5 when it is J20 155 <*_>SIGMA<*/>5 (Theorem 5). Thus, apart from the J20 156 exceptional case, the stabilizers of the Clebsch hexagons of a J20 157 conic provide the subgroups A5 or J20 158 <*_>SIGMA<*/>5 of PGL2(K); and their J20 159 stabilizers in PSL2(K) are A5 or J20 160 <*_>SIGMA<*/>5, the latter if and only if J20 161 GF(52) is a subfield of K. Nothing better can be J20 162 hoped for. In the exceptional case PSL2(K) does not J20 163 contain A5 (Theorem 5).

J20 164 1.5 The A5 of a Clebsch hexagon H of a conic J20 165 C is (Section 2.8) transitive on the vertices, Brianchon J20 166 points and edges of H. It is their nature that accounts for J20 167 the different possibilities for orbits on C described in J20 168 Section 1.3. The six vertices (10 Brianchon points) lie on a conic J20 169 if and only if K has characteristic 5 (3), when this conic is J20 170 C: there is then an orbit of size 6 (10). Otherwise (Theorem J20 171 6), provided they are chords and not non-secants, the 15 edges, 10 J20 172 polars of the Brianchon points, and six polars of the vertices J20 173 intersect C in orbits of 30, 20, 12 points respectively. The J20 174 group PGL2(K) is transitive on the Clebsch hexagons J20 175 of C; if K is GF(q) then they form (Theorem 7) two J20 176 orbits under PSL2(q) unless the characteristic of J20 177 K is 5 and GF(52) is a subfield of K. And, J20 178 except in characteristic 5, each triangle of H is shared with J20 179 one other Clebsch hexagon of C. This sharing gives rise over J20 180 those GF(q) with formula or formula to a J20 181 bipartite 5-valent graph on formula vertices on which J20 182 PGL2(q) acts. If formula or formula J20 183 each orbit under PSL2(q) is a 5-valent graph with J20 184 formula vertices on which PSL2(q) acts J20 185 (Theorem 7). The smallest non-trivial case occurs for q=31 J20 186 when there are 248 vertices.

J20 187 One other feature deserves highlighting for characteristic not J20 188 5. We say, in analogy with the classical case, that two conics have J20 189 double contact if over K or a quadratic extension of K J20 190 they touch at two points; equivalently they define a pencil J20 191 containing a repeated non-tangent line. The six conics through five J20 192 of the vertices of H have double contact with C. Further, J20 193 if M is a vertex of H, then all the other vertices of all J20 194 the Clebsch hexagons of C with M for one vertex lie on J20 195 the same conic, and each of its points, apart from any contact J20 196 point with C, is the vertex of one Clebsch hexagon of C J20 197 (Theorem 8).

J20 198 1.6 What of the geometry of the Clebsch hexagons, when they J20 199 occur, in the plane? They are all of the same type; J20 200 PGL3(K) acts transitively on them (Theorem 1). Any J20 201 two triangles of a Clebsch hexagon H are (Theorem 9) in triple J20 202 perspective from three collinear centres, which are vertices, one J20 203 from each, of the other three triangles. (There is an additional J20 204 centre for characteristic 5.) This is a consequence of there being J20 205 six lines each containing five vertices of the triangles of H J20 206 and 10 lines containing three such vertices. The full figure of J20 207 these lines together with the vertices of H, the Brianchon J20 208 points, the vertices of the triangles and the edges makes a Clebsch J20 209 hexagon a self-dual configuration (Theorem 9).

J20 210 And there is more to be told. Clebsch hexagons also account for J20 211 the existence, for certain K, of A6 as a subgroup of J20 212 PSL3(K). For reasons of length the details must J20 213 await a sequel to this paper, but it is pertinent to whet the J20 214 appetite and to indicate this part of the full story. J20 215 J21 1 <#FLOB:J21\>ON ANTIPLECTIC PAIRS IN THE HAMILTONIAN J21 2 FORMALISM OF EVOLUTION EQUATIONS

J21 3 By GEORGE WILSON

J21 4 Introduction

J21 5 IT is well known that evolutionary systems of the type of the J21 6 Korteweg-de Vries (KdV) equation can (under suitable boundary J21 7 conditions) be regarded as examples of infinite dimensional J21 8 integrable Hamiltonian systems. Recently (see [9]) I have J21 9 argued that these systems typically occur as part of a pair of J21 10 equations related by a transformation of the dependent variables. J21 11 The relationship between the two equations of the pair is analogous J21 12 to the relationship between a Lie group and its Lie algebra. From J21 13 the Hamiltonian viewpoint, these pairs have a curious feature: one J21 14 of the flows in question takes place on a (slightly degenerate) J21 15 symplectic manifold, the other on a Poisson J21 16 manifold. The purpose of this paper is to lay the foundations J21 17 for that part of the theory of such pairs that is independent of J21 18 the boundary conditions, and hence of a purely algebraic nature. J21 19 The appropriate language for this is what Gel'fand and Dikii call J21 20 the formal calculus of variations (see [3]). We begin by J21 21 reviewing the relevant background material.

J21 22 In the literature on finite dimensional Hamiltonian systems, J21 23 one can observe some mild disagreement as to whether the notion of J21 24 a symplectic manifold or a Poisson manifold is the better starting J21 25 point. We recall first the latter notion. Let M be a smooth J21 26 manifold, l a skew tensor of type (2,0) on M. We can J21 27 think of l as a skew form on the cotangent bundle T*M; J21 28 or, equivalently, we can consider the corresponding homomorphism J21 29 formula. Thus l# assigns to each 1-form on J21 30 M a vector field. Let A denote the algebra of all smooth J21 31 real valued functions on M. Then for each f <*_>unch<*/> J21 32 A we have the 1-form df, and hence the vector field J21 33 l#(df) : we shall think of this vector field as J21 34 a derivation of A and denote it by <*_>delta<*/>f. J21 35 Thus we have, for any f, g <*_>unch<*/> A

J21 36 formula

J21 37 The function formula is also denoted by {f, J21 38 g} and called the Poisson bracket of f and J21 39 g. It is evidently skew, since l is. If in addition we J21 40 have, for every f, g <*_>unch<*/> A,

J21 41 formula&caption

J21 42 or, equivalently, if the bracket satisfies the Jacobi identity J21 43 and the map formula is a homomorphism of Lie algebras, then J21 44 we say that (M, l) is a Poisson manifold. The J21 45 condition (1.1) translates into a quadratic condition on the tensor J21 46 l, namely, the vanishing of a certain skew 3-tensor, the J21 47 Schouten-Nijenhuis bracket of l with itself: see, for example, J21 48 [4, 8].

J21 49 If now in addition to all the above l is non-degenerate, J21 50 that is, l# is an isomorphism, we can consider the J21 51 inverse isomorphism formula, and the corresponding J21 52 non-degenerate 2-form <*_>lambda<*/>. The pair (M, J21 53 <*_>lambda<*/>) is then called a symplectic manifold. One J21 54 advantage of formulating the theory in terms of <*_>lambda<*/>, J21 55 rather than the inverse tensor l, is that the crucial J21 56 condition (1.1) becomes a very simple linear condition on the J21 57 2-form <*_>lambda<*/>, namely that it should be closed: J21 58 d<*_>lambda<*/> = 0. Nevertheless, from this point of view, a J21 59 symplectic manifold is just a special case of a Poisson manifold. J21 60 On the other hand, any Poisson manifold (M, l) comes J21 61 equipped with a kind of foliation by symplectic manifolds in which J21 62 the tangent spaces to the leaves (which in general may not all have J21 63 the same dimension) make up the image of the map l#. J21 64 All the Hamiltonian flows on M (that is, the integral curves J21 65 of the vector fields <*_>delta<*/>f) preserve the J21 66 symplectic leaves of this foliation. The effect of all this is that J21 67 it usually makes little difference whether we regard <*_>lambda<*/> J21 68 or l as the more fundamental object.

J21 69 Before passing to the infinite dimensional case, let us observe J21 70 that all the discussion above could be formulated in more algebraic J21 71 language in terms of the algebra A of functions on M; J21 72 this would be particularly appropriate in the variant of the theory J21 73 where M was an affine algebraic variety and A the algebra J21 74 of regular functions on M. Instead of the cotangent bundle to J21 75 M we should then speak of the universal A-module J21 76 with derivation <*_>OMEGA<*/>; the tensor l above would become J21 77 a skew bilinear form formula, and a 2-form would be a skew J21 78 form on the dual A-module <*_>OMEGA<*/>*.

J21 79 Now let us consider the prototype example of the infinite J21 80 dimensional systems that we want to study, the KdV equation

J21 81 formula&caption

J21 82 We want to regard this equation as defining a flow on some J21 83 infinite dimensional manifold M of functions u(x) J21 84 (we prescribe the initial value u(x, 0) and (1.2) then J21 85 determines u(x, t)). For that we have to fix J21 86 suitable boundary conditions; for example, we can take M to be J21 87 the space of all smooth periodic functions u(x) (of some fixed J21 88 period). The interpretation of the KdV equation as a J21 89 Hamiltonian flow on M is based on the fact that (1.2) can J21 90 be written in the Hamiltonian form

J21 91 formula&caption

J21 92 where

J21 93 formula&caption

J21 94 and formula; here we write formula, and J21 95 formulais the Euler-Lagrange operator: in general, if J21 96 f is a function of u, ux, uxx, J21 97 ... then

J21 98 formula&caption

J21 99 The justification for calling (1.3) 'Hamiltonian' is as J21 100 follows. On the vector space M we have the usual J21 101 L2 inner product formula (where the integral J21 102 is taken over a period) which identifies M with a subspace of J21 103 its dual M*; we call this subspace the 'smooth part' of J21 104 M*. If formula is the function

J21 105 formula

J21 106 one easily finds that the exterior derivative dH belongs J21 107 to the smooth part of M*, and that its value at a point u J21 108 <*_>unch<*/> M, regarded as an element of formula, is J21 109 just formula. Thus if we regard the skew differential J21 110 operator l in (1.4) as analogous to the map l# J21 111 in our discussion of finite dimensional Poisson manifolds, then J21 112 (provided we can check the vital condition (1.1)) it is clear that J21 113 (1.3) is indeed the vector field on the Poisson manifold M J21 114 corresponding to the function H.

J21 115 We shall not pursue here the question of making this discussion J21 116 rigorous, but shall pass at once to the algebraic formulation of J21 117 the theory. By that I do not mean simply replacing the infinite J21 118 dimensional manifold M by its algebra of functions: that would J21 119 achieve no simplification, and would ignore two important features J21 120 of our situation. The first is that our manifold M is not just J21 121 an abstract manifold, but consists of functions of x. Thus on J21 122 M we have a distinguished vector field formula, and we J21 123 are interested only in vector fields formula that commute J21 124 with <*_>delta<*/>. Still more significant is the remark that, J21 125 because (1.3) involves only the 'density' f, and not the J21 126 function formula, the boundary conditions used in our J21 127 discussion are for many purposes irrelevant: for example, if we J21 128 were to replace M by a suitable space of functions on the J21 129 whole line, decaying at infinity, a large part of the theory would J21 130 remain unaffected. The basic question of whether (1.1) is satisfied J21 131 belongs to this purely local part of the theory. For these reasons J21 132 we follow the influential paper [3] of Gel'fand and Dikii, and J21 133 proceed as follows. Let formula be the algebra of J21 134 differential polynomials in u: that is, polynomials J21 135 in the infinite set of variables formulae (we write J21 136 formulae, etc.). We equip A with the unique derivation J21 137 <*_>delta<*/> such that formula. We now try to think of the J21 138 differential algebra (A, <*_>delta<*/>) as the algebra of J21 139 functions on the manifold M (although it is not), and imitate J21 140 the algebraic version of the finite dimensional theory. The role of J21 141 the vector fields on M is played by the Lie algebra of J21 142 evolutionary derivations of A, that is, derivations J21 143 commuting with <*_>delta<*/>. Clearly such a derivation of A J21 144 is entirely determined by its value on u, and this value may J21 145 be chosen arbitrarily. Thus equation (1.2) (or (1.3)) can be J21 146 interpreted as defining an evolutionary derivation of A. Of J21 147 course, this way of studying the algebraic properties of an J21 148 evolutionary system is quite classical, and has long been a J21 149 standard point of view in discussing the local conservation laws of J21 150 such systems; however, the fact that much of the Hamiltonian J21 151 formalism too can be described within this framework seems first to J21 152 have been made clear in [3]. Given the skew differential operator J21 153 l, the basic construction of the theory now assigns to each J21 154 f <*_>unch<*/> A the unique evolutionary derivation J21 155 <*_>delta<*/>f of A such that formula is the J21 156 right-hand side of (1.3). We then call l Hamiltonian if a J21 157 suitably formulated version of (1.1) holds. In this algebraic J21 158 context the idea of considering the inverse of the differential J21 159 operator l seems very unnatural; and it is doubtful if we J21 160 could make any sense of discussing the symplectic leaves of a J21 161 foliation defined by l. Therefore, the following attitude J21 162 towards the Hamiltonian formalism has become standard: we should J21 163 simply accept that l defines a Poisson structure in the way J21 164 sketched above, and abandon any idea of relating it to a symplectic J21 165 structure, as we do in the classical case.

J21 166 In [9], however, I have argued that this may not be the best J21 167 attitude. For the KdV equation and its generalizations, the J21 168 operator l, although not invertible, is in a certain sense J21 169 obtained by inverting a 2-form <*_>lambda<*/>; moreover, it is J21 170 <*_>lambda<*/> that is the more fundamental object. The J21 171 justification for this last claim is that <*_>lambda<*/> is J21 172 invariant with respect to a symmetry group G (for the KdV J21 173 equation G is PSL(2)) and l is obtained from J21 174 <*_>lambda<*/> essentially by passing to the quotient by G. J21 175 The construction of <*_>lambda<*/> from l is less J21 176 straightforward, especially since in practice the group G is J21 177 not given to us in advance. As an example, the 2-form J21 178 <*_>lambda<*/> corresponding to the l in (1.4) is

J21 179 formula&caption

J21 180 where the variables <*_>psi<*/> and u are related by

J21 181 formula&caption

J21 182 (this last expression is the Schwarzian derivative of J21 183 <*_>psi<*/>: it is invariant with respect to the group G = J21 184 PSL(2) acting on <*_>psi<*/> by linear fractional J21 185 transformations). We shall call such a pair of variables J21 186 (<*_>psi<*/>, u) with their associated skew operators J21 187 (<*_>lambda<*/>, l) an antiplectic pair. As we shall J21 188 see, a consequence of the existence of <*_>lambda<*/> is that each J21 189 equation of the form (1.3) for u is implied by a J21 190 symplectic G-invariant equation for <*_>psi<*/>. For J21 191 example, the KdV equation (1.2) follows from the 'Ur-KdV' J21 192 equation

J21 193 formula&caption

J21 194 for <*_>psi<*/>. This equation is symplectic in the sense that J21 195 it can formally be written as

J21 196 formula&caption

J21 197 where H = u2 is the Hamiltonian for the KdV J21 198 equation. It is a basic part of the machinery that we are going to J21 199 discuss that the appearance of <*_>lambda<*/>-1 here does J21 200 not actually require us to integrate anything: in fact for any J21 201 Hamiltonian H depending only on u, ux, ... J21 202 we have

J21 203 formula

J21 204 so that the <*_>lambda<*/> here cancels the J21 205 <*_>lambda<*/>-1 in (1.9) to give (1.8). This example is J21 206 considered in more detail in [9], and in <*_>section<*/>7 below.

J21 207 The main purpose of the present paper, then, is to give a J21 208 precise formulation of the notion of an antiplectic pair, and in J21 209 particular to clarify the process by which the type of the J21 210 G-invariant tensor <*_>lambda<*/> reverses when we divide J21 211 out by G. The paper is arranged as follows. Section 2 is J21 212 preliminary, and is intended for reference only. It sets out some J21 213 elementary facts (and conventions) about modules over rings with an J21 214 anti-involution. The motivation for that is the following: the J21 215 basic derivation <*_>delta<*/> of the algebra A acts also on J21 216 the universal A-module <*_>OMEGA<*/> that plays the role, J21 217 in the algebraic theory, of the contangent bundle T*M. Thus J21 218 <*_>OMEGA<*/> acquires a structure of left module over the algebra J21 219 R of (ordinary) differential operators with coefficients in J21 220 A: this structure plays an important role in our version of J21 221 the formalism. On R we have the anti-involution J21 222 formula, where L* is the formal adjoint of L. J21 223 This makes possible certain constructions that could not be J21 224 performed for a general non-commutative ring R. J21 225 J22 1 <#FLOB:J22\>Illness behaviour therefore includes variables such as J22 2 self-assessment of ill-health, general-practice consultations, J22 3 consumption of over-the-counter medication, consumption of J22 4 prescribed medication and sickness absence (both certified and J22 5 uncertified).

J22 6 Early studies of stress outcomes have tended to infer the J22 7 presence of illness by measuring illness behaviours such as J22 8 sickness absence. However, the relationship between illness and J22 9 illness behaviour is not a direct one-to-one relationship, and it J22 10 is now well known that many complex, social, cultural and J22 11 demographic factors contribute to the causation of sickness absence J22 12 besides illness per se (Johns and Nicholson, 1982). We J22 13 therefore find that the demographic aspects of illness are not J22 14 necessarily the same as the demographic aspects of illness J22 15 behaviour.

J22 16 Theoretical and Methodological Issues

J22 17 In stress research the concepts of stress, coping and J22 18 well-being are frequently confounded (Edwards and Cooper, 1988). J22 19 The authors give the example in life-events research, where some J22 20 'stressful' life events may also be construed as an inability to J22 21 cope (e.g. divorce). Personal illness may be regarded as a J22 22 stressful life event or as a health outcome. Furthermore, some J22 23 definitions of stress, such as those which describe stress as a J22 24 situation where demands exceed abilities (e.g. French, Rodgers and J22 25 Cobb, 1974), confound stress with the inability to cope. It is J22 26 therefore essential, in stress research, for authors to clearly J22 27 present the theoretical model which they are investigating.

J22 28 SEX DIFFERENCES

J22 29 It has been argued that women, by virtue of the roles they J22 30 occupy, experience more life events and chronic social stresses, J22 31 and less social support than men, and that this differential J22 32 exposure to risk factors explains women's greater vulnerability to J22 33 depression.

J22 34 Sex Differences in Social Stress

J22 35 The empirical data available so far suggest that there is no J22 36 difference in the rates at which men and women experience acute J22 37 life events or adversity (Meyers et al., 1971; Dekker and Webb, J22 38 1974; Newman, 1975, Henderson et al., 1980; Markush and Favero, J22 39 1974). However, the possibility remains to be tested that women in J22 40 general may experience more undesirable life events by virtue of J22 41 their low socio-economic status overall (Myers, Lindenthal and J22 42 Pepper, 1975), since there is much evidence that women in general J22 43 enjoy lower status than men, both at home and at work, and J22 44 frequently earn less even when in comparable jobs (Office of J22 45 Population Censuses and Surveys, Social Trends). There is J22 46 no evidence that life events have more impact on women than on men J22 47 (Paykel, Prusoff and Uhlenhuth, 1971; Personn, 1980). However, J22 48 there is evidence that women experience more chronic social stress J22 49 than men. Radloff and Rae (1979) reported that women were more J22 50 exposed than men to low education, low income, low occupational J22 51 status, fewer leisure activities, and more current and recent J22 52 physical illness.

J22 53 Furthermore, there is evidence that men and women respond J22 54 differently to the same number of stresses (e.g. Russo, 1985). In J22 55 addition, there is ample evidence for sex differences in J22 56 stress-induced physiological responses (e.g. Frankenhauser, 1983; J22 57 Stoney, Davis and Matthew, 1985) and also for differences in the J22 58 ways people cope cognitively, emotionally and behaviourally in J22 59 response to stress.

J22 60 According to Kessler, Price and Wartman (1985) gender J22 61 differences in health are to a large extent attributable to J22 62 differences in the appraisal of stresses and the selection of J22 63 coping strategies.

J22 64 Vingerhoets and Van Heck (1990) explored gender differences in J22 65 coping and found that males preferred problem-focused coping J22 66 strategies, planned and rational actions, positive thinking, J22 67 personal growth and humour, day-dreaming and fantasies. J22 68 Women preferred emotion-focused coping solutions, self-blame, J22 69 expression of emotions, seeking of social support and wishful J22 70 thinking/emotionality. The same authors also found that men and J22 71 women do not differ in terms of the amount of stressful events J22 72 experienced.

J22 73 Sex Differences in Social Support

J22 74 There are few studies which specifically address the question J22 75 of whether women experience less social support J22 76 <}_><-|>then<+|>than<}/> men. Miller and Ingham (1976) found that J22 77 casual, less intimate friends as well as intimates afforded J22 78 protection from developing illness, and that "psychological J22 79 symptom levels probably vary with social support even when there is J22 80 no serious life event present". It is therefore apparent J22 81 that contacts with colleagues at work may also be supportive to the J22 82 individual, and it may be that the housewife often experiences J22 83 relative isolation in the home, experiencing less frequent daily J22 84 verbal exchanges with other individuals than does her counterpart J22 85 in the office.

J22 86 Henderson et al. (1979) found that males reported more J22 87 availability of social integration than females, while females J22 88 scored higher on the quality or adequacy of the social integration. J22 89 Females scored more on availability of attachment than males, but J22 90 there was no sex difference on the quality or adequacy of the J22 91 attachment. It was the author's view that special attention should J22 92 be paid to those social bonds which promote self-esteem - both the J22 93 esteem of self in terms of appearance, abilities, competence and J22 94 position in a dominance hierarchy, as well as the degree to which J22 95 one believes one is lovable to others. The question is, therefore, J22 96 whether such self-esteem is more likely to be derived from social J22 97 integration within a group, while the extent to which one believes J22 98 one is lovable may be obtained from both kinds of social bond. If J22 99 the important aspects of self-esteem are more likely to be derived J22 100 from social integration, then Henderson's finding that males J22 101 reported quantitatively more availability of social integration J22 102 than females may be of crucial significance to the question of J22 103 whether women experience less social support than men. While J22 104 females report a better quality of social integration, in terms of J22 105 self-esteem thus engendered, quality may not make up for quantity. J22 106 Henderson found that for minor psychiatric morbidity social J22 107 integration had a stronger association with symptom level than did J22 108 attachment for women. For men, the strength of the association of J22 109 symptom level with social integration and with attachment was the J22 110 same. Henderson concluded that "social bonds" appear to be J22 111 related to morbidity in a manner independent of the challenge of J22 112 adversity". While these primary questions afford some hope J22 113 of elucidating the nature of the sex difference in the prevalence J22 114 of minor psychiatric morbidity, it is clear that further work is J22 115 required. In the mean time the evidence suggests that women do J22 116 experience more chronic social stresses, e.g. low occupational J22 117 status and low income, than men and also experience less J22 118 availability of social integration - a factor with a strong J22 119 negative association with minor psychiatric morbidity.

J22 120 Brugha et al. (1990) found that the explained variance in J22 121 recovery from depression due to social support was equal in men and J22 122 women. But according to subset analyses, the aspects of personal J22 123 relationships and perceived support that predict recovery in men J22 124 and women appear to be different. In women, the significant J22 125 predictors of recovery appeared to be the number of primary group J22 126 members named and contacted, and satisfaction with social support, J22 127 while in men it appeared to be living as married, and the number of J22 128 non primary group social contacts named as acquaintances or J22 129 friends.

J22 130 Sex Differences in Illness

J22 131 Men and women certainly differ in terms of the balance of role J22 132 obligations in the occupational, marital and parental domain. If J22 133 these distinctive role patterns are responsible for gender J22 134 differences in health, it should be the case that where gender J22 135 equality is achieved, gender differences in health should be J22 136 reduced. Studies of men and women in similarly responsible and J22 137 demanding jobs do seem to find a reduction in the substantially J22 138 lower mortality rates among women (e.g. Detre et al., 1987). J22 139 Jenkins (1985), in a study of young male and female executive J22 140 officers, found that there was no sex difference in prevalence of J22 141 minor psychiatric morbidity. Other studies of true homogeneous J22 142 samples have found the same thing (e.g. Parker, 1979; Golin and J22 143 Hartz, 1977; Hammen and Padesky, 1977).

J22 144 A paradox which has attracted considerable attention ever since J22 145 John Graunt, the founder demography, commented on it in his J22 146 'Natural and political observations', published in 1662, is that J22 147 while women attend doctors more often than men, their life J22 148 expectancy is no less than that of men (indeed, is now about 8% J22 149 longer). Graunt concluded that either the women were generally J22 150 cured by their physicians or that the men suffered form untreated J22 151 morbidity. In Western countries, where women's life expectancy is J22 152 greater than men's, women are nonetheless reported to suffer more J22 153 illness than men, are higher users of medical services and J22 154 prescriptions, and take more time off work for sickness (Verbrugge, J22 155 1976; Nathanson, 1977; Wingard, 1984; Verbrugge, 1985; Strickland, J22 156 1988; Jenkins, 1985).

J22 157 Table 5.1 illustrates this paradox with figures taken from UK J22 158 sources. It can be seen that, while females' life expectancy in J22 159 England and Wales exceeds that of men by six years, women consult J22 160 general practitioners (GPs) more often than men, they take J22 161 prescribed drugs more often than men, they take more frequent J22 162 spells of sickness absence (although the total duration of the J22 163 certified absence is not greater) and, despite attending J22 164 out-patient facilities in roughly equal numbers, women are admitted J22 165 to hospital more often than men. In general it can be said that J22 166 women suffer more from psychological distress and minor somatic J22 167 disorders, whereas men seem to be especially vulnerable to J22 168 life-threatening diseases, e.g. myocardial infarction and cancer J22 169 (e.g. Rice at al., 1984; Bush and Barrett-Conner, 1985).

J22 170 table&caption

J22 171 Looking now at mental illness, GPs diagnose more episodes of J22 172 mental illness in women than in men, women take more certified J22 173 sickness absence for mental illness than do men (both in terms of J22 174 frequency and duration) and women have more psychiatric admissions J22 175 to hospital than do men. These comparisons are illustrated in Table J22 176 5.2. When specific diagnostic categories of mental illness are J22 177 examined, using the International Classification of Diseases the J22 178 picture becomes rather more subtle. Table 5.3 presents the general J22 179 practice episode rates for psychiatric illness and the admission J22 180 rates to psychiatric hospitals for the year 1972 by diagnosis and J22 181 sex. Males predominate in the areas of alcoholism and personality J22 182 disorders. Sex differences in reported rates of schizophrenia are J22 183 negligible. However, for affective psychoses and the J22 184 psychoneuroses, women predominate over men.

J22 185 tables&captions

J22 186 These sex differences in illness have been variously ascribed J22 187 to a number of explanations which may be categorised as in Table J22 188 5.4.

J22 189 These theories have been considered in relation to mental J22 190 health in Jenkins (1985), who concluded that:

J22 191 Clearly the relative importance of each of these variables is J22 192 likely to vary from illness to illness. For mental health, we now J22 193 have evidence that both manic depressive psychosis and J22 194 schizophrenia have a multifactorial aetiology, involving both J22 195 genetic and environmental factors. Both diseases are so severe that J22 196 sex differences in reporting behaviour and diagnostic habits have J22 197 minimal impact on reported rates. However for minor psychiatric J22 198 morbidity, the depressions and anxiety states, the evidence for a J22 199 genetic aetiology is small. While there is evidence that changes in J22 200 gonadal hormones are sometimes linked to mood changes, there is no J22 201 direct evidence that reproductive physiology is responsible for J22 202 women's excess of reported depression. The evidence for the J22 203 importance of environmental stress and support in the aetiology of J22 204 minor psychiatric morbidity is much stronger, although the variance J22 205 explained by such factors is not large.

J22 206 This chapter will therefore concentrate on environmental J22 207 factors, rather than the biological factors of genes and hormones J22 208 responsible for sex differences in illness. Sex differences in J22 209 stress and support have already been discussed above. Differences J22 210 in sex roles and their effects on health will be discussed J22 211 below.

J22 212 Sex Roles and Their Influence on Constitutional and J22 213 Environmental Vulnerability to Illness, and on Reporting J22 214 Behaviour

J22 215 Sex roles and their interaction with constitution

J22 216 It has been suggested that sex differences in the early J22 217 upbringing and social environment of males and females place a J22 218 permanent stamp on the phenotype of the individual, thus affecting J22 219 constitutional vulnerability to psychiatric illness in adult life J22 220 (Chesler, 1971, 1972; Chodorow, 1974). The learned helplessness J22 221 model proposes that helplessness is the salient characteristic of J22 222 depression and that it results from learning that one's actions do J22 223 not produce predictable responses (Seligman, 1975). Cochrane and J22 224 Stopes (1980) <}_><-|>aregued<+|>argued<}/> that women are J22 225 traditionally more sheltered than boys, women have less initiative J22 226 in selecting their spouses than do men, their life-styles face more J22 227 disruption with the advent of children, and they have to follow J22 228 their husbands geographically and socially. J22 229 J23 1 <#FLOB:J23\>Processes J23 2 All theories of behaviour make some assumptions about J23 3 innateness and, thus, about the architecture of cognition, even J23 4 though the issue has not always been explicitly addressed when the J23 5 theory sets out its assumptions. In particular, all theories assume J23 6 that the mechanisms of learning or development are themselves J23 7 invariant, and are innate endowments of the organism. Behaviourist J23 8 theories attribute all of behaviour to learned responses. Such a J23 9 theory presumes an organism that is disposed to learn; an organism J23 10 that is innately endowed with the ability to associate a stimulus J23 11 with a response. While the association between a particular S and a J23 12 particular R is learned, behaviourists do not argue that organisms J23 13 learn how to make associations; associations are part of the J23 14 organism's innate repertoire. Similarly the ability to discriminate J23 15 among stimuli and to generalize across stimuli are innate and even J23 16 though the basis of particular discriminations and generalizations J23 17 may have to be learned.

J23 18 For Piaget's constructivist theory the mechanisms of J23 19 assimilation and accommodation are innate and unchanging J23 20 mechanisms. Piaget is quite explicit on this issue. He sees these J23 21 mechanisms as evolution's answer to adaptation to the environment J23 22 (Piaget, 1967a/1971). These mechanism support changing cognitive J23 23 structures.

J23 24 The information processing framework also postulates a variety J23 25 of innate mechanisms for learning and development. Many J23 26 developmental theorists within the information processing framework J23 27 have borrowed Piaget's constructs of assimilation and J23 28 accommodation. However, the major mechanism of learning presumed by J23 29 information processing theories in induction (Holland, Holyoak, J23 30 Nisbett and Thagard, 1986). Induction is the formulation of general J23 31 laws from particular cases as for example:

J23 32 I have seen lots of white swans.

J23 33 I have never seen any other colour of swan.

J23 34 Therefore, all swans are white.

J23 35 This reasoning is an impeccable piece of induction, although it J23 36 is not logically valid. Philosophers have been much preoccupied by J23 37 the fact that it is not logically valid. However, from a J23 38 psychological viewpoint it is the mechanism itself that is of J23 39 interest, not its logical status. Clearly, much human knowledge is J23 40 based on much less than certainty. Bertrand Russell (1912: p. 38) J23 41 concluded that the inductive process itself is "incapable J23 42 of being proved by an appeal to experience" but went J23 43 on to argue that "we must either accept the inductive J23 44 principle on the ground of its intrinsic evidence, or forgo all J23 45 justification of our expectations about the future". Any J23 46 theory that postulates rule learning needs a mechanism for deriving J23 47 rules from environmental regularities and induction serves this J23 48 purpose admirably. The importance of induction is that it permits J23 49 extrapolation from particular instances and provides a rule-based J23 50 framework for cognitive processing. An example of what is commonly J23 51 regarded as inductive learning is the child's derivation of the J23 52 linguistic rule 'add -ed to the verb stem to form the past J23 53 tense'. Initially children learn past tense forms on a word-by-word J23 54 basis and thereby acquire the correct past tense form for both J23 55 regular and irregular verbs (e.g., walked and took). J23 56 However, as more verbs are learnt the child notices that the J23 57 majority of verbs form the past tense by adding -ed to the J23 58 verb root. This results in an inductive generalization that all J23 59 verbs form the past tense by adding -ed to the root. J23 60 Application of this rule results in an overregularization of J23 61 previously correct forms, so that taked will now replace J23 62 took. The phenomenon will be discussed in more detail in J23 63 chapter 7. Here, the major point is to illustrate that an inductive J23 64 inference can have powerful effects on the child's cognitive J23 65 representations.

J23 66 There seems to be general agreement that the basic cognitive J23 67 mechanisms of change are innate evolutionary products and are not J23 68 themselves subject to developmental change. The content to which J23 69 they apply does, of course, change with changing cognitive J23 70 representations. Associations formed late in development may have a J23 71 considerably more complex representational structure than J23 72 associations formed in the first few days of life. Similarly, J23 73 induction may produce relatively simple rules in early development J23 74 but more complex rules in later development. The difference in J23 75 complexity in both cases lies in the representation not in the J23 76 mechanism that produced it.

J23 77 Structure J23 78 For classical behaviourists the structure of cognition was an J23 79 associative chain of 'mediating responses'. As the content of J23 80 associations was entirely the product of learning, the structure of J23 81 cognition was thus entirely the product of learning.

J23 82 For Piaget the structure of cognition was also a developmental J23 83 product. However, Piaget thought that cognition was more than a J23 84 copy of environmental associations, and he thought that it J23 85 developed but was not necessarily learned (Piaget, 1964). The J23 86 Piagetian view is that through the processes of assimilation and J23 87 accommodation new cognitive structures emerge. Piaget argued that J23 88 reflexes were the initial innate structures from which all other J23 89 structures eventually develop.

J23 90 Information processing theories have a much more complex theory J23 91 of the structure of cognition than either behaviourist or Piagetian J23 92 theories. According to information processing theories, the J23 93 cognitive system consists of a variety of interacting components. J23 94 At a very general level, the major components are memory and J23 95 attention. These components have various processes associated with J23 96 them that transform and encode the input received. The system also J23 97 has components that control the processing of information in J23 98 accordance with the organism's goals and plans.

J23 99 Behaviourist, Piagetian, and many Information Processing J23 100 theories regard the content of cognition as a developmental J23 101 product. This seems an entirely reasonable position. Evolution J23 102 could not possibly have prepared us for the particular information J23 103 content that we receive. Content is a function of the particular J23 104 environment that we inhabit - be it desert, or forest, or city; of J23 105 our particular culture; and of our particular experience within an J23 106 environment and a culture.

J23 107 At one level this analysis is true but it fails to do justice J23 108 to the fact that regardless of the variation in content there are J23 109 certain requirements placed by evolution on how content is J23 110 processed. For example, different inputs must be treated as J23 111 functionally similar by an organism in order for behaviour to be J23 112 adaptive. This requires an innate ability to treat input in a J23 113 categorical fashion. If an organism could not treat some stimuli as J23 114 functionally equivalent to others, there would be no possibility of J23 115 learning from experience because every encounter with the world J23 116 would be unique. As we shall see in chapter 2, the empirical J23 117 evidence is completely in accord with this line of reasoning.

J23 118 Categorization is only one among a variety of structures J23 119 imposed on information. During the course of development the child J23 120 acquires a variety of increasingly complex organizational J23 121 structures. The most notable example of this is human language. It J23 122 has been a matter of considerable debate in recent years to what J23 123 extent the structure of language (and of other cognitive J23 124 representations) is learnt and to what extent it is innate. The J23 125 issue was first raised by Chomsky (1959, 1965) and it has not yet J23 126 been completely resolved. We shall return to this issue in chapter J23 127 7.

J23 128 What and How: A Second Pass

J23 129 The question of how development occurs can be approached in J23 130 several ways. Firstly, it is necessary to identify the learning J23 131 mechanisms possessed by the cognitive system. Secondly, it is J23 132 necessary to describe how these mechanisms act over time. Thirdly, J23 133 there is the issue of how the environment in which a child learns, J23 134 affects development. The first two issues are central to a theory J23 135 of cognitive development; the third is most relevant to issues of J23 136 individual differences.

J23 137 The nature of learning mechanisms postulated by a theory is J23 138 very largely dependent on its view of what develops. It is easiest J23 139 to illustrate this by example. In some behaviourist theories of J23 140 development (e.g., Bijou and Baer, 1961) what develops is a series J23 141 of discriminative responses to environmental stimuli. Such a theory J23 142 places no more demands on learning than that a child be capable of J23 143 associating a stimulus and a response. Thus, the answer to how J23 144 development occurs is by forming new associations or refining old J23 145 ones. By contrast Piaget's theory of development regards the child J23 146 as engaged in a continuous attempt to construct a theory of the J23 147 world, while, at the same time, interpreting the world in the light J23 148 of the child's current theory. Such a theory requires a give and J23 149 take between the child's existing theory and new information, J23 150 otherwise the system would never progress. Piaget accordingly J23 151 postulates mechanisms of assimilation and accommodation as the J23 152 primary learning mechanisms. Assimilation is the process by which J23 153 new information is interpreted in the light of existing cognitive J23 154 structures; accommodation is the process by which cognitive J23 155 structures change in the light of new information.

J23 156 Information processing theories place great emphasis on changes J23 157 in the child's representation of the world. Accordingly, mechanisms J23 158 of change are needed that shift the child from one representation J23 159 of a task to another, more sophisticated representation of that J23 160 task. The mechanisms of change proposed will vary as a function of J23 161 the relation between the representations. Flavell (1972) discusses J23 162 some of the varieties of representational change from an J23 163 information processing perspective. However, it is noteworthy that J23 164 the issue of how change occurs has rarely been given serious J23 165 theoretical discussion within the information processing J23 166 framework.

J23 167 As one further example consider Chomsky's (1965) view of J23 168 language development. Chomsky argued that what develops during J23 169 language development is a system of rules for a particular J23 170 language. These rules are derived from a system of abstract J23 171 grammatical rules shared by all languages. He argued further that J23 172 this system could not be learnt from the environment by any J23 173 conventional learning procedure. Nevertheless it must J23 174 <}_><-|>he<+|>be<}/> possible for children to acquire these rules J23 175 because children evidently do learn a language. Consequently, J23 176 Chomsky proposed that, in some sense, general rules of language J23 177 were innate although obviously the grammar and vocabulary of a J23 178 particular language must be a product of interaction between J23 179 environmental input and innate general rules. This interaction is J23 180 conducted through the mechanism of testing hypotheses (derived from J23 181 the innate general rules) about the structure of the particular J23 182 language the child hears.

J23 183 It is evident from the three theoretical sketches that there is J23 184 a strong inter-dependency between the what and the how of J23 185 development. Chomsky is the most notable example here because he J23 186 was led to question whether conventional accounts of how J23 187 development proceeds could adequately explain the acquisition of J23 188 the rules of grammar. Finding a negative answer Chomsky then J23 189 proposed a greater role for biological inheritance than was common J23 190 in theories of development. We shall return to the issue of J23 191 innateness and biological inheritance in due course. For now, it J23 192 can be remarked that the basic mechanisms of development postulated J23 193 by most, if not all, theories are regarded as the product of J23 194 evolution rather than of development. That is to say, the child is J23 195 assumed to have an innate ability to make associations, or to J23 196 engage in assimilation and accommodation. These mechanisms do not J23 197 themselves require a developmental explanation within a theory of J23 198 cognitive development.

J23 199 How do mechanisms act over time to produce change? In a J23 200 behaviourist theory associations are strengthened or weakened over J23 201 time as a function of the child's history of reinforcement. In a J23 202 Piagetian account of cognitive development the mechanisms of J23 203 assimilation and accommodation serve as a filter between cognitive J23 204 structures and new information. They act continually over time to J23 205 modify existing cognitive structures. For Chomsky the major J23 206 mechanism at the child's disposal is hypothesis-testing about the J23 207 particular structure that language might have. The hypotheses to be J23 208 tested are generated from the child's innate knowledge of the J23 209 general rules of language. By continuous hypothesis testing the J23 210 child gradually arrives at the correct particular structure of the J23 211 language that he or she hears.

J23 212 Theories of cognition emphasize the creation and manipulation J23 213 of representations of the environment. Cognitive development is, to J23 214 a large extent, the development of the ability to create J23 215 increasingly complex and sophisticated representations of the J23 216 environment. The next two chapters will examine the origins of J23 217 mental representations.

J23 218 2

J23 219 Perpetual Development in Infancy

J23 220 Method and Interpretation

J23 221 Writing in 1890, William James declared: "The baby, J23 222 assailed by eyes, ears, nose, skin and entrails at once, feels it J23 223 all one great blooming buzzing confusion" (p. 488). J23 224 J24 1 <#FLOB:J24\>More recent studies have examined the effects of J24 2 moderate intensity noise on cognitive vigilance tasks. An example J24 3 of this type of task is the Bakan task where single digits are J24 4 presented on the screen and subjects have to detect a particular J24 5 sequence such as three successive odd digits. Poulton (1977) J24 6 suggested that all the detrimental effects of noise could be J24 7 explained either by the masking of acoustic cues or, in the case of J24 8 cognitive tasks, by the masking of inner speech. The first part of J24 9 this view has been strongly opposed (see Broadbent, 1978). The J24 10 second part led to an interest in the effects of noise on cognitive J24 11 vigilance, for post-experimental reports from subjects having J24 12 completed the Bakan task clearly suggest that they say the digits J24 13 to themselves. If noise masks internal speech such tasks should be J24 14 impaired and this issue was examined by Jones, Smith & Broadbent J24 15 (1979). They carried out four experiments using slightly different J24 16 versions of the Bakan task. In each case there was evidence of a J24 17 detrimental effect of noise, although the exact nature of the noise J24 18 effect depended on specific features of the task. This, and the J24 19 finding that the noise had no differential effect on digits which J24 20 were rehearsed together and those which were rehearsed separately, J24 21 argues against noise producing effects solely by interfering with J24 22 subvocal articulation.

J24 23 These results do, however, show that vigilance tasks involving J24 24 memory are susceptible to noise below 95 dB, which is consistent J24 25 with results obtained by Benignus, Otto & Knelson (1975). This J24 26 finding has also been confirmed in subsequent studies by Miles, J24 27 Auburn & Jones (1984) and Smith (1988a). The importance of the J24 28 nature of the tasks will be discussed again in a later section, J24 29 after results from other attention paradigms have been J24 30 summarized.

J24 31 Noise and the five-choice serial reaction time task

J24 32 Another task which has been widely used to study the effects of J24 33 noise is the five-choice serial reaction time task. J24 34 Broadbent (1979) has reviewed the effects of very loud noise on J24 35 this task and concluded that noise increases momentary J24 36 inefficiency, in that there are more errors and gaps (occasional J24 37 very long reaction times) than in quiet. Again, this increased J24 38 momentary inefficiency depended on the noise being over 95 dB and J24 39 the subject being in the noise for at least half an hour.

J24 40 Recent research has shown that a long exposure of moderate J24 41 intensity noise can produce comparable results to the older studies J24 42 using levels over 95 dB. For example, Smith & Miles (1985) asked J24 43 subjects to carry out a serial choice reaction time task when they J24 44 had been in 75 dB noise for two hours and when they had been in the J24 45 noise for five hours. The results showed that when subjects had J24 46 been in the noise for five hours there was an increase in the J24 47 number of errors.

J24 48 Poulton (1977) has tried to explain the effects of noise on the J24 49 five-choice serial reaction time task in terms of masking. The J24 50 explanation of the early results has been opposed by Broadbent J24 51 (1978), and recent data also argue against it. For example, Jones J24 52 (1983) was able to confirm the effect of noise on errors and show J24 53 that this was still apparent after the noise had been switched off. J24 54 It is clear that something other than the masking of acoustic cues J24 55 was responsible for this effect of noise on the accuracy of J24 56 performing the task. Smith (1985) and Smith (1986b) changed J24 57 the nature of the serial reaction time task and examined the J24 58 effects of noise on a biased probability task and a version where J24 59 the stimulus-response arrangement was incompatible. The results J24 60 showed that the effects of noise depended on the task parameters, J24 61 with noise slowing responses to stimuli which were presented less J24 62 frequently, but reducing reaction times to those with a high J24 63 probability of occurrence. Incompatible responses were found to be J24 64 slower in noise and it is difficult to explain this result, or the J24 65 previous one, in terms of masking of acoustic cues.

J24 66 Noise and selectivity in attention

J24 67 In multiple tasks or tasks with several subcomponents, noise J24 68 often leads to increased concentration upon the dominant or high J24 69 probability component at the expense of other features. For J24 70 example, Hockey (1970) showed that noise improved performance on a J24 71 central tracking task but led to slower reactions to those lights J24 72 which had a lower probability of occurrence.

J24 73 Broadbent (1971) cites several lines of evidence to support the J24 74 view that noise influences selectivity in memory and attention. For J24 75 example, studies of the effects of noise on the Stroop task have J24 76 shown that noise may reduce the amount of interference from J24 77 irrelevant colour names (e.g. Houston & Jones, 1967). Hockey & J24 78 Hamilton (1970) have also demonstrated that noise aids intentional J24 79 recall but impairs incidental recall, and Smith (1982) has J24 80 replicated this effect using priority instructions rather than the J24 81 intentional/incidental manipulation.

J24 82 Unfortunately, other studies have failed to replicate the J24 83 effects of noise on selectivity in memory and attention. For J24 84 example, Forster & Grierson (1978) and Loeb & Jones (1978) were J24 85 unable to replicate the results of Hockey (1970), although it J24 86 should be noted that they used tasks with a different level of J24 87 difficulty from the original study. Similarly, Smith (1982) J24 88 demonstrated that the effects of task priority are easily modified, J24 89 which in turn changes the effects of noise. Indeed, Smith (1982) J24 90 suggested that Broadbent's (1971) view needs modifying and that J24 91 noise biases the allocation of effort towards the operation which J24 92 appears best to repay the investment of effort. This may take the J24 93 form of a bias towards the high priority task but it is also going J24 94 to depend on other factors such as the difficulty of each part of J24 95 the task, the subject's prior experience and the salience of the J24 96 stimuli. This suggests that most theories which argue that J24 97 performance is shifted by noise in an invariant or mechanical J24 98 fashion are going to be inadequate, and that it may be more J24 99 profitable to consider the strategies of performance in noise.

J24 100 Strategies of performance in noise

J24 101 Smith (1983) has reviewed evidence which demonstrates that, J24 102 when subjects carry out a task which can be performed in different J24 103 ways, noise may lead to the adoption of certain strategies in J24 104 preference of others. If task parameters are changed then the J24 105 method of doing the task often changes, and this plausibly explains J24 106 why noise effects vary when different versions of a task are used. J24 107 It is also known that choice of an appropriate strategy can J24 108 eliminate the detrimental effects of noise. Pollock & Bartlett J24 109 (1932) carried out an experiment in which subjects were given J24 110 groups of letters and told to form as many words as possible from J24 111 these letters. At first the subjects were impaired by noise but J24 112 later they were able to perform as well as in the quiet condition. J24 113 The subjects reported that as time went on they discovered rules J24 114 and mechanical techniques of solving these problems, and once they J24 115 had developed these strategies the noise ceased to have an J24 116 effect.

J24 117 In many tasks it is obvious that one strategy should be used in J24 118 preference to others. This may be because of instructions, previous J24 119 experience, or some other feature of the task. Studies by Smith J24 120 (1982) and Wilding, Mohindra & Breen-Lewis (1982) have shown that J24 121 noise often reinforces the use of the dominant strategy.

J24 122 Smith (1988a, 1990a, b) has examined resource J24 123 allocation in noise by studying dual task performance. The J24 124 difficulty of each task was varied, the probability of having to do J24 125 the tasks was manipulated, and in other conditions one of the tasks J24 126 had a higher priority than the other. The effect of noise depended J24 127 on the nature of the task, in that monitoring tasks involving J24 128 active strategies were impaired by noise whereas tasks performed J24 129 passively were unimpaired. Changing task parameters only influenced J24 130 the noise effect if the two tasks competed for common resources. It J24 131 was also important whether task parameters were constant or changed J24 132 rapidly. This last point is described in more detail in the J24 133 following section on noise and flexibility of changing strategy.

J24 134 There is evidence that noise may influence the efficiency of J24 135 the control processes which monitor and alter performance. For J24 136 example, Rabbitt (1979) has suggested that the effects of noise on J24 137 the five-choice serial reaction time task (an increase in errors J24 138 and/or gaps) can be explained in terms of noise producing J24 139 inefficient control of the speed-error trade-off function. The J24 140 effects of noise on control processes may be of short duration and J24 141 reflect initial coping with the task. However, other effects may J24 142 continue even when the noise is switched off (see Cohen, 1980, for J24 143 a review of the after-effects of noise) and may still be apparent J24 144 when the subject is transferred to the quiet condition (see J24 145 Poulton, 1982, for an account of asymmetric transfer).

J24 146 Poulton (1982) considers asymmetric transfer to represent the J24 147 transfer of the strategy used in one condition to another J24 148 condition. Dornic & Fernaeus (1982) have shown that subjects in J24 149 noise are less able to switch between strategies. Smith J24 150 (1990a) asked subjects to carry out a dual task involving a J24 151 running memory task and a proportion perception task. The J24 152 difficulty, probability and priority of the tasks changed J24 153 frequently and noise produced a general impairment of the running J24 154 memory task. Overall, these results suggest that subjects in noise J24 155 are rather inflexible and unable to deal efficiently with changing J24 156 task demands.

J24 157 Another area which is relevant to the present discussion is the J24 158 effects of noise on response criteria. Broadbent & Gregory (1963, J24 159 1965) found that very loud noise (over 90 dB) produced more J24 160 confident assertions or denials in the second half of a vigilance J24 161 task, but reduced the number of reports of intermediate confidence. J24 162 Jones, Thomas & Harding (1982) examined the effects of noise on J24 163 recognition memory for prose items. They used a signal detection J24 164 analysis and showed that noise decreased values of beta for rare J24 165 names and increase beta for common names. Smith (1989b) also J24 166 found that noise influenced the use of response categories, J24 167 although, again, the nature of the noise effect was easily modified J24 168 by changing task parameters.

J24 169 Many of the studies of strategies of performance and noise have J24 170 been carried out with verbal memory tasks, because they offer a J24 171 variety of strategies, and shifts of dominance or preference can be J24 172 easily effected. An alternative approach is to develop tasks where J24 173 certain strategies are clearly used and examine whether noise J24 174 influences the efficiency of the different types of processing. J24 175 This technique was adopted here to investigate whether noise has J24 176 selective effects on different aspects of attention. The following J24 177 section gives a brief rationale for the choice of tasks in this J24 178 study of attention.

J24 179 Dimensions of attention

J24 180 Many of the details of the mechanisms of selective attention J24 181 are controversial (see Broadbent, 1982). However, two main types of J24 182 task have been used to study selective attention (see Kahneman & J24 183 Treisman, 1984), the first involving focused attention or J24 184 'filtering', and the second categoric search or 'pigeonholing'. The J24 185 first type of task involves selecting on the basis of some J24 186 characteristic simple feature such as location in space. The other J24 187 type requires the person to select a member of a category of J24 188 events. If one is trying to measure the efficiency of attention it J24 189 is clearly important which kind of task one uses. Indeed, J24 190 Broadbent, Broadbent & Jones (1989) have shown that the two J24 191 paradigms respond differently to changes in state (in their J24 192 experiment this was examined by comparing performance at different J24 193 times of day) and the main aim of the present study was to examine J24 194 the effects of noise on these different aspects of attention.

J24 195 The experiment reported here obtained a number of measures of J24 196 attention, the principal one being the difference between a J24 197 filtering and search paradigm. The filtering paradigm was based on J24 198 the method of Eriksen & Eriksen (1974). They required subjects to J24 199 perform a choice reaction time task to visual letters in a known J24 200 location. If distracting non-target letters were also presented, J24 201 the latency of the response to the target was increased. This J24 202 effect was greatest if the distractors were close to the target, J24 203 and this result has given rise to the analogy with a zoom lens or J24 204 spotlight (Broadbent, 1982). J24 205 J25 1 <#FLOB:J25\>The latter is particularly important as a number of J25 2 studies (including that of Oborne & Holton) appear to overestimate J25 3 the sensitivity of global measures such as reading comprehension or J25 4 measures of speed based on sentences or even whole texts. On J25 5 balance, the evidence points to a genuine disadvantage in J25 6 processing text on a VDU, particularly when task demands are high J25 7 (Wright & Lickorish, 1983; Zwahlen, Hartman, Ranjarajula, & J25 8 Escontrela, 1985). However, there is no generally accepted J25 9 explanation of the effects. Visual discomfort and headache appear J25 10 to be a consequence of staring at a flickering striped surface J25 11 (Wilkins, 1985a). Changes in reading speed must obviously relate in J25 12 some way to the distribution of eye movements over the screen or J25 13 page, involving, for example, changes in the frequency and/or J25 14 duration of fixations and the extent of saccades. However, these J25 15 variables have not previously been systematically examined using a J25 16 task approaching normal reading.

J25 17 An important observation regarding the effects of flicker on J25 18 eye movement control was made by Wilkins (1985b). He examined eye J25 19 movements as subjects made repeated saccades between two letters J25 20 embedded in a piece of text. The display was illuminated at 50 Hz J25 21 or 100 Hz. The results showed reliably longer saccades at 50 Hz J25 22 than at 100 Hz, the size of the effect being about one character J25 23 position. A further, similar experiment contrasted 100-Hz with J25 24 20-KHz flicker, and again saccades were longer at the lower rate - J25 25 although the effect was only reliable when the target letters were J25 26 wide apart. Wilkins is tempted to interpret this finding as J25 27 indicating an increase in saccade length in flicker, but it is J25 28 important to note that absolute accuracy differed in the two J25 29 conditions. At 50 Hz saccades overshot the target slightly, and at J25 30 100 Hz slightly undershot it. Without a baseline condition of J25 31 steady illumination it is not possible to interpret the results in J25 32 terms of a flicker-induced extension of saccade length. J25 33 Further, since Wilkins employed an external event (a tone) as a J25 34 signal for a saccade to take place, the possibility arises that the J25 35 obtained inaccuracy could relate to an increased proportion of J25 36 prematurely triggered saccades (this is discussed in greater detail J25 37 below).

J25 38 There are, in fact, two sources of evidence converging on the J25 39 prediction that interfering transient events, of the kind that may J25 40 arise in flicker, may disturb eye movement control. First, a study J25 41 by Becker, Kieffer, and Jurgens (1987) showed that both auditory J25 42 and visual signals delivered after the onset of a saccade may J25 43 modify its extent and velocity. The effect of the transient signal J25 44 varied somewhat between subjects but led overall to a change in the J25 45 final target location, shortening the saccade's length. Similarly, J25 46 Neary and Wilkins (1989) showed that when saccades are made between J25 47 two points on a raster display with a short-persistence J25 48 <}_><-|>phospor<+|>phosphor<}/>, transient 'ghost images' may be J25 49 seen in the vicinity of the target point. When lines rather than J25 50 points are used, the fixated target may appear to tilt momentarily J25 51 as a result of the top-to-bottom scan of the raster display. J25 52 Clearly, these data also show that brief stimulation during a J25 53 saccade can interfere with the normal suppression processes. The J25 54 second line of evidence derives from the possibility that saccades J25 55 may be triggered by transient external events that are J25 56 independent of the process of saccade computation itself. We assume J25 57 that when such external triggers occur, saccades are launched J25 58 towards their destination on the basis of the current (incomplete) J25 59 state of the on-going computation (O'Regan, 1989). For each J25 60 saccade, target location is continually updated on the basis of J25 61 available visual information. This becomes more refined with time J25 62 and is modulated by attentional processes (Deubel, Wolf, & Hauske, J25 63 1984). Thus, high spatial frequency information becomes available J25 64 later than low-frequency information. A 'premature' trigger will J25 65 result in a saccade launched effectively to the available centre of J25 66 gravity of the target region as a whole (Findlay, 1981, 1982). In J25 67 principle, such an effect may result in either undershoot or J25 68 overshoot, depending on the particular visual configuration (see J25 69 O'Regan, 1989, for an interesting discussion of this point), but it J25 70 is important to note that in normal reading, for the majority of J25 71 cases, large saccades over the page or screen surface are J25 72 likely to be directed towards a boundary, in which case the 'centre J25 73 of gravity' effect will lead to a preponderance of undershoots.

J25 74 It is possible that both mechanisms operate while a person is J25 75 reading flickering text. By definition, saccades interrupted in J25 76 flight will be of reduced extent. Similarly, but for a quite J25 77 different reason, large saccades prematurely triggered will tend to J25 78 land short. On the other hand, if the effects of premature J25 79 triggering on saccades of normal extent is to increase the J25 80 frequency of both under- and over-shoot, the overall result will be J25 81 a change in the variability in landing position. To this will J25 82 be added the possible effect of interference during a saccade, J25 83 although this may be more likely to affect those saccades directed J25 84 to remote targets, which will be longer in duration. The key J25 85 question is whether an increase in the number of mislocated J25 86 saccades produced in this way has important consequences for the J25 87 reader? The results from a number of recent eye movement studies J25 88 suggest a positive answer.

J25 89 It is now well established that the eye movement control system J25 90 is relatively intolerant of deviations from what is variously J25 91 termed the 'convenient' or 'preferred' viewing position for a given J25 92 word (O'Regan & Levy-Schoen, 1987; Rayner, 1979). Under laboratory J25 93 conditions, involving the inspection of words in isolation, the J25 94 'convenient viewing position' for the majority of words of average J25 95 length lies slightly to the left of the centre. The penalty for J25 96 landing outside this point is approximately 20 msec per letter. J25 97 However, this figure is a statistical average, primarily resulting J25 98 from increases in the probability of making additional corrective J25 99 saccades when the eye falls outside the convenient viewing J25 100 position. The degree to which these effects are found in normal J25 101 reading has recently been examined by McConkie, Kerr, Reddix and J25 102 Zola (1988). The 'preferred viewing position' (i.e. where the eye J25 103 actually falls, as distinct from what may be the best place for it J25 104 to be) shows a similar curvilinear relationship with letter J25 105 position, although the curve is much flatter (see also Vitu & J25 106 O'Regan, 1989). When account is taken of the 'launch site' (the J25 107 distance from the target word) from which a saccade is made, the J25 108 probability of landing on a particular letter, for words of all J25 109 lengths, is a roughly normal function about the word centre. J25 110 McConkie et al. (1988) identify a number of visual and oculomotor J25 111 processes that might underlie the obtained variability in landing J25 112 position, but for present purposes it is only necessary to note (a) J25 113 that a 'preferred viewing position' (i.e. the word centre) exists J25 114 for normal reading; and (b) that an increase in variability about J25 115 this position or a systematic shift away from it will change the J25 116 probability of additional corrective saccades (McConkie, Kerr, J25 117 Reddix, Zola, & Jacobs, 1989).

J25 118 In previous work we have shown that one class of eye movements J25 119 distinguishes a successful reading style. This involves the use of J25 120 large saccades (e.g. spanning more than 40 character positions) to J25 121 reinspect selected parts of previously read text. The reader can, J25 122 by this means, locate antecedents of anaphoric expressions (Murray J25 123 & Kennedy, 1988); deal with local syntactic ambiguity (Kennedy & J25 124 Murray, 1984; Kennedy, Murray, Jennings, & Reid, 1989); or J25 125 reanalyse an initially incorrect semantic reading (Pynte, Kennedy, J25 126 Murray, & Courrieu, 1988). Since the targets of such reinspections J25 127 are remote, their control cannot be based on physical J25 128 identification of the target: it must, rather, depend on the reader J25 129 maintaining a form of spatially coded information (Kennedy, J25 130 1987). Obviously, if flicker acts to perturb saccade control, this J25 131 class of saccades provides an ideal litmus test of its effect since J25 132 the saccades in question are large, increasing the probability of J25 133 disturbance; they have a distinct role to play in the reader's J25 134 process of comprehension; and their use effectively characterizes J25 135 an efficient reading style.

J25 136 The experiment reported here examined the effects of flicker on J25 137 landing position and on a number of other eye movement measures J25 138 under experimental conditions that called for the execution of a J25 139 number of large reinspecting saccades. Following Wilkins, two J25 140 flicker frequencies were used (50 Hz and 100 Hz), but in the J25 141 present study a control condition was introduced using steady J25 142 illumination. Two populations were studied - a group of typists who J25 143 reported relatively severe symptoms of headache and visual fatigue J25 144 associated with using word-processors; and a group of students with J25 145 little experience of VDUs. The procedure employed was based on that J25 146 of Kennedy and Murray (1987). Subjects read a sentence displayed on J25 147 a single horizontal line, and this was then followed by the J25 148 presentation of a stimulus word to the right of the sentence on the J25 149 same line. The subject's task was to indicate whether or not the J25 150 word was present in the displayed sentence. This task produces a J25 151 relatively high number of large saccades, involving both J25 152 reinspections of the target word and returns to the stimulus. It is J25 153 a situation more closely approximating normal reading than that J25 154 used by Wilkins (1985b) in that it encouraged, but did not demand, J25 155 that specific reinspections took place. The task is thus a good J25 156 analogue of some aspects of normal reading since when subjects do J25 157 reinspect, it is to a variable target position that is not J25 158 pre-defined (i.e. not involving a visual search) and yet is J25 159 specifiable for each item.

J25 160 Four hypotheses were tested: (1) that flicker will act to alter J25 161 the distribution of landing positions within the word in the 'first J25 162 pass' (i.e. in the phase of the experimental procedure before J25 163 presentation of the stimulus word); (2) that flicker will act to J25 164 alter the accuracy of large saccades directed towards a remote J25 165 target, causing an increased proportion to fall short; (3) that the J25 166 effects of an increased probability of mislocated saccades will be J25 167 an increase in the number of small 'corrective' eye movements; (4) J25 168 that the disruptive effects of flicker will be greater in a sample J25 169 of typists reporting symptoms of visual discomfort. The first three J25 170 hypotheses arise naturally from the earlier discussion. The fourth J25 171 requires some additional justification. Whether or not an increase J25 172 in the proportion of mislocated saccades seriously disrupts the J25 173 reader may relate to both subject and task demands. There is J25 174 evidence that professional typists adopt specific reading J25 175 strategies to optimize performance of a complex motor skill J25 176 (Shaffer, 1975). Typists pay more attention to local orthographic J25 177 features of the text and exhibit an 'eye-hand' span that is J25 178 much shorter than the normal eye-voice span (Butsch, 1932). Average J25 179 fixation duration during transcription typing is much longer than J25 180 normal, and average saccade extent is much shorter (Inhoff, Morris, J25 181 & Calabrese, 1986). We hypothesize that these habits will carry J25 182 over to unconstrained reading - in which case, small deviations J25 183 from intended eye position will be particularly disruptive for this J25 184 group of subjects. The choice of an appropriate comparison groups J25 185 is, however, problematic. Completely asymptomatic typists were J25 186 extremely rare in the population sampled. In any case, it is not J25 187 clear that such a group would comprise and effective control if, as J25 188 both Shaffer and Inhoff et al. suggest, typists, in general, J25 189 exhibit a characteristic reading style. In the present study, J25 190 therefore, comparisons were drawn between a group of typists and a J25 191 group of normal readers (students). The latter group permits direct J25 192 comparisons to be made between performance on the task employed in J25 193 the present study and the extensive literature on eye movement J25 194 control, which has almost exclusively made use of this population J25 195 of subjects. Analysis of the results will focus on differences in J25 196 reading strategy between the two groups and on the effects of J25 197 flicker for each group.

J25 198 Method J25 199 Materials and Design. A set of 60 items was J25 200 constructed, 30 of the form shown in (1), where the stimulus word J25 201 had occurred in the sentence, and 30 like (2), where the sentence J25 202 contained a close synonym of the target.

J25 203 1. The novels in the library had started to go mouldy with the J25 204 damp. novels

J25 205 J26 1 <#FLOB:J26\>It may be an over-simplification to conceptualize all J26 2 the different interests in terms of class struggle, particularly as J26 3 some of the interest groups involved and some of the alliances of J26 4 interests forged may defy analysis in conventional Marxist terms. J26 5 Nevertheless, there are class interests involved even though they J26 6 may not always conveniently reduce to one labouring class versus J26 7 one capitalist class.

J26 8 The transnational capitalist class, fractions of the labour J26 9 force, and other support strata that the TNCs have created, will J26 10 all increasingly identify their own interests with those of the J26 11 capitalist global system and, if necessary, against the interests J26 12 of their 'own' societies as the transnational practices of the J26 13 system penetrate ever deeper into the areas that most heavily J26 14 impact on their daily lives. The specific function of the agents of J26 15 transnational political practices is to create and sustain the J26 16 organizational forms within which this penetration takes place and J26 17 to connect them organically with those domestic practices that can J26 18 be incorporated and mobilized in the interests of the global J26 19 capitalist system. In order to do this the transnational capitalist J26 20 class must promote, outside the First World heartlands of J26 21 capitalism, a 'comprador' mentality throughout society.

J26 22 'Comprador' mentality

J26 23 A comprador mentality is the attitude that the best practices J26 24 are invariably connected with the global capitalist system. J26 25 Comprador mentality is either a 'cost' or a 'benefit' and whichever J26 26 way we look at it we are bound to beg a very important question. J26 27 This is the point at issue in the ideological struggle between J26 28 those who believe that TNCs will inevitably damage Third World J26 29 development prospects in the long run, as against those who believe J26 30 that there will be no development prospects without the TNCs. This J26 31 struggle revolves around opposing material interests of competing J26 32 classes and groups in all countries.

J26 33 There are those who see the destiny of the Third World as bound J26 34 up with the adoption of all that is 'modern', often embodied in the J26 35 products and practices of the TNCs. On the other hand, there are J26 36 those who are deeply suspicious of the modernization represented by J26 37 the TNCs, particularly where this is perceived as Western or US J26 38 dominance in culture, industry, warfare, science and technology. J26 39 This is a field riddled with dilemmas. Modern warfare and the J26 40 modern economy, for example, require increasingly higher levels of J26 41 technology and this cannot be avoided if any state wishes to J26 42 participate, or if threatened, survive in the contemporary J26 43 world.

J26 44 A battery of concepts, some of which have migrated from social J26 45 science jargon to the mass media, identify those on either side of J26 46 the divide. The academically discredited distinction between J26 47 traditional and modern is now common currency, while the J26 48 notions of inward-oriented and outward-oriented describe J26 49 those who look for guidance and sustenance to the resources of J26 50 their own groups as opposed to those who look outside, usually to J26 51 the West. Much the same idea is expressed by the distinction J26 52 between local and cosmopolitan orientation. As it is, J26 53 more or less by definition, an extra-First World phenomenon, the J26 54 question of the comprador mentality will be resumed in the next J26 55 chapter.

J26 56 It is difficult to think productively about 'modernization' for J26 57 many reasons, not least the problem of what is appropriate in J26 58 the adoption of innovation. The proper approach to development J26 59 lies, no doubt, somewhere between a slavish attachment to all J26 60 things foreign and an atavistic distaste for any type of change. An J26 61 illustration of how appropriate choices for the different groups J26 62 and classes in the capitalist global system can and do change is J26 63 provided by the analysis of counter movements to the global J26 64 capitalist project. These are movements that aim to undermine the J26 65 power of the TNCs, and the transnational capitalist class, and to J26 66 force people to think critically about the ways in which the system J26 67 as a whole promotes the culture-ideology of consumerism.

J26 68 Counter movements

J26 69 Although the agents of global capitalism are firmly in control J26 70 and the TNCs are clearly the dominant institutional force, there J26 71 are, as we approach the millennium, two counter movements that J26 72 could represent real threats to the global capitalist system. The J26 73 first, rich country protectionism, comes from within global J26 74 capitalism, but in the form of national as opposed to systemic J26 75 interests. The second, the Green movement, is a very widely based J26 76 and variegated collection of individuals and groups that includes J26 77 those on the fringes of the global capitalist system as well as J26 78 some who are fundamentally opposed to it, mainly from the J26 79 libertarian left but also from the authoritarian right.

J26 80 Protectionism J26 81 Protectionism is, of course, not a new phenomenon. Indeed, the J26 82 most potent argument against it may be that we know only too well J26 83 how protectionism contributed to the great depression of the 1930s. J26 84 Nevertheless, as the World Bank and other august proponents of the J26 85 perpetual increase of global trade never tired of reminding us J26 86 throughout the 1980s, many First World countries began to step up J26 87 protectionist measures in that decade. The free entry of goods J26 88 (particularly consumer goods) from abroad has never been a feature J26 89 of global trade. Restrictive measures have been directed at Third J26 90 World manufacturers whose electronic and electrical products, J26 91 garments, shoes, toys, sporting and household goods were said to be J26 92 unfairly flooding vulnerable First World markets. The interesting J26 93 twist to this issue is that it was Japan, clearly now a First World J26 94 country, and in the opinion of many now the most dynamic economic J26 95 power, that was often identified in the United States and in J26 96 Europe, as the worst offender, with the four East Asian NICs not J26 97 far behind.

J26 98 The tendency to protectionism is increased by the belief that a J26 99 substantial part of TNC manufacturing industry is 'footloose'. J26 100 Transnational corporation production tends to be globally J26 101 integrated into vertically organized production processes. Offshore J26 102 plants tend to be financially controlled from abroad, they tend to J26 103 be rented rather than owned, and their managers tend to have J26 104 cosmopolitan rather than local perspectives. All of these factors J26 105 weaken the ties that such businesses have with the communities in J26 106 which they are located and make it less difficult for them to close J26 107 down and/or relocate if and when business conditions deteriorate in J26 108 one country relative to other countries. This happens in the First J26 109 World as well as in the Third World. For example, Hood and Young J26 110 (1982) document the closure of several old established US firms in J26 111 Scotland to seek cheaper production sites elsewhere. The mobility J26 112 of the TNCs, the job losses that usually follow, and the J26 113 identification of 'cheap imports' with goods previously produced at J26 114 home, increase protectionist pressures among labour and small J26 115 domestic capitalists alike.

J26 116 While protectionism as a transnational political force appears J26 117 to have little likelihood of success in the foreseeable future, the J26 118 threat of it is ever present as a reminder that the orderly J26 119 progress of global trade in the interests of the TNCs has to be J26 120 maintained and those who transgress will be punished. This works J26 121 both ways. First World markets are open to Third and Second World J26 122 goods only as long as their markets are also open to First World J26 123 goods and, increasingly, services.

J26 124 Protectionaism is not a serious counter movement to global J26 125 capitalism because if it was successful it could do great damage to J26 126 the system and, ultimately, destroy it. All parties realize this, J26 127 and so protectionism acts as a bargaining counter for the rich, and J26 128 a bluff for the poor, and mainly comes to life in its use as a J26 129 rhetorical device to satisfy domestic constituencies. For example, J26 130 desperate politicians tend to fall back on it to appease working J26 131 class voters in the United States and the United Kingdom. The Green J26 132 movement is much more serious, actually and potentially.

J26 133 The Green movement

J26 134 With the sole exception of the global communist movement, the J26 135 Green or environmentalist movement presents the greatest J26 136 contemporary challenge to the global capitalist system. This is J26 137 paradoxically confirmed by the fact that both capitalist and J26 138 socialist economists, politicians and ideologues are increasingly J26 139 trying to jump on to the Green bandwaggon and to appropriate its J26 140 policies for themselves. This is not surprising, because although J26 141 capitalists and socialists are usually reluctant to spell out their J26 142 plans of global domination, Green politics are largely based on a J26 143 straight-forward conception of planet earth and what needs J26 144 to <}_><-|>de<+|>be<}/> done at the global level to sustain human J26 145 life on it.

J26 146 The key threat that Green politics poses to the capitalist J26 147 global system is in the matter of the consumption of non-renewable J26 148 resources. While Green politics are based on the belief that the J26 149 resources of the planet are finite and have to be carefully tended, J26 150 global capitalist politics are based on the belief that the J26 151 resources of the planet are virtually infinite, due to the J26 152 scientific and technological ingenuity released by the capitalist J26 153 system that will ensure unlimited replacement or substitution of J26 154 resources as they are used up.

J26 155 Green politics are closely connected with the emergence of a J26 156 critical consumer movement. The idea of consumerism has J26 157 experienced an important dialectical inversion in recent years. It J26 158 is commonly used in two senses. In this book I use it to denote an J26 159 uncritical obsession with consumption. It is, however, also J26 160 commonly used in an opposite sense, as in the consumer movement's J26 161 version of consumerism, to denote suspicion of consumer goods, a J26 162 wish to know more about how they are produced and who produces J26 163 them. This version of consumerism can lead to a radical critique of J26 164 consumption. To minimize confusion, I shall use consumerism J26 165 where I mean the first, and the consumer movement to J26 166 denote the second.

J26 167 Kaynak (1985, p.15) defines the consumer movement "as a J26 168 movement seeking to increase the rights and powers of buyers in J26 169 relation to sellers" and presents a useful historical J26 170 account from the first cooperative society, founded in Scotland in J26 171 1769, to the present. He connects the rise of consumer movements in J26 172 different countries with the position of the country in the world J26 173 market, but from a marketing rather than a political economy point J26 174 of view. His discussion of consumer movements in the Third World, J26 175 however, does confuse the two meanings of consumerism. Food riots J26 176 in North Africa, he writes, "are examples of what LDC J26 177 consumers are concerned with - the right to consume" J26 178 (p.20). But this misses the real distinction between diametrically J26 179 opposed beliefs based on entirely different conceptions of the J26 180 satisfaction of human needs. Thus, it is to the culture-ideology of J26 181 consumerism, and how it is broadcast in the global capitalist J26 182 system through a variety of transnational practices, that we must J26 183 now turn.

J26 184 Cultural-ideological transnational practices

J26 185 There are many who argue that the key to hegemonic control in J26 186 any societal system lies not in the economic nor in the political J26 187 sphere, but in the realm of culture and ideology. Those for whom J26 188 this idea is a novelty may be surprised to learn that is was the J26 189 writings and political practice of a Marxist, and a communist J26 190 militant at that, which were largely responsible for the present J26 191 currency of this view among radical thinkers. Antonio Gramsci, who J26 192 spent most of his adult life in one of Mussolini's prisons, J26 193 elaborated on Marx's insight that the ruling ideas of an epoch are J26 194 the ideas of its ruling class, to create a theory of hegemony and a J26 195 theory of classes of intellectuals whose function it is in any J26 196 literate society to propagate or to challenge these leading ideas. J26 197 Gramsci's Prison Notebooks represent not only a stirring J26 198 monument to the human spirit under adversity, but a significant J26 199 turning-point in the history of Marxist ideas and their relevance J26 200 for the twentieth century. This is partly because in the sphere of J26 201 culture and ideology the material conditions have changed to such J26 202 an extent that what Gramsci was arguing about hegemonic processes J26 203 in the 1930s has become more, not less, relevant today than it was J26 204 then. To put the point graphically, while Marx and his J26 205 nineteenth-century comrades would have no great difficulty in J26 206 recognizing the economic and the political spheres today, despite J26 207 the major changes that have undoubtedly taken place in the last J26 208 hundred years, in the cultural-ideological sphere the opportunities J26 209 for hegemonic control on a global scale have changed out of all J26 210 recognition.

J26 211 J27 1 <#FLOB:J27\>THE CLAIMS OF A DOG

J27 2 Our environment has many unpleasant features. One is dog J27 3 excrement. Piles and pats of it abound, waiting to slip up the J27 4 inattentive pedestrian. But what do we do if we see a squatting J27 5 dog? We either express haughty contempt, turn away, or blame the J27 6 owner. Would any of us kick or punch the dog? Probably not; instead J27 7 we are more likely to shout at the person holding the lead. On the J27 8 one hand we seem to believe that the dog is not entirely J27 9 responsible for its actions, on the other we feel that it should J27 10 not be deliberately hurt simply because it is doing what it must. J27 11 We are not certain what we should do about the dog that fouls our J27 12 front garden, indeed we are not sure we can rightly do anything. J27 13 Our treatment of the animal is difficult. Animals present society J27 14 with ambiguities; with companionship, food, clothing, fun, but also J27 15 with demands for compassion, abstinence and kindness. They disgust J27 16 and please us; we can do with them what we will yet pull back with J27 17 horror from open cruelty. The dog makes claims upon us, although we J27 18 are not entirely sure what those claims are. But they still J27 19 profoundly influence our treatment of the animal.

J27 20 This book is about those claims and how they affect the way J27 21 individuals relate to, and understand, animals. At its broadest, J27 22 this book is concerned with the relationship between animals and J27 23 society, but mostly it is about the special claims which are made J27 24 for animals' rights. This explanation of animal rights will J27 25 discover many surprising incidents in the attitude and beliefs of J27 26 humans. I will emphasise the situation in Britain - perhaps J27 27 England is better - and only refer to events in other J27 28 countries if they support any case or illustrate it particularly J27 29 vividly. This is not because Britain is more respectful of animal J27 30 rights than anywhere else (just because we see ourselves as a J27 31 nation of 'animal lovers', we should be wary of falling into any J27 32 national chauvinism), but simply because Britain is the environment J27 33 I know best. I hope that the story told in this book has some J27 34 relevance elsewhere. In a sentence, I will explore the social J27 35 processes and relationships which lie behind the many assertions J27 36 which are made in Britain for the moral relevance of animals, and J27 37 show how those relationships influence people's lives.

J27 38 Concern for animals has a long history. To give just one J27 39 example, when he wrote the great radical Romantic poem Queen J27 40 Mab in about 1812, Shelley bewailed the treatment of animals J27 41 by which "the bull must be degraded into the ox, and the J27 42 ram into the wether" (Shelley 1905: 818). Yet he was only J27 43 interested in the sufferings of animals because they reflected J27 44 "manhood blighted with unripe disease" (Shelley J27 45 1905: 770). He was not concerned with the relevance of animals so J27 46 much as with the decay of humanity. The idea of animal rights is J27 47 rather different precisely because it does say something tangible J27 48 about animals and advocates that individuals should think about how J27 49 they treat them. Animal rights does not seem to be as selfish as J27 50 Shelley's self-serving pity for a castrated ram. The wish J27 51 to say something morally irreducible, something real, about animals J27 52 and then establish that statement as a guide for human behaviour is J27 53 a relatively recent phenomenon. In Britain the concern with J27 54 animals' rights has been an especially widespread and lively social J27 55 issue since the middle of the 1970s (although, as we will see, the J27 56 roots of the idea are to be found in the eighteenth century). The J27 57 surge of interest can be attributed to the debates stimulated by J27 58 the book Animal Liberation, written by the Australian J27 59 philosopher Peter Singer and first published in Britain in 1976.

J27 60 This chapter will give an introductory survey of the three J27 61 clearest, most influential, and sustained statements of the claims J27 62 of animals: that by Singer and later contributions by Tom Regan and J27 63 Stephen Clark. It might seem curious that the whole complex issue J27 64 of the relationship between animals and society is so quickly J27 65 reduced to just three philosophy books: surely they cannot say J27 66 everything that is important? No, they cannot, but between them J27 67 Singer, Regan, and Clark do cover virtually all the ground which J27 68 must be mapped in order to understand modern moral attitudes J27 69 towards animals.

J27 70 The argument of Animal Liberation is powerfully J27 71 simple. Singer writes from within the utilitarian tradition of J27 72 moral philosophy. Broadly speaking, utilitarianism holds to the J27 73 central theme that pleasure is good and pain is bad. As John Stuart J27 74 Mill put it, utilitarianism asserts that "pleasure and J27 75 freedom from pain, are the only things desirable as ends" J27 76 (Mill 1910: 6). It goes beyond individualistic hedonism. Certainly, J27 77 Mill tells us, the individual should follow the path of pleasure J27 78 and avoid pain (such a life is morally good), but the individual J27 79 lives in a social world and, consequently, should act to promote J27 80 global pleasure, or at least defend the preference to avoid pain. J27 81 Utilitarianism asks the individual to aggregate the consequences of J27 82 his or her actions for the promotion of pleasure and avoidance of J27 83 pain, and demands that she or he should morally only follow that J27 84 course which causes more pleasure than it does pain. This is a J27 85 morality which can be imagined as a pair of scales, and the moral J27 86 act is the one which tips the individual and social balance in J27 87 favour of pleasure. An act is not moral as such; rather, J27 88 morality is a product of the consequences which the act has on the J27 89 balance of pleasure and pain. Utilitarianism is inherently social. J27 90 Singer holds to a slightly modified version of Mill's position; for J27 91 Singer the moral act is that which takes account of the preference J27 92 of all sentient creatures not to experience increases in pain or J27 93 suffering.

J27 94 Now, utilitarianism can have important things to say about the J27 95 social treatment of animals because, if animals are able to J27 96 experience pleasure and pain, it is logical to assume that they J27 97 prefer not to undergo any increase of suffering, and they should J27 98 therefore be included in the calculations which must be made to J27 99 deduce the moral consequences of any act. For utilitarians, the J27 100 experience of pleasure and pain cannot be ignored simply on the J27 101 grounds that the experiencing subject is an animal. Utilitarians J27 102 like Singer assert that the preference for pleasure rather than J27 103 increased pain demands equal consideration whether the experiencing J27 104 subject is the Queen of England or a laboratory rabbit. Peter J27 105 Singer calls this the principle of equality and, J27 106 obviously, it is a prescription for equal ethical J27 107 consideration, and not a description of equality. The J27 108 principle refers to the equal consideration of interests and J27 109 preferences (Singer 1976: 5). All beings that can suffer pain have J27 110 an equal interest in avoiding it. Singer has no doubt that this is J27 111 the essence of morality:

J27 112 If a being suffers there can be no moral justification J27 113 for refusing to take that suffering into consideration. No matter J27 114 what the nature of the being, the principle of equality requires J27 115 that its suffering be counted equally with the like suffering - in J27 116 so far as rough comparisons can be made - of any other being. If a J27 117 being is not capable of suffering, or of experiencing enjoyment or J27 118 happiness, there is nothing to be taken into account. So the limit J27 119 of sentience (using the term as a convenient if not strictly J27 120 accurate shorthand for the capacity to suffer and/or experience J27 121 enjoyment) is the only defensible boundary of concern for the J27 122 interests of others. To mark this boundary by some other J27 123 characteristic like intelligence or rationality would be to mark it J27 124 in an arbitrary manner. Why not choose some other characteristic, J27 125 like skin color? (Singer 1976: 9)

J27 126 The only morally relevant characteristic is the ability to J27 127 suffer and the preference not to, and so according to Singer the J27 128 moral terrain is a plateau of morally equal preferences in which J27 129 features such as skin colour, sex or species are as morally J27 130 important as the differences between green and brown eyes. Indeed, J27 131 to subordinate the question of sentience to some secondary feature J27 132 is utterly reprehensible, and just as overriding sentience on the J27 133 basis of skin colour is morally condemned as racism, or on the J27 134 basis of sex as sexism, any qualification of the principle of the J27 135 equal consideration of interests on the grounds of species should, J27 136 Singer asserts, be rejected as speciesism. The word was first J27 137 used by Richard Ryder in a critique of the use of live animals in J27 138 medical experiments (Ryder 1975) and Peter Singer defines the J27 139 speciesist as a person who "allows the interests of J27 140 his own species to override the greater interests of members of J27 141 other species" (Singer 1976: 9). Animal J27 142 Liberation asserts that the truth of animals is their ability J27 143 to suffer on a morally equal basis to ourselves, and they have a J27 144 preference in the avoidance of suffering as we likewise do. To J27 145 argue the contrary is to give voice to speciesism. Singer intended J27 146 his book to be read and adopted as a persuasive repudiation of the J27 147 acts of speciesism: the point is not so much that animals have a J27 148 right to be treated well, rather it is that if we are to be good J27 149 utilitarians, we should not act in any way that violates their J27 150 preference not to suffer.

J27 151 This is important. Singer does not believe that animals have J27 152 rights. No, he is more concerned with the morality (or otherwise) J27 153 of the acts of society and individuals. This is why the book is J27 154 called Animal Liberation; Singer argues for moral, J27 155 utilitarian acts, not a respect for intrinsic rights. He wants the J27 156 acts which will liberate animals from speciesism. Indeed, in J27 157 subsequent essays, Singer has often taken pains to distance himself J27 158 from any talk of rights: "when I talk of rights, I do it J27 159 ... as a concession to popular rhetoric (Animal J27 160 Liberation was not written primarily for J27 161 philosophers)" (Singer 1980a: 327). Singer is more J27 162 concerned with acts, a position which has caused him to fall J27 163 between two stools. Popularly he has been taken up as a prime J27 164 advocate of animal rights (it must be said that Animal J27 165 Liberation is not as explicit as it might have been on why it J27 166 rejects talk of rights: Singer was quite prepared to play the J27 167 rights card when it leant the case greater strength). Meanwhile, J27 168 rights theorists have expressed grave doubts over the applicability J27 169 of utilitarian calculations to the relationship between society and J27 170 animals. This is especially true of Tom Regan.

J27 171 Regan tears out the heart of Singer's thesis. Singer asks that J27 172 account be taken of the consequences of an act for the preference J27 173 of sentient beings not to suffer, before that act is performed. It J27 174 should only be carried out if, on aggregate, it does not increase J27 175 the sum of pain amongst morally relevant creatures. For Regan, such J27 176 an approach not only betrays a "significant conservative J27 177 bias" because it considerably reinforces existing social J27 178 attitudes by making them the yardstick against which all else is J27 179 measured (Regan 1984: 138), but, more importantly, the utilitarian J27 180 emphasis on aggregates can provide no foundation for the claims of J27 181 an individual over the group.

J27 182 Singer does not talk about rights; he gives the individual no J27 183 priority outside of the global equality of the preference or J27 184 interest not to suffer. Tom Regan believes that this position is J27 185 quite unsatisfactory. He is right to say that according to the J27 186 utilitarian approach to morality:

J27 187 we must choose that option which is most likely to J27 188 bring about the best balance of totalled satisfactions over J27 189 totalled frustrations. Whatever act could lead to this outcome is J27 190 the one we ought morally to perform - it is where our moral duty J27 191 lies. And that act clearly might not be the same one that would J27 192 bring about the best results for me personally, or for my family or J27 193 friends, or for a lab animal. The best aggregated consequences for J27 194 everyone concerned are not necessarily the best for each J27 195 individual. (Regan 1985: 20)

J27 196 Think of Raskolnikov in Crime and Punishment. J27 197 J28 1 <#FLOB:J28\>As we have seen, this authority extends beyond the J28 2 economic realm to management of the reproductive process itself, J28 3 and the business of deciding who may marry when.

J28 4 Why do younger Karimojong put up with their subservience to the J28 5 elders? (Why do sons everywhere put up with their subservience to J28 6 their fathers?) Why do they not rebel and seize power and wealth? J28 7 In a sense they do, as time moves on and the pressure to turn over J28 8 the generation sets increases. The phase of domestic fission is J28 9 therefore very evocative of class conflict, raising as it does J28 10 questions about access to productive resources. The elder J28 11 generation is concerned to hang on to its authority and privileges, J28 12 but passing time, the aging and the death of the elders, whittles J28 13 down their numbers and their claims to authority. They certainly do J28 14 their best to hang on as long as possible; one ideological tactic J28 15 at their disposal is their power to collectively curse juniors who J28 16 have displeased them. But pressure builds up inexorably among the J28 17 lower ranks for a succession ritual - which only the elders can J28 18 sanction. The aging juniors resent their economic and political J28 19 subordination to the old brigade, and are likely to be particularly J28 20 hostile to their 'poaching' brides from an age cohort which the J28 21 juniors consider to be rightfully theirs. If promotion to elderhood J28 22 may be thought of as a kind of upward class mobility, the J28 23 Karimojong are remarkable in that they synchronize this process J28 24 ritually across the entire population. The Red-Yellow opposition J28 25 could be thought of as an uneasy intergenerational truce. The J28 26 competing interests are the familiar ones of domestic fission, but J28 27 here they are given a concerted voice. In most societies the J28 28 generations do not have this sort of solidarity, and the youngsters J28 29 have to struggle on a more individual basis to make their way in J28 30 society. In Karimojong social organization, the intergenerational J28 31 tensions which are a dynamic force in all societies become a J28 32 coherent central motif in their historical understanding of J28 33 themselves.

J28 34 Our own thinking may not be so very different. But in our J28 35 societies each generation, each overlapping cohort of subjective J28 36 viewpoints, is left to form its own collective vision of society J28 37 and experience of history. It draws on certain generally available, J28 38 chronological cues to categorize itself as a generation: the J28 39 Depression, Vietnam, etc. There is a moral dimension: each J28 40 generation likes to claim that it has virtue on its side, and has J28 41 had to face more serious economic and political challenges. This J28 42 can go beyond a way of speaking to a way of acting. If we doubt the J28 43 political force of generational conflict we should recall the J28 44 worl-wide revolutionary fervour of the junior 'underclass' in the J28 45 years 1968-72, and the 1989 revolt against the gerontocracy in J28 46 China. These movements are so impressive because they translate the J28 47 squabbles between parents and children into a much wider arena than J28 48 the household, pitching one generation as a mass against the other. J28 49 Husbands can feel much the same categorical alienation when their J28 50 wives resort to the rhetoric of feminism.

J28 51 Periodically, a generation makes itself evident by force of J28 52 numbers: by dying in a war, or by deferring marriage in a period of J28 53 economic slump. These effects are amplified by the demographic J28 54 scale of industrial societies. As Reuben Hill puts it, J28 55 "Each cohort encounters at marriage a unique set of J28 56 historical constraints and incentives which influence the timing of J28 57 its crucial life decisions" (1970: 322). How, we may ask, J28 58 did households at different stages of development respond to the J28 59 Great Depression of 1932-3? It seems likely that those in decline J28 60 suffered most heavily, while those in the early stages of expansion J28 61 were better placed to take advantage of Roosevelt's New Deal J28 62 programmes. The point is that people neither make history nor J28 63 respond to historical events as an undifferentiated mass. The J28 64 reproductive cycle produces categorically distinct interests across J28 65 a whole population, and how each generation responds will depend on J28 66 its particular life experiences as well as its sheer size. As the J28 67 proportion of old people increases in our industrial countries we J28 68 can be sure that they will do their <}_><-|>bast<+|>best<}/> to J28 69 hang on to their wealth and voting power to secure their J28 70 welfare.

J28 71 Although competition between generations may not be as J28 72 concerted as it is among the Karimojong, the 'motor' of the J28 73 reproductive process is massive in its effects. Most influential J28 74 are the acquisitive pressures on the younger generation, which give J28 75 society much of its historical momentum. Juniors must accumulate, J28 76 must displace their seniors, and must in their turn ultimately J28 77 relinquish control, divest, decumulate. But resources are not J28 78 passed on mechanically and without a struggle. They must also be J28 79 created: new lands must be colonized, livestock bred and J28 80 technical innovations made. This progressive force is J28 81 continually pitted against the constraints of environment and the J28 82 limitations of technology. But all around us we see evidence of the J28 83 extent to which youthful ingenuity has prevailed. This rhythm of J28 84 progress is thwarted only by the countervailing rhythm of aging and J28 85 death.

J28 86 Consider the possibility of a whole society which consisted J28 87 exclusively of a younger generation: that, surely, would be J28 88 extremely dynamic and progressive. Perhaps surprisingly, quite a J28 89 few such societies exist. Almost invariably they have one central J28 90 purpose: the pursuit of rapid economic development. An example is J28 91 the Malaysian government's enormous land development programme. J28 92 Between 1957 and 1980 more than a hundred thousand households from J28 93 poor rural areas were resettled on two hundred rubber and oil palm J28 94 plantations cleared from the jungle. The many applicants were J28 95 screened on the basis of 'need' (family poverty) and 'suitability' J28 96 (efficient, able-bodied labourers). Whole communities were thus J28 97 created consisting of younger married couples with several J28 98 children, additionally motivated by the fact that they had little J28 99 or no land at home to support them. This active and acquisitive J28 100 labour force helped to guarantee the remarkable economic success of J28 101 the Malaysian land development schemes for several decades. The J28 102 abnormality of these communities was evident in many ways, for J28 103 example in the chronic overloading of the local schools. The Bukit J28 104 Besar scheme, which I visited in 1972, had ten simultaneous first J28 105 grade classes spread out over morning and afternoon sessions.

J28 106 By the 1980s, however, the schools were emptying and the J28 107 children were taking their educational advantages to the expanding J28 108 towns and cities. After reaping the short-term benefits of this J28 109 'unnaturally' skewed population, the Malaysian government is now J28 110 confronting the long term costs: both the settlement scheme J28 111 populations and the tree crops they have been tending are aging J28 112 simultaneously. Dwindling crop yields are exacerbated by the J28 113 declining physical capacity of the settlers themselves, and J28 114 additional labour (much of it drawn from neighbouring Indonesia) J28 115 has to be hired to do the work. Billions of dollars of investment J28 116 in more than two hundred schemes is now under threat, and the J28 117 government is understandably vexed to see its ardent young J28 118 proletarians turn into aged, underproductive capitalists. Recently J28 119 it has even felt obliged to renege on original contracts to hand J28 120 the land over to the settlers. The dream of profitable, J28 121 self-sustaining smallholder communities has evaporated, and the J28 122 settlement schemes remain government-managed estates.

J28 123 Evidently, the Malaysian planners either did not understand or J28 124 chose to ignore the significance of the reproductive process in the J28 125 making and breaking of their land development schemes. Experience J28 126 tells us repeatedly that the success of economic projects is not J28 127 guaranteed solely by the efficiency of productive arrangements. J28 128 Either optimism, shortsightedness or plain ignorance blinds us to J28 129 the fact that the generational cycle involves decumulation as well J28 130 as accumulation. Folk wisdom may know better - 'riches to rags in J28 131 three generations' goes the old saying. Individuals move out of one J28 132 social class and into another, but the consolidation of a J28 133 dominant class undoubtedly depends on its capacity to resist this J28 134 sort of mobility. For a minority to retain control of capital it J28 135 has to keep outsiders at bay, and prevent its own offspring from J28 136 dropping out. This is hard work, involving the sort of economic, J28 137 political and cultural tactics we shall consider in the next J28 138 two chapters. Reproduction poses the most serious threat to the J28 139 maintenance of class power. But if reproductive processes are J28 140 excluded from a definition of class, then class itself is doubly J28 141 deficient as an analogy for understanding the inequalities of J28 142 generation and gender, and their impact on modern society.

J28 143 However conspicuous they may be, political and economic J28 144 definitions of generation and gender are always secondary J28 145 social constructions of the fundamental reproductive J28 146 distinctions. This is why trying to explain inequalities of gender J28 147 and generation from a political-economic point of view has proved J28 148 so frustrating. For a start, the two sorts of categorization (class J28 149 on the one hand, and generation and gender on the other) are not J28 150 mutually exclusive. Thus, Dyson-Hudson is at pains to point out J28 151 that Karimojong generation sets do not equalize wealth: there are J28 152 still richer and poorer elders, more and less advantaged youths. In J28 153 a broad survey of age-set organization in Eastern Africa, Baxter J28 154 and Almagor make the point that the sets do not corporately own or J28 155 control productive property: that remains vested in households and J28 156 patrilineal descent groups. Consequently, "When the J28 157 interests of domestic and familial loyalties diverge from those of J28 158 set loyalties, set loyalties are likely to be the ones which J28 159 give." Indeed, as domestic issues claim more of their J28 160 attention, older men may "ooze away" from J28 161 involvement in their age set (1978: 22,13).

J28 162 In a zealous effort to apply Marxian explanations to J28 163 pre-industrial societies, the French anthropologist P.P. Rey (1979) J28 164 has used the model of class conflict in industrial societies to J28 165 explain the relations between "exploiters and J28 166 exploited" in societies like the Karimojong. Juniors keep J28 167 their own company in a kind of working-class solidarity in the face J28 168 of the economic and political domination of the elders. In reaction J28 169 to Rey's argument, but still from a materialist perspective, J28 170 Meillassoux points out that people distinguished as a generation J28 171 within a community cannot constitute a class, because the one is J28 172 defined by the passage of time and the other by access to material J28 173 resources. Elders can only be elders by virtue of having been J28 174 juniors; but people do not have to be proletarians before they can J28 175 become capitalists (1981: 80). Classes, Meillassoux implies, are J28 176 fixed, but people are not: "Age, even understood in its J28 177 social sense, is only a transitional moment in the life of an J28 178 individual." However, classes are only 'fixed' because we J28 179 choose to think of them so. A central interest in maintaining class J28 180 status is to resist downward mobility within and between the J28 181 generations - in other words, to counteract the turbulence of the J28 182 reproductive process.

J28 183 Gender and political economy

J28 184 Similar problems have arisen with political-economic J28 185 explanations of gender. The central dilemma for feminists thus J28 186 becomes one of loyalties: to a 'class' of women, or an economic J28 187 class which is comprised of both sexes. Classes are historical J28 188 developments, but women's subordination seems timeless. Playing men J28 189 at their own materialist game, likening women to a revolutionary J28 190 proletariat, may actually obscure the issues, for example by J28 191 reducing explanations of women's subordination to arguments about J28 192 the nature and value of their labour, and whether or not it should J28 193 be 'waged'. It seems that feminists are gradually returning to the J28 194 view that an explanation of women's subordination must be sought in J28 195 the organization of reproduction. Rossi declares firmly, J28 196 "Gender differentiation is not simply a function of J28 197 socialization, capitalist production, or patriarchy. It is grounded J28 198 in a sex dimorphism that serves the fundamental purpose of J28 199 reproducing the species" (1985: 161). However real and J28 200 potent class distinctions may be, they are also historically more J28 201 recent than reproductive categories of gender and generation, and J28 202 are thus in some sense more superficial. Generation and gender are J28 203 categories which divide households, social classes and whole J28 204 societies. Social classes, on the other hand, divide whole J28 205 societies but do not divide households. Class conflict J28 206 widens the gap between households; it does not necessarily close J28 207 the generation or gender gap within families. Nor does class J28 208 differentiation eradicate distinctions of gender and generation, it J28 209 superimposes itself on them. J28 210 J29 1 <#FLOB:J29\>In her more recent work Pahl has further divided J29 2 the categories by identifying female and male dominated 'patterns J29 3 of control' (Pahl, 1989 and 1990; see also Vogler, 1989).

J29 4 Following Pahl's work the focus of most research in this area J29 5 to date has been on access to, and the control and management of, J29 6 financial resources. However, we would argue that a broader J29 7 definition of the resources involved is needed to analyse J29 8 satisfactorily the internal workings of the household. These J29 9 include the domestic division of labour; the allocation of time to J29 10 household members; the allocation of consumption goods and J29 11 services; and access to consumer durables which nominally belong to J29 12 all household members. While this broader definition provides a J29 13 more comprehensive account of the internal workings of the J29 14 household it also poses considerable problems when constructing a J29 15 short series of survey questions. Most questionnaire surveys have J29 16 neither the space nor time to cover the full range of items which J29 17 would be needed to cover satisfactorily all the areas mentioned. It J29 18 is in establishing the key indicators from which we can infer how J29 19 and why people allocate the resources available to them, and how J29 20 this process changes over the life course in relation to factors J29 21 which are both internal and external to the household, that J29 22 qualitative information should prove useful.

J29 23 The qualitative data already collected has produced a wealth of J29 24 information on the often subtle and complex negotiations between J29 25 household members concerning the allocation of money and resources. J29 26 As one <}_><-|>women<+|>woman<}/> describes:

J29 27 I allot myself approximately about 250 (pounds per J29 28 month) for food, which might sound a lot for three of us but we've J29 29 got two big dogs that eat a lot and the two men eat like they've J29 30 got hollow legs and then he (her husband) has a set amount which is J29 31 what we sort of discuss, I mean I don't pin him down, he said a set J29 32 amount which I think is extremely generous and I don't know what he J29 33 does with it because he never buys clothes, but anyway, he has this J29 34 money and he doesn't drink and he only smokes about five cigarettes J29 35 a day and I don't actually have a set amount of pocket money but if J29 36 I want something and I don't often buy things, but I really want a J29 37 new dress or something and there's money in the account I'll buy J29 38 it. But I always say to him, look, you know I've bought a new J29 39 dress, I hope you don't mind.

J29 40 It would be foolhardy to suggest that this depth and complexity J29 41 of interpersonal negotiations can be successfully replicated in the J29 42 context of a small number of survey J29 43 <}_><-|>questons<+|>questions<}/>. However, this does not imply J29 44 that asking about the distribution of resources within the J29 45 household in the context of a survey is a useless exercise. Indeed J29 46 there is a definite need for a large scale data set in this area of J29 47 research. What it does highlight however is the need for J29 48 qualitative and quantitative research to be mutually informative J29 49 and provide the possibility for analyses in which both types of J29 50 data contribute to and illuminate the analysis as a whole. While J29 51 the majority of social researchers would probably agree with these J29 52 aims, in practice they are often not as easy to reconcile, even J29 53 though the methodological divisions are by no means as clear cut in J29 54 the practice of social research as the epistemological debates J29 55 which posit a necessary link between theoretical perspective and J29 56 method of data collection assert. As Bryman (1988) argues there may J29 57 not be a necessary or one to one correspondence between J29 58 epistemology and methods as is often asserted in the methodological J29 59 literature. We discuss this point further in the final section.

J29 60 Problems of classification

J29 61 The qualitative project has highlighted several problems with J29 62 existing models of intra-household allocations, some of which are J29 63 common to all classifications of social phenomena designed to J29 64 clarify the world we live in, while others relate specifically to J29 65 the needs and objectives of the panel study. The first point J29 66 relates to the general problems of categorisation and J29 67 classification and the second to the specific issues raised by the J29 68 longitudinal nature of the panel study. When empirical work based J29 69 on an already existing taxonomy begins it becomes almost J29 70 immediately apparent that there are considerable numbers of cases J29 71 which do not fit neatly into any one category. The preliminary J29 72 analysis of our interviews (Ritchie and Thomas, 1990:3) indicates J29 73 that of the nineteen couples interviewed individually only four J29 74 (three common pool, and one independent manager) could certainly be J29 75 assigned to one of the categories of the model outlined above. J29 76 Another two (one independent manager, one common pool) could be J29 77 assigned if the category was broadly interpreted. In the remaining J29 78 thirteen cases there were either significant features of a category J29 79 missing or there seemed no appropriate category at all. As Jan Pahl J29 80 observes in relation to her own data:

J29 81 In reality the different types were not mutually J29 82 exclusive, and for many couples it was a case of choosing the type J29 83 of allocative system which most closely resembled the way in which J29 84 they organised their money. As so often in the social sciences, J29 85 classification involved simplification. It is therefore justifiable J29 86 to be sceptical about the adequacy of the typology. J29 87 (1989:79)

J29 88 Thus, by simplifying in order to clarify, the subtleties of the J29 89 interactions between household members are subsumed within general J29 90 categories which have a tendency to become reified. Instead of J29 91 proceeding via a process of analytic induction, by which J29 92 categorisation as hypothesis is subject to falsification, the J29 93 temptation is to either squeeze recalcitrant cases into J29 94 pre-existing boxes or to find reasons for omitting or ignoring them J29 95 altogether. Of course the tendency to view existing classifications J29 96 as given is shared by respondents, and may be reinforced by the J29 97 design of a survey question. When faced with a categorisation such J29 98 as that employed by the SCELI project (which used an elaborated J29 99 version of Pahl's original model) and asked: "People J29 100 organise their household finances in different ways. Which of the J29 101 methods on this card comes closest to the way that J29 102 <}_><-|>your<+|>you<}/> organise yours? It doesn't have to fit J29 103 exactly - you should choose the nearest one", very few J29 104 respondents are inclined to say "none" (apparently zero per J29 105 cent in the SCELI survey, see Vogler, 1989:46, Table 1). But the J29 106 difficulty is not automatically overcome by resorting to depth J29 107 interviewing since it is likely that the researcher's eventual J29 108 categorisation or description of responses will correspond to J29 109 her/his existing conception of what such a categorisation should J29 110 look like - and what it should look like is of course something J29 111 like an already existing classification. This is the nature of any J29 112 conceptual system - it simplifies and organises, but it constrains J29 113 the possibilities of thinking beyond it.

J29 114 The second major problem, relating in particular to J29 115 longitudinal analysis, is that a classification into types is J29 116 inherently static in its form; people may of course move between J29 117 boxes, but the boxes themselves represent states rather than J29 118 processes. While for the purposes of analysis it may be necessary J29 119 to identify 'types', just as it is necessary to identify events or J29 120 outcomes, we must remember that we may be applying relatively J29 121 arbitrary delineations upon what are essentially continua. J29 122 Following the same line of argument, a movement from one box of a J29 123 classification to another for an individual, or even an entire J29 124 change of classificatory typology, does not constitute an J29 125 explanation. Too often, in quantitative research particularly, the J29 126 development of a new classification or the identification of J29 127 movement between states is taken as the end of analysis. The J29 128 development of a classification and the demonstration of its J29 129 usefulness is frequently not perceived in terms of the necessity to J29 130 provide a theoretical justification - the identification of ordered J29 131 associations is regarded as sufficient. Beyond this solid ground J29 132 lies the dangerous and unstable terrain of the explanation of human J29 133 action - the realm of interpretavism<&|>sic!. Here the familiar J29 134 language of science - of objectivity, of rationality - must be J29 135 replaced by the shifting undercurrents of far less tangible J29 136 discourses involving 'meanings' and 'motivations'. But it is this J29 137 realm which must be entered if an attempt is to be made at J29 138 understanding change. Any simple identification of movement between J29 139 the categories of a classification is bound to fail at the level of J29 140 explanation unless some understanding of the processes which J29 141 underlie observed outcomes is arrived at, and these processes J29 142 themselves can only be fully understood with reference to actor's J29 143 accounts.

J29 144 Couples, families or households?

J29 145 These general considerations lead us to two issues of direct J29 146 relevance to the BHPS. Both relate to problems of definition. J29 147 Firstly there is the question of how one sets the analytical J29 148 boundaries of intra-household allocations. It is clear that J29 149 households do not exist in isolation. Individuals within households J29 150 have relationships with individuals in other households and J29 151 transfer resources of time, money or care which may materially J29 152 affect how those households arrange their internal affairs. For J29 153 example we have found that it is common for women with young J29 154 children who return to work part-time to rely on their mothers to J29 155 provide free childcare. The availability of free childcare was seen J29 156 by many women as an important factor influencing their decision to J29 157 return to work. As Brannen and Moss have observed, women engage in J29 158 a "mental accounting procedure of whether 'it's worth my J29 159 while working'" (1987:86) where the cost of childcare is J29 160 seen as a direct deduction from their own earnings. As one woman J29 161 told us:

J29 162 "It's funny you know, because I pay the child-minding J29 163 fees and I think that I pay it because I feel that it's, the child J29 164 or children are really my responsibility. If I go out to work then J29 165 it should come out of my salary to pay someone else to look after J29 166 them. I wouldn't dream of asking him to pay the childminding J29 167 fee."

J29 168 The fact that women are able to go out to earn is, however, of J29 169 considerable benefit to all household members, particularly in view J29 170 of the gendered nature of spending patterns in which women's J29 171 earnings are more likely to be spent on items of communal rather J29 172 than purely personal benefit even where women's earnings are J29 173 irregular or part-time (see Morris, 1989). Qualitative data is able J29 174 to highlight the extent of inter-household transfers and wider J29 175 support networks or relationships which may be vital to J29 176 understanding the constraints and possibilities open to particular J29 177 individuals in a given household. While these remain theoretically J29 178 distinct from purely intra-household allocations it is clear that J29 179 any analysis of the internal dynamics of the household cannot be J29 180 conducted without reference to inter-household relationships.

J29 181 Secondly, the qualitative research has made us question whether J29 182 it is valid to extend a model developed to categorise financial J29 183 arrangements between married or cohabiting couples to the J29 184 distribution of a variety of resources within households which may J29 185 or may not be made up of married couples or nuclear families.

J29 186 We have already noted the importance of recognising that J29 187 households exist within the framework of broader social J29 188 relationships. In addition however, it is clear that we need to J29 189 conceptualize the household not only in terms of the conjugal or J29 190 'standard' nuclear family household but to take a variety of J29 191 different household compositions with varying cultural backgrounds J29 192 into account. An approach which focuses exclusively on the way in J29 193 which couples allocate resources between them, while being a useful J29 194 exercise in its own right, loses important dimensions which J29 195 influence the internal workings of the household. For example where J29 196 teenage or young adult children are present their involvement or J29 197 contribution to the running of the household, both in terms of J29 198 money and labour within the home, is relevant. It is possible that J29 199 the informal exchange of resources in kind such as free child-care J29 200 and housework may materially affect a household's ability to J29 201 survive financially. Equally important is the fact that households J29 202 may be composed of unrelated individuals or contain an extended J29 203 family of two or more generations, while ethnic minority households J29 204 may operate on very different assumptions to the 'Western European' J29 205 model.

J29 206 The distinction between the household and the family is one J29 207 which is continually blurred in much of the literature in this area. J29 208 J30 1 <#FLOB:J30\>In other words, he has no adequate basis on which to J30 2 determine whether, in consequence of the deviance he refers to, J30 3 there was, or was not, a stronger definition of the moral and J30 4 social boundaries of the community. So far as popular perceptions J30 5 and evaluations are concerned, he is without means of access.

J30 6 Likewise, in treating the second hypothesis, on the constant J30 7 level of deviance, Erikson has to rely on official crime J30 8 statistics, which, for well-known reasons, give only a very J30 9 uncertain indication of the actual level of social deviance, and J30 10 are influenced in their trend by a variety of other factors. J30 11 However, unlike the sociologist of deviance working in contemporary J30 12 society, Erikson cannot investigate in any detail the processes J30 13 through which the official statistics were constituted, nor can he J30 14 collect data of his own which could provide alternative estimates - J30 15 as, say, through some form of 'victim survey'.

J30 16 To be sure, the hypotheses that Erikson addresses are not ones J30 17 that would be easily tested under any circumstances. But, given J30 18 that they derive from a theory that pretends to a very high level J30 19 of generality, there is all the more reason to ask why Erikson J30 20 should impose upon himself the limitations that must follow from J30 21 choosing a historical case. Why should he deny himself the J30 22 possibility of being able to generate his own evidence, to his own J30 23 design, and under conditions in which problems of reliability and J30 24 validity could best be grappled with? Any sociologist, I would J30 25 maintain, who is concerned with a theory that can be tested in J30 26 the present should so test it, in the first place; for it is, in J30 27 all probability, in this way that it can be tested most J30 28 rigorously.

J30 29 I would now like to move on to consider cases where the J30 30 recourse of sociologists to history would appear to have the J30 31 good reasons which, I earlier maintained, should always be present. J30 32 Here my aim is to illustrate what such reasons might be, but also - J30 33 when they are acted upon - the difficulties that may be J30 34 expected.

J30 35 Sociologists, one might think, will most obviously need to turn J30 36 to history where their interests lie in social change. However, it J30 37 should be kept in mind that a recourse to the past - or, that is, J30 38 to the relics thereof - is not the only means through which such J30 39 interests may be pursued: life-course, cohort or panel studies, for J30 40 example, are all ways of studying social change on the basis of J30 41 evidence that is, or has been, collected in the present. J30 42 Sociologists, I would argue, are compelled into historical research J30 43 only where their concern is with social change that is in fact J30 44 historically defined: that is, with change not over some J30 45 analytically specified length of time - such as, say, 'the J30 46 life-cycle' or 'two generations' - but with change over a period of J30 47 past time that has dates (even if not very precise ones) and that J30 48 is related to a particular place. Sociologists have a legitimate, J30 49 and necessary, concern with such historically defined social change J30 50 because, as I have earlier suggested, they wish to know how widely J30 51 over time and space their theories and hypotheses might apply.

J30 52 One illustration of what I have in mind here is provided by J30 53 Michael Anderson's book, Family Structure in Nineteenth J30 54 Century Lancashire. Anderson is concerned with the hypothesis J30 55 that in the process of industrialisation, pre-existing forms of J30 56 'extended' family and kinship relations are disrupted. J30 57 Specifically, he is interested in whether or not this hypothesis J30 58 holds good in the British case - that of the 'first industrial J30 59 nation'. Thus, to pursue this issue, Anderson aims to examine just J30 60 what was happening to kinship relations in Britain at the time J30 61 when, and in the place where, the 'take-off' into industrialism is J30 62 classically located. In contrast, then, with Erikson, Anderson has J30 63 a quite clear rationale for turning to historical research.

J30 64 A second illustration is provided by Gordon Marshall's book, J30 65 Presbyteries and Profits. Marshall is concerned with the J30 66 'Weber thesis' - that a connection exists between the secular ethic J30 67 of ascetic Protestantism and 'the spirit of capitalism'. In the J30 68 long-standing debate on this thesis, the case of Scotland has J30 69 several times been suggested as a critical one, in that, in the J30 70 early modern period, Scotland had a great deal of ascetic J30 71 Protestantism - that is, Calvinism - yet showed little in the way J30 72 of capitalist development. Marshall's aim is then to re-examine the J30 73 Scottish case for the period from around 1560 down to the Act of J30 74 Union of 1707. Marshall points out that Weber himself always J30 75 emphasised that his argument on the role of the Protestant ethic in J30 76 the emergence of modern capitalism was intended to apply only J30 77 to the early stages of this process: once a predominantly J30 78 capitalist economy was established, its own exigencies - in the J30 79 workplace and market - would themselves compel behaviour generally J30 80 consistent with the 'spirit of capitalism' without need of help J30 81 from religion. Again, then, Marshall, like Anderson, has obviously J30 82 good grounds for his recourse to history.

J30 83 Now before proceeding further, I should make it clear that I J30 84 have the highest regard for the two studies to which I have just J30 85 referred. Both make signal contributions to the questions they J30 86 address; and, for me, they stand as leading examples of how in fact J30 87 historical sociology should be conceived and conducted. I say this J30 88 because I want now to go on to emphasise the severe limitations to J30 89 which the analyses of both authors are subject: not because of J30 90 their deficiencies as sociologists, but simply because of the fact J30 91 that they were forced into using historical evidence - forced into J30 92 a reliance on relics - rather than being able to generate their own J30 93 evidence within a contemporary society.

J30 94 The relics on which Anderson chiefly relies are the original J30 95 enumerator's books for the censuses of 1841, 1851 and 1861. On this J30 96 basis, he can reconstruct household composition according to age, J30 97 sex and kinship relations, and he can also to some extent examine J30 98 the residential propinquity of kin. But this still leaves him a J30 99 long way short of adequate evidence on the part actually played by J30 100 kinship in the lives of the people he is studying and on the J30 101 meanings of kinship for them. He attempts to fill out the J30 102 essentially demographic data that he has from the enumerators' J30 103 books by material from contemporary accounts. But these would, I J30 104 fear, have at best to be categorised as 'casual empiricism' and at J30 105 worst as local gossip or travellers' tales. Titles such as J30 106 Walks in South Lancashire and on its Borders, A Visit to J30 107 Lancashire in December 1862, and Lancashire Sketches J30 108 give the flavour.

J30 109 Anderson is in fact entirely frank about the problem he faces. J30 110 "It must of course be stressed", he writes, J30 111 "that just because interaction with kin occurred it is no J30 112 necessary indication that kinship was important. The real test, J30 113 which is quite impossible in any precise way in historical work, J30 114 would be to examine the extent to which kinship was given J30 115 preference over other relational contacts (and the reasons for this J30 116 preference), and the extent to which contacts with kin fulfilled J30 117 functions which were not adequately met if kin did not provide J30 118 them".

J30 119 The point I want to make here would perhaps best be brought out J30 120 if one were to compare Anderson's study of kinship with one carried J30 121 out in contemporary society - let us say, for example, Claud J30 122 Fischer's study of kinship and of other 'primary' relations in J30 123 present-day San Francisco, To Dwell Among Friends. The J30 124 only conclusion could be that the latter is greatly superior in the J30 125 range and quality of data on which it draws, and in turn the rigour J30 126 and refinement of the analyses it can offer. And this point is, of J30 127 course, not that Fischer is a better sociologist than Anderson but J30 128 that he has an enormous advantage over Anderson in being able to J30 129 generate his own data rather than having to rely on whatever relics J30 130 might happen to be extant.

J30 131 Turning to Marshall, one finds that he has problems essentially J30 132 the same as those of Anderson. One of Marshall's main concerns is J30 133 that Weber's position should be correctly understood - following J30 134 the vulgarisations of Robertson, Tawney, Samuelson and other J30 135 critics; and in this respect Marshall makes two main points. First, J30 136 Weber was not so much concerned with official Calvinist doctrine on J30 137 economic activity as with the consequences of being a J30 138 believing Calvinist for the individual's conduct of everyday life - J30 139 consequences which the individual might not even fully realise. In J30 140 other words, Weber's thesis was ultimately not about theology but J30 141 subculture and psychology. Secondly, Weber's argument was that the J30 142 Protestant ethic was a necessary, but not a sufficient cause of the J30 143 emergence of modern capitalism; there were necessary 'material' J30 144 factors also - such as access to physical resources and to markets, J30 145 the availability of capital and credit etc.

J30 146 Thus, Marshall argues, in evaluating the Weber thesis, it is J30 147 not enough to look simply for some overt association between J30 148 theology, on the one hand, and the development of capitalist J30 149 enterprise on the other. What is required is more subtle. It is J30 150 evidence that believing Calvinists, on account of their acceptance J30 151 of a Calvinist world-view, were distinctively oriented to work in a J30 152 regular, disciplined way, to pursue economic gain rationally, and J30 153 to accumulate rather than to consume extravagantly - so that, J30 154 if other conditions were also met, capitalist enterprise J30 155 would then flourish.

J30 156 Marshall's position here is, I believe, entirely sound. But it J30 157 leads him to problems of evidence that he can in fact never J30 158 satisfactorily overcome - despite his diligence in searching out J30 159 new sources and his ingenuity is using known ones. And the basic J30 160 difficulty is that relics from which inferences can systematically J30 161 be made about the orientations to work and to money of early modern J30 162 Scots are very few and far between.

J30 163 In other words, what is crucially lacking - just as it was J30 164 lacking for Anderson and indeed for Erikson - is material from J30 165 which inferences might be made, with some assurance of J30 166 representativeness, about the patterns of social action J30 167 that are of interest within particular collectivities. As Clubb has J30 168 observed, the data from which historians work only rarely allow J30 169 access to the subjective orientations of actors en masse, J30 170 and inferences made in this respect from actual behaviour tend J30 171 always to be question-begging. And Marshall, it should be said, J30 172 like Anderson, sees the difficulty clearly enough. He acknowledges J30 173 that it may well be that "the kind of data required in J30 174 order to establish the ethos in which seventeenth-century Scottish J30 175 business enterprises were run simply does not exist " - or, J30 176 at least, not in sufficient quantity to allow one to test J30 177 empirically whether Calvinism did indeed have the effect on mundane J30 178 conduct that Weber ascribed to it.

J30 179 III J30 180 Let me at this point recapitulate. I have argued that history J30 181 and sociology differ perhaps most consequentially in the nature of J30 182 the evidence on which they rely, and that this difference has major J30 183 implications for the use of history in sociology. I have presented J30 184 a case of what, from this standpoint, must be seen as a perverse J30 185 recourse to history on the part of a sociologist; and I have now J30 186 discussed two further cases where, in contrast, such a recourse was J30 187 justifiable, indeed necessary, given the issues addressed, but J30 188 where, none the less, serious difficulties arise because of the J30 189 inadequacy of the relics as a basis for treating these issues.

J30 190 To end with, however, I would like to move on from these J30 191 instances of sociologists resorting to history in the pursuit of J30 192 quite specific problems to consider - with my initial argument J30 193 still in mind - a whole genre of sociology which is in fact J30 194 dependent upon history in its very conception. I refer J30 195 here to a kind of historical sociology clearly different to that J30 196 represented by the work of Anderson or Marshall, and which has two J30 197 main distinguishing features. First, it resorts to history because J30 198 it addresses very large themes, which typically involve the tracing J30 199 out of long-term 'developmental' processes or patterns or the J30 200 making of comparisons across a wide range of historical societies J30 201 or even civilisations. J30 202 J31 1 <#FLOB:J31\>Regional Inequalities in Infant Mortality in J31 2 Britain, 1861-1971: Pattern and Hypotheses

J31 3 C.H. LEE

J31 4 Variations in health inequality, and the debate about whether J31 5 they have increased or decreased during recent decades, have J31 6 generated considerable investigative activity and querulous J31 7 discussion. A considerable part of this literature has been J31 8 focussed upon health inequalities between social classes, but J31 9 debate has foundered on difficulties in defining and measuring J31 10 acceptable categories and ensuring consistency over time. Illsley J31 11 has argued that differences between death rates of different J31 12 classes are primarily a reflection of the classificatory scheme J31 13 itself and bear little resemblance to actual changes in health or J31 14 death inequality. Others have averred that such reservations about J31 15 the use of class data based on occupational groups are exaggerated J31 16 and, consequently, have felt able to reiterate their conclusions J31 17 derived from this source. It seems unlikely that this avenue of J31 18 investigation can be developed so as to reconcile these various J31 19 positions. The data on occupations and the class categorisations J31 20 derived from them do not admit further refinement, certainly not J31 21 sufficient to persuade those who, like the present writer, regard J31 22 this material as inherently unsatisfactory for meaningful J31 23 statistical analysis.

J31 24 In this paper the problem of health inequality is approached J31 25 from a different perspective, that of regional variations. This has J31 26 two advantages. First, the data are available for a long period of J31 27 time, and thus allow analysis from the 1860s to the present, for J31 28 the constituent counties of Great Britain which remained J31 29 substantially the same throughout the period. The principal change J31 30 in regional categorisation occurred during the 1970s, a fact which J31 31 led to the decision to conclude this investigation at the beginning J31 32 of that decade. By then, the principal features of the most recent J31 33 phase of historical development were clearly established. Even the J31 34 fact that such county-based regionalisation does not allow division J31 35 between rural and urban areas can be tolerated in view of the J31 36 compensating advantage this classification allows in long-term J31 37 internal consistency of the data. Failure to achieve such J31 38 consistency has been, of course, one of the principal criticisms J31 39 levelled against analysis based on class determined by occupational J31 40 groupings. The second justification for adopting a regional J31 41 perspective lies in the fact that this aspect of health inequality J31 42 has been little explored, although it has occasioned several J31 43 general, and usually inaccurate, interpretative observations.

J31 44 The measure of health inequality adopted here is the infant J31 45 mortality rate. Infant mortality has been widely accepted as an J31 46 important and significant indicator of health achievement, because J31 47 infancy has always been one of the most vulnerable periods of human J31 48 life, and because the scale of infant mortality has important J31 49 consequences. Furthermore, infant mortality has been shown to be J31 50 associated, in various studies, with a number of important economic J31 51 and social indicators, such as income per head, equality of income J31 52 distribution, and material deprivation. With regard to the latter, J31 53 it was suggested in the Black Report "that any factors J31 54 which increase the parental capacity to provide adequate care for J31 55 an infant will, when present, increase the chance of survival, J31 56 while their absence will increase the risk of premature death. The J31 57 most obvious such factors fall within the sphere of material J31 58 resources: sufficient household income, a safe, uncrowded and J31 59 unpolluted home, warmth and hygiene, a means of rapid communication J31 60 with the outside world, for example a telephone or car, and an J31 61 adequate level of manpower - or womanpower (two parents would J31 62 normally provide more continuous care and protection than J31 63 one)". Variations between infant mortality rates in J31 64 different regions and in the long term provide, therefore, a J31 65 significant indicator of variations in basic economic and social J31 66 well-being.

J31 67 The long-run pattern of infant mortality in the advanced J31 68 industrial nations is generally familiar, the fluctuating levels J31 69 which obtained through the second half of the nineteenth century J31 70 gave way to widespread and rapid decline through the twentieth J31 71 century. In the case of Britain, much attention has been paid by J31 72 historians to the sharp drop in the infant mortality rate at the J31 73 beginning of the present century and considerable effort has been J31 74 expended in attempts to explain this phenomenon. One of the most J31 75 popular theses of recent years has been Beaver's argument that J31 76 improvements in the milk supply reduced gastro-enteritis after the J31 77 turn of the century, a development which was happily co-incident J31 78 with the emergence of welfare services which emphasised safe J31 79 feeding. Purification of the water supply, sewage disposal, safer J31 80 milk, and greater food hygiene all contributed to reduce the risk J31 81 of infection from food and drink. Beaver's important contribution J31 82 reinforced attention on the downturn in the national infant J31 83 mortality rate at the turn of the century as the key to the J31 84 long-term pattern of change.

J31 85 In Table 1 infant mortality rates are shown at ten-year J31 86 intervals from 1861 to 1971 for each of the 55 regions of Great J31 87 Britain. The composition of these regions which comprise English J31 88 counties, and aggregations of Welsh and Scottish counties, is J31 89 indicated in Table 2. These regional infant mortality rates vary J31 90 considerably more than might be expected from the national J31 91 aggregates for England and Wales and for Scotland, which fell J31 92 rather dramatically after 1901. Furthermore, these regional rates J31 93 do not support the popular notion that trends in infant mortality J31 94 change were "remarkably consistent" throughout the J31 95 country. Nor do these estimates sustain the view advanced in the J31 96 Black Report that "In the middle of the nineteenth century, J31 97 the south-east of England recorded comparatively high rates of J31 98 death, while other regions like Wales and the far north had a J31 99 rather healthier profile". While London just fell into the J31 100 group of 17 regions with an infant mortality rate of 150 or above J31 101 in 1861, the majority of such disadvantaged areas were the J31 102 industrialising counties of the English midlands and north. J31 103 Lancashire, the East Riding, the West Riding, Staffordshire and J31 104 Leicestershire were among regions with the highest rates. At the J31 105 opposite extreme, in 12 regions infant mortality was 120 or less. J31 106 Half of these were located in the south and west of England, while J31 107 the rest were in Scotland where in Strathclyde North, Dumfries and J31 108 Galloway, Grampian and Highland very low mortality rates between 86 J31 109 and 95 were recorded.

J31 110 More significant, and contrary to conventional wisdom, is the J31 111 pattern of change in these regional infant mortality rates. They J31 112 did not fall uniformly, nor was there a clearly marked hiatus J31 113 around the turn of the century. In fact, several groups followed J31 114 different and divergent patterns of change during the second half J31 115 of the nineteenth century. In some regions the highest mortality J31 116 rate was recorded in 1861, and there was a continuous improvement J31 117 thereafter. Buckinghamshire, Bedfordshire, Cambridgeshire, J31 118 Wiltshire and Dorset comprised this favoured group. In other J31 119 regions mortality peaked in 1871 followed by continuous improvement J31 120 thereafter, namely Hertfordshire, Northamptonshire, Rutland and J31 121 Strathclyde South. In many other regions the peak infant mortality J31 122 rate occurred in either 1861 or 1871 and was followed by a fall, J31 123 and a modest increase in 1891 or 1901 as a temporary setback. In no J31 124 fewer than 40 of the 55 regions was the highest infant mortality J31 125 rate registered in 1861 and 1871, a group which together accounted J31 126 for 80.0 per cent of the national population at the latter date. J31 127 Furthermore, many of the regions with an early peak in infant J31 128 mortality were those with the lowest rates. This was especially J31 129 pertinent in the south and west of England where only in Hampshire, J31 130 Essex and Devonshire was a peak reached after 1871. Even in some of J31 131 the industrial regions with the highest infant mortality rates, J31 132 like Lancashire, the West Riding and Durham the highest rates were J31 133 recorded in 1871.

J31 134 tables&caption

J31 135 In those regions which did not fall into the large group in J31 136 which there was a peak in 1861 or 1871 the highest rates were found J31 137 in either 1891 or 1901. After the turn of the century, except for a J31 138 small number of occasional and temporary increases, all regional J31 139 infant mortality rates fell continuously. There is no doubt that J31 140 after 1901 the downturn was both universal and substantial. The J31 141 most obvious group of regions in which infant mortality rates J31 142 increased, rather than decreased during the later decades of the J31 143 nineteenth century was found in Scotland. In seven of the ten J31 144 Scottish regions a peak infant mortality rate was found in 1891 or J31 145 1901, in some of them rates rose during the later decades of the J31 146 century, substantially so in the case of Grampian. Elsewhere, in J31 147 Monmouth, South Wales and Essex mortality rates also rose during J31 148 the closing decades of the century, while in Northumberland, J31 149 Cheshire, the North Riding and Worcestershire rates were high, but J31 150 stable.

J31 151 The pattern of change at national aggregate level, confirmed by J31 152 the unweighted average infant mortality rate for the aggregated J31 153 regions shown in Table 3, covers a diversity of regional patterns. J31 154 The aggregate rate suggests fluctuation around a slowly falling J31 155 mean value until 1901, and the improvement around 1881 and the J31 156 setback in 1891 was experienced in most regions. At regional level, J31 157 there were two phases of downturn some 30 years apart, with much of J31 158 Scotland, South Wales and some English counties, like J31 159 Northumberland, the North Riding and Essex, lagging behind the J31 160 rest. With these exceptions, and admitting the temporary setback J31 161 suffered during the 1890s when, it has been suggested, long, hot J31 162 summers produced conditions conducive to a resurgence of diarrhoeal J31 163 complaints, much of Britain experienced a downturn in infant J31 164 mortality rates from the 1860s and 1870s. From the turn of the J31 165 century, decline was universal and persistent.

J31 166 The long-term pattern of change does not, of course, indicate J31 167 increase or decrease in the variation of infant mortality rates J31 168 between different regions. Estimates of four different measures of J31 169 inequality are shown in Table 3. Each has its own peculiar J31 170 strengths and limitations. The coefficient of variation is J31 171 sensitive to change throughout a given distribution, while the J31 172 variance of logarithms is sensitive to changes at the lower ranges J31 173 of a scale, and the Gini coefficient is particularly responsive to J31 174 transfers affecting the middle values of a distribution. Theil's J31 175 entropy index is responsive to changes throughout the distribution J31 176 and measures deviations from a state of equality in which each J31 177 variable in the distribution has a share equivalent to its relative J31 178 size. Since it measures inequality exclusively in terms of the J31 179 'distance' between variables, it satisfies, unlike the other J31 180 measures quoted, the "strong principle of J31 181 transfers". But, most importantly, all four measures are J31 182 invariant if all values in a set are raised or lowered in the same J31 183 proportion, so that each measure is an appropriate indicator of J31 184 inequality in infant mortality rates. As can be seen from Table 3, J31 185 with minor exceptions, all the four measures of inequality show the J31 186 same pattern of change over the century examined.

J31 187 table&caption

J31 188 Three of the four measures show increasing inequality between J31 189 1861 and 1871, as the downturn in infant mortality rates in J31 190 healthier regions was not replicated elsewhere. All measures J31 191 converge during the following decade, 1871-1881, as the decline in J31 192 the mortality rate became widespread. Thereafter, with the single J31 193 exception of the Gini coefficient for 1911, all the indicators show J31 194 an increase in inequality until 1921 or 1931. Even the downturn in J31 195 the mean infant mortality rate after 1891, and in all the regional J31 196 rates after 1901, did not prevent this increase. The reason for J31 197 this increase in inequality is quite clear; the decline in infant J31 198 mortality rates was fastest in those regions in which the rates had J31 199 been lowest at mid-century. Thus, the rate of decline between J31 200 1881-1921 was only 0.50 per cent per year in Lanarkshire, and 0.80 J31 201 per cent in Durham, compared to 1.48 per cent in Surrey and 1.25 J31 202 per cent in Wiltshire. From the peak of 1921 and 1931, when J31 203 measured inequality was almost the same, there was a continuous J31 204 convergence of regional infant mortality rates as inequality J31 205 decreased. This trend was reflected in each of the four measures. J31 206 By 1961, two measures indicated a level of inequality similar to J31 207 that in 1861, while Gini's index suggested that it was greater and J31 208 Theil's that it was smaller. By 1971, all four indicators showed J31 209 the lowest level of inequality recorded during the entire J31 210 period.

J31 211 J32 1 <#FLOB:J32\>Theoretical linguistics, like theoretical physics, J32 2 theoretical chemistry or theoretical biology, is, of itself, J32 3 non-empirical. It is free to create its own theoretical constructs J32 4 as it will. But, also like these other theoretical sciences to J32 5 which I have just referred, it originates with the observation and J32 6 systematization of identifiable phenomena which appear J32 7 pre-theoretically to have something in common. In so far as it J32 8 retains its internal coherence and distinctive identity - in so far J32 9 as theoretical linguistics is to be distinguished from the J32 10 theoretical branches of other sciences - it maintains, and must J32 11 maintain, its connexion with what is pre-theoretically J32 12 identifiable, across all societies and cultures, as the referent of J32 13 Saussure's 'langage'.

J32 14 Theoretical linguistics - more precisely, theoretical general J32 15 linguistics - is that branch of the subject which sets out to J32 16 provide a non-trivial, intellectually satisfying, answer to what I J32 17 referred to earlier as the central defining question of J32 18 linguistics: 'What is language?' (construed, as we shall see below, J32 19 in a particular way).

J32 20 Although this question contains the ontological presupposition J32 21 that there is such a thing as language ('langage'), of J32 22 itself it says nothing about its ontological status. It does not J32 23 necessarily imply that language is empirically and J32 24 pre-theoretically separable from non-language. Both general and J32 25 descriptive linguistics have always operated, however, with the J32 26 assumption that this is so. To quote W.S. Allen, on this point: J32 27 "We presume that there is a particular mode of human J32 28 behaviour which it is legitimate to isolate and to label as J32 29 'language'; we assume also that this behaviour is such that J32 30 systematic statements may be made about its various J32 31 manifestations" (1957b: 13). Once again, if we wish to be J32 32 precise, we need to be clear about the distinction between process J32 33 and product, and consequently about the different ways in which J32 34 language 'manifests' itself to us in the physical world. Most J32 35 branches of linguistics draw their data from the products of the J32 36 process, not from the process itself (various kinds of muscular and J32 37 neurophysiological activity). This is an important point which has J32 38 been dealt with above (see Chapter 2): I will not elaborate upon it J32 39 further. For present purposes, let us simply note that the two J32 40 assumptions, or postulates, made explicit in the passage just J32 41 quoted - the postulate of isolability and the postulate of J32 42 systematicity - have proved their worth over the centuries (the J32 43 history of linguistics, in some of its branches at least, and of J32 44 linguistic theory is measured in centuries) and need not be J32 45 justified in detail here. Nor is there any need to labour the point J32 46 that the isolability of the "particular mode of human J32 47 behaviour" that is pre-theoretically identifiable as J32 48 language ('langage') rests, operationally, upon the J32 49 relatively clear, empirically determinate and theory-neutral, J32 50 difference between speech and non-speech.

J32 51 The question 'What is language?' can be addressed from several J32 52 points of view and can be answered in several different, but J32 53 equally legitimate, ways according to the point of view that is J32 54 adopted. Theoretical linguistics, founded upon the Saussurean and J32 55 post-Saussurean trichotomy of 'langage', J32 56 'langue' and 'parole', interprets the question J32 57 as meaning 'What is a language?' ('Qu'est-ce qu'une J32 58 langue?'). The different branches of theoretical J32 59 linguistics adopt characteristically different points of view and J32 60 consequently postulate different kinds of theoretical constructs in J32 61 the answers they give to the question. Theoretical microlinguistics J32 62 (often called autonomous linguistics: see Chapter 2) adopts the J32 63 point of view expressed by Saussure, or rather his editors, in the J32 64 famous final sentence of the Cours: "la J32 65 linguistique a pour unique et v<*_>e-acute<*/>ritable objet la J32 66 langue envisag<*_>e-acute<*/>e en elle-m<*_>e-circ<*/>me et pour J32 67 elle-m<*_>e-circ<*/>me" (1916: 317). It is the J32 68 controversial "en elle-m<*_>e-circ<*/>me et pour J32 69 elle-m<*_>e-circ<*/>me", of course, which J32 70 distinguishes theoretical microlinguistics from the various J32 71 sub-branches of theoretical macrolinguistics. But they too, as I J32 72 shall argue, have their own distinctive conceptions of the J32 73 language-system. It must not be thought that sociolinguistics or J32 74 psycholinguistics, or the other branches of macrolinguistics, can J32 75 dispense with the distinction between the system and the process J32 76 (or its products). I will come back to this point. But first let me J32 77 make explicit something else which is not immediately obvious.

J32 78 This is the fact that 'la langue' in the J32 79 famous passage from the Cours that I have just quoted can be J32 80 interpreted either generically or specifically. Its generic J32 81 interpretation defines the field of theoretical general J32 82 linguistics; its specific interpretation, that of theoretical J32 83 descriptive linguistics. This is the gloss that I would add to J32 84 Katz's formulation of the goals of what he calls linguistics J32 85 tout court (I would call it theoretical linguistics) in J32 86 his recent defence of Platonic realism (as an alternative to both J32 87 American structuralism and Chomskyan cognitivism): J32 88 "linguistics tries to construct theories to answer the J32 89 questions, first, 'What is English, Urdu, and other natural J32 90 languages?' and second, 'What is language in general?'" J32 91 (Katz, 1981: 21). Two further terminological comments may be made J32 92 about this passage, by way of exegesis: (i) by "natural J32 93 languages" Katz, like most philosophers and linguists, J32 94 clearly means N-languages; (ii) by "language in J32 95 general" he means, in Saussurean terms, not J32 96 'langage', but 'langue' (construed J32 97 generically). His two questions are in fact post-Chomskyan J32 98 reformulations of Saussure's "la linguistique a J32 99 pour ... objet ..." (Chapter 4). Although J32 100 theoretical general linguistics existed long before Chomsky J32 101 published his seminal work in the mid-1950s, modern theoretical J32 102 descriptive linguistics is very much his creation. A generative J32 103 grammar of any N-language - English, Urdu, etc. - is a theory of J32 104 that language: more specifically, a theory of the well-formedness J32 105 of the sentences of the language.

J32 106 So far, I have been concerned, first of all, to point out that, J32 107 although until recently there was no need to distinguish between J32 108 'general linguistics' and 'theoretical linguistics', nowadays there J32 109 is; and, second, to prepare the ground for the distinction that I J32 110 am drawing between theoretical linguistics and linguistic theory J32 111 and for the necessarily brief presentation of my own approach to J32 112 the definition of the field of theoretical linguistics, on the J32 113 basis of alternative, equally legitimate, conceptions of J32 114 language-systems.

J32 115 But why, it may now be asked, is it not possible to operate J32 116 with a single notion of the language-system valid in all branches J32 117 of linguistics, micro- and macro-, theoretical and non-theoretical? J32 118 This is a question that has been dealt with in some detail in the J32 119 preceding chapter. The answer, as we have seen, derives partly from J32 120 the apparently sui generis properties of N-languages and J32 121 partly from the complexity and heterogeneity of the J32 122 pre-theoretically isolable phenomena identifiable as J32 123 'language' (i.e., as Saussure's 'langage').

J32 124 The multiplicity and heterogeneity of the connections that can J32 125 be established between what are pre-theoretically classifiable as J32 126 language-data ('des donn<*_>e-acute<*/>es J32 127 langagi<*_>e-grave<*/>res', if I may employ this useful J32 128 post-Saussurean adjective) and other data, natural and cultural, J32 129 constituting the subject matter of other disciplines are such that, J32 130 in my view at least, there is no immediate possibility, perhaps J32 131 even no ultimate possibility, of constructing a unified theory of J32 132 the natural and social sciences within which a unitary theory of J32 133 language (of 'langue' construed generically) would find J32 134 its place and be descriptively and explanatorily adequate to the J32 135 data that it systematizes and accounts for. As to the apparently J32 136 unique, or sui generis, character of what are commonly J32 137 referred to as natural languages, this may well have been J32 138 exaggerated at times by proponents of so-called autonomous J32 139 linguistics. The discontinuity between language and non-language, J32 140 on the one hand, and the determinacy, arbitrariness and closedness J32 141 of grammatical structure, on the other, have certainly been greatly J32 142 exaggerated by linguists of various schools, generativist and J32 143 non-generativist. The fact remains that nothing remotely resembling J32 144 a comprehensive, intellectually interesting and empirically J32 145 satisfactory account of the grammatical structure of N-languages in J32 146 terms of the theoretical concepts and explanatory principles of J32 147 other disciplines has yet been provided by any of those who have J32 148 challenged the sui generis character of languages. There J32 149 is every reason therefore to continue to subscribe to a working J32 150 hypothesis that has proved its heuristic value in the practical J32 151 description of languages over the centuries and has been, more J32 152 recently, the foundation-stone of what is so far the most J32 153 sophisticated branch of theoretical linguistics, both general and J32 154 descriptive: theoretical (synchronic) microlinguistics. This does J32 155 not mean, however, that we should, as practitioners of J32 156 microlinguistics, whether general or descriptive, close our minds J32 157 to those aspects of language that are not, or do not appear to be, J32 158 sui generis or deny the validity of alternative views of J32 159 the nature and ontological status of language-systems.

J32 160 The ontological status of the language-system (Saussure's J32 161 'langue') has been controversial ever since the J32 162 publication of the Cours. Saussure's own views are unclear and J32 163 perhaps contradictory. At one time, he says that they are J32 164 supra-individual social facts; at another time, that they are J32 165 stored in the brains of individual members of the J32 166 language-community (1916: 23-32). And each of these conflicting J32 167 views is incompatible with the view, recently advocated by Katz J32 168 (1981), that language-systems are purely abstract, mathematical J32 169 (so-called Platonic) objects (see Chapter 4). As will be obvious J32 170 from what has been said earlier, Katz's view (which is close to J32 171 Hjelmslev's, 1943a), is the one that I accept for microlinguistics, J32 172 though not for psycholinguistics, sociolinguistics or other J32 173 branches of macrolinguistics.

J32 174 Some part of the confusion and controversy that has surrounded J32 175 the Saussurean distinction of 'langue' and J32 176 'parole' over the last half-century or so is to be J32 177 attributed to the fact that both words are used in the Cours J32 178 non-technically (i.e., pre-theoretically) as well as J32 179 technically; and, since the theoretical distinction (or, as we J32 180 shall see, distinctions) that Saussure draws between them J32 181 correlates with differences of meaning in everyday French, it is J32 182 not always clear in what sense they are being employed in J32 183 particular contexts. It must also be admitted that Saussure's own J32 184 comments (or those of his editors) about the rough equivalence J32 185 between French 'parole' and German 'Rede' (and J32 186 Latin 'sermo' in contrast with 'lingua') are J32 187 less than helpful (1916: 31). They must have encouraged, even for J32 188 those who have read the Cours in French, what has undoubtedly J32 189 been, over the years, by far the most serious misunderstanding of J32 190 the technical distinction between 'langue' and J32 191 'parole': the view that it relates basically, or J32 192 primarily, to the distinction between language and speech. It does J32 193 indeed cover one dimension or one part of the semantic difference J32 194 between 'language' and 'speech' (between German 'Sprache' J32 195 and 'Rede', between Russian 'jazyk' and J32 196 'recj', etc.): or rather, to be more precise, between J32 197 'language' construed as a count noun and 'speech' understood as J32 198 referring to the product, rather than the process, of speaking. But J32 199 it does so, as we have seen, only secondarily. The primary J32 200 distinction is between a language and utterances (spoken, written, J32 201 or whatever: i.e., products, not processes, inscribed in some J32 202 appropriate physical substance or medium) which, by virtue of their J32 203 structure (and independently of their physical manifestation), are J32 204 identifiable as utterances of the language in question (see Chapter J32 205 1). It is unfortunate that the beginnings of theoretical J32 206 linguistics should have coincided, for historically explicable J32 207 reasons, with a period of extreme phonocentrism. But no more needs J32 208 to be said on that score.

J32 209 Much of the controversy, if not confusion, that still attaches J32 210 to the Saussurean, or post-Saussurean, distinction between J32 211 'langue' and 'parole' (or the Chomskyan J32 212 distinction of 'competence' and 'performance', which is valid for J32 213 psycholinguistics but not for microlinguistics) must, however, be J32 214 attributed to what in this and the previous chapter I have J32 215 characterized as a false assumption: the assumption that there is J32 216 only one kind of reality and that so-called natural languages, J32 217 N-languages, must be either psychological or social entities, or, J32 218 in terms of an alternative dichotomy, that they must be either J32 219 physical or non-physical. It is my contention that J32 220 microlinguistics, on the one hand, and the several branches of J32 221 macro-linguistics, on the other, start from the same J32 222 pre-theoretical notion of N-languages and that, according to their J32 223 own viewpoint and the alliances that they forge with other J32 224 disciplines (mathematics, psychology, sociology, anthropology, J32 225 etc), they each practise a particular kind of abstraction and J32 226 idealization in the construction of the ontologically appropriate J32 227 model of the underlying language-system.

J32 228 I have already referred to one common misunderstanding of J32 229 Saussure's terminological distinction between 'langue' J32 230 and 'parole': a misunderstanding based on the view that J32 231 it correlates directly with the distinction between language and J32 232 speech. J32 233 J33 1 <#FLOB:J33\>Knowledge 3: acknowledging and sharing J33 2 the courtesies of the game, such as how long to wait between moves, J33 3 whether or not to talk, move about, how seriously to take it, what J33 4 to do if a piece gets lost, whether or not to penalise one another J33 5 if a move is retracted. And so on.

J33 6 Knowledge 4: how to play with skill. In chess, as in J33 7 all games many of those who play don't play very well, hence the J33 8 prevarications: 'Yes I play but not very well' or 'I don't really J33 9 play' or 'I know how to play'.

J33 10 Knowledge 2 may be less clear since the pieces may be quite new J33 11 to the claimant. However, since s/he knows what sorts of pieces to J33 12 expect (two bishops, eight pawns and so on), even if the pieces in J33 13 this new set are distinguished only by colours s/he will very J33 14 quickly identify them for what they represent, and will also be J33 15 able to negotiate with the inviting player who may have begun with J33 16 a different piece identification. That doesn't matter as long as J33 17 they agree as to which object stands for which piece in the J33 18 game.

J33 19 So much for the rules: without the knowledges indicated above, J33 20 playing a game of chess would not be possible. If the two players J33 21 have different ideas about which moves may be made, in which order J33 22 or which piece stands for what, then there can only be chaos or J33 23 randomness, or of course total and instant victory for one side.

J33 24 5.6 Describing the four kinds of Knowledge

J33 25 Knowledge 1, in spite of what has just been said, is a form of J33 26 convention. Rules may in fact be formalised conventions, whether J33 27 they are the operations of a computer or a motor or the rules of a J33 28 nation state or the rules of a game, but once fixed they cease to J33 29 be conventions (or thought of as conventions) and become law-like. J33 30 Rules may originate in conventions but they must then be elevated J33 31 above conventions in order for the game to proceed at all.

J33 32 Knowledge 2 may appear to be more convention-like than J33 33 rule-like but as we shall see later it is safer to regard J33 34 it as rule-like. Of course, like Knowledge 1 it takes its origin in J33 35 convention, in tacit agreement as to what shape indicates which J33 36 piece: this is the Queen, this the King, Bishop and so on. But J33 37 although at the outset which counts as which is immaterial, as time J33 38 goes on the player forgets that it was only a convention and for J33 39 him/her the connection between the object and the chess piece J33 40 becomes unquestionable. Of course the new player may find this J33 41 uncomfortable (and if s/he doesn't know chess incomprehensible) but J33 42 will readily accept the distribution of object to piece because J33 43 this is what s/he expects.

J33 44 Knowledge 2 like Knowledge 1 draws on a set of routines and J33 45 their combinations in skilled and planning ways, leaving the J33 46 neophyte lost because s/he does not understand what or why it is J33 47 happening, and still puzzling the learning chess player.

J33 48 Knowledge 3 is more obviously conventional and overlaps with J33 49 very local arrangements which can be negotiated separately for each J33 50 chess encounter. No doubt in some cases the types of convention J33 51 listed here (for example how long to wait between moves, talking, J33 52 moving about, whether or not to allow a retraction of a move) are J33 53 more rule-like and may indeed be governed by a rule book just like J33 54 Knowledge 1 and 2. But other aspects of Knowledge 3, for example J33 55 how seriously to take it, whether to bet on a game, how generously J33 56 to interpret the 'rules' and whether to behave sympathetically to J33 57 one's opponent - these features of behaviour must be left to J33 58 individuals. Indeed the problem with Knowledge 3 is that it tends J33 59 to spread over into very personal and quite individual J33 60 <}_><-|>characterstics<+|>characteristics <}/> like how often to J33 61 smile during a game, whether to eat, smoke, drink. These move J33 62 beyond even the negotiated interpersonal and local, becoming wholly J33 63 idiosyncratic.

J33 64 Knowledge 4 is of a different order. It is possible to say, in J33 65 answer to the question we first asked, 'Yes, I play chess'. But J33 66 that does not imply well or badly. Modesty normally requires a J33 67 simple answer without qualification: or rather the unmarked form J33 68 would be 'I play, but not very well', whereas 'Yes I play' could J33 69 mean I play very well indeed. Modesty here is conventional except J33 70 in the Stephen Potter gamesmanship ploys which deliberately J33 71 downplay beyond the reach of modesty, for example 'Do you play J33 72 chess?' - 'Hardly at all', meaning 'Yes I'm a Grandmaster!'

J33 73 But Knowledge 4 is really about level of skill and the point we J33 74 must make is that this is a quite separate attribute from J33 75 Knowledges 1-3: Knowledges 1-3 are required for chess players. J33 76 Knowledge 4 is not required in the same way. Knowledge 4 is of J33 77 course necessary to some modest degree. To take a different sort of J33 78 game, to have only Knowledge 1-3 for tennis is of no value if you J33 79 are invited to join a friend on the tennis court for a game since J33 80 you will not find it possible to translate your knowledges into J33 81 some kind of performance, however low level. Knowing the rules of J33 82 tennis (or of chess) is no guarantee (indeed no assurance of any J33 83 kind) of being able to play the game for real.

J33 84 The game analogy helps in two ways. First, it indicates the J33 85 distinction between performance and competence. Performance (in J33 86 chess, tennis, games, - and language) means putting into action J33 87 Knowledge 1-3, that is playing the game, producing, using the J33 88 language. There are, of course, different levels of performance and J33 89 I will return to these. Second, it indicates that what the J33 90 performance shows (for the moment, again, leaving aside its level J33 91 of skill) is the extent of the informing Knowledges 1-3, whether J33 92 the player or user knows the rules (Knowledge 1), is familiar with J33 93 their representation (Knowledge 2) and observes the interactional J33 94 courtesies (Knowledge 3).

J33 95 What Knowledge 4 of course indicates brings me to the heart of J33 96 the 'how well do you play?' question. First of all let me dispense J33 97 with the explicit fallacy. Just as Knowledges 1-3 may be present J33 98 without Knowledge 4, so that the player or user knows in theory but J33 99 can't in practice, so Knowledge 4 may be present alone, the player J33 100 can play or use but is not able to explain this understanding in J33 101 terms of Knowledges 1-3. In both cases we probably need to suspend J33 102 disbelief and assume, given subjective normality, that the J33 103 unpractised player who has Knowledges 1-3 can through practice J33 104 articulate Knowledges 1-3; and similarly the player or user who has J33 105 Knowledge 4 can acquire Knowledges 1-3.

J33 106 However, this need not be the case both ways. Knowledges 1-3 J33 107 are possible for armchair players who never acquire or articulate J33 108 Knowledge 4.

J33 109 In the reverse case we know that Knowledge 4 is possible with J33 110 no explicit Knowledges 1-3. Of course we assume that Knowledges 1-3 J33 111 must implicitly underlie Knowledge 4, that is that no-one can play J33 112 chess or tennis or another game, without some knowledge of the J33 113 rules, the moves, the conventions and the courtesies. Even more J33 114 interesting in terms of Knowledge 4 is the threshold question. J33 115 Should I assume that Knowledge 4 necessarily requires knowing not J33 116 to play skilfully? No doubt we do make major distinctions J33 117 among players, we provide hierarchies and championships and J33 118 honours, we choose teams and we (probably) distinguish even when J33 119 choosing opponents ourselves. And yet although this is common and J33 120 indicates an important aspect of game knowledge it is not, I J33 121 suggest, necessary. We can all be chess players however badly we J33 122 play, (although there is one caveat which is that we do need to J33 123 have some small acquaintance with Knowledges 1-3.) Knowledge 4 is J33 124 paradoxically less necessary. Some modicum is necessary, but no J33 125 more than a limited amount. Perhaps in addition to the possession J33 126 of Knowledges 1-3 there also needs to be some motivation to develop J33 127 Knowledge 4.

J33 128 The point of the analogy should now be obvious. Knowledges 1-3 J33 129 have to do with competence, Knowledge 4 with a combination of J33 130 performance and proficiency; and what the performance of Knowledge J33 131 4 demonstrates in illustrating levels of proficiency is precisely J33 132 the extent to which Knowledges 1-3 have been internalised.

J33 133 5.7 Knowledges and language

J33 134 Let me now move back from the game analogy to language, J33 135 indicating the parallels of Knowledges 1-4 and, relating in each J33 136 case, the different kinds of knowledge available to the fugitive J33 137 native speaker I am attempting to capture.

J33 138 I must say, first of all, that, as with all analogies, the J33 139 parallels between games and language do not easily hold up. J33 140 Nevertheless, I will make whatever connections are possible and, J33 141 when necessary, point to the discrepancies.

J33 142 Metalinguistic knowledge

J33 143 Knowledge 1 is metalinguistic knowledge, knowledge about the J33 144 language. Native speakers may or may not have this explicit J33 145 knowledge though it is customary to say that they have internalised J33 146 it in some sense. What it means is the ability to talk about the J33 147 language, to know and describe in however elementary a way, the J33 148 parts of the sentence, to have some awareness, which can inform J33 149 discussion, of accent, style, register, linkages in discourse and J33 150 so on. But Knowledge 1 in language refers more importantly to a J33 151 manipulative ability with these structures, to be able to put J33 152 together sounds, intonation, stressing, rhythm, sentences, J33 153 discourses, registers, styles, perhaps within a very limited range J33 154 (especially at above sentence levels). Notice that we are teetering J33 155 here on the very edge of rule-governed behaviour (and are already J33 156 moving into the arena of Knowledge 2 and Knowledge 3).

J33 157 Knowledge 1 involves having the construction ability to J33 158 assemble the parts of common sentence types or texts and to J33 159 recognise them receptively as meaning bearing whether or not they J33 160 are understood. What matters crucially then is a recognition of J33 161 language use as being an exemplification or realisation of the J33 162 structural resource which they do have control over. Of course this J33 163 is a strong argument in support of the centrality of grammar: it J33 164 assumes that all language use is a particular, local or contextual J33 165 adaptation of the grammar. As I will show, the power of the local J33 166 or contextual is not so easily dismissed. But for the moment, to J33 167 use another analogy, it is generally accepted that skills are J33 168 transferable in activities such as reading, (whatever script is J33 169 used, given constancy of code) or driving a car, whatever J33 170 intricacies and developments the car may have, or farming, medicine J33 171 and so on. In all such cases there are constants, the important J33 172 core remains and what changes is how to use that core. Similarly J33 173 with language: the grammar of any one 'language' remains quite (if J33 174 not fully) impermeable to change but what uses it is put to vary, J33 175 with time and demand.

J33 176 Discriminating knowledge

J33 177 Knowledge 2, which I will call discriminating knowledge, J33 178 enables the native speaker to recognise what counts or what does J33 179 not count as being part of the language. There are perhaps three J33 180 aspects to this, none of them foolproof for reasons discussed J33 181 earlier in connection with the L1-L2-FL (Foreign Language) J33 182 relation, that is that any one native speaker is vastly limited in J33 183 what s/he knows of his/her own language, but it does include a J33 184 recognition ability of the rote kind of idiom, metaphor and so J33 185 on.

J33 186 However, given those constraints, I propose these three J33 187 attributes to Knowledge 2. First, the native speaker knows what is J33 188 his/her language and what is not (it's English, say, not French). J33 189 Second, the native speaker knows that a sentence/text/sound could J33 190 be his/her language but it doesn't sound quite right. It belongs J33 191 elsewhere but is not somehow familiar; in other words it must J33 192 belong to some other dialect. Even if in my idiolect: 'term starts J33 193 again on Monday already' is not possible I recognise that it is J33 194 possible in some idiolects of English.

J33 195 Third, the native speaker knows that a new word or expression, J33 196 one that s/he has not heard before or even one that s/he chooses to J33 197 invent 'belongs' to the language. J33 198 J34 1 <#FLOB:J34\>The only factor which might argue against this J34 2 characterization of <*_>oe-ligature<*/>-Tensing as lexical in this J34 3 set of dialects is its contravention of Structure Preservation J34 4 (SP), a constraint which states that "lexical rules may not J34 5 mark features which are non-distinctive, nor create structures J34 6 which do not conform to the basic prosodic template of the J34 7 language" (Borowsky, 1986: 29). SP is therefore intended to J34 8 prohibit the introduction of non-contrastive features into the J34 9 lexical phonology: as Kiparsky (1985: 87) says, "if a J34 10 certain feature is non-distinctive, we shall say that it may not be J34 11 specified in the lexicon. This means that it may not figure in J34 12 non-derived lexical items, nor be introduced by any lexical rule, J34 13 and therefore may not play any role at all in the lexical J34 14 phonology". Since Harris (1989a) assumes J34 15 [<*_>plus-minus<*/>tense] to be non-distinctive for J34 16 English, and since <*_>oe-ligature<*/>-Tensing clearly refers to J34 17 this feature, the rule contravenes SP. However, Harris tentatively J34 18 suggests that newly-lexicalised rules may violate SP J34 19 temporarily, with the reassertion of SP perhaps influencing the J34 20 future direction of change, although he provides no clear evidence J34 21 of this determinative role of SP. In other dialects, J34 22 <*_>oe-ligature<*/>-Tensing has moved even further from its J34 23 Neogrammarian source: in RP, for instance, J34 24 <*_>oe-ligature<*/>-Tensing is no longer a J34 25 <}_><-|>synhronically<+|>synchronically<}/> productive rule, but J34 26 has caused restructuring at the underlying level, such that the J34 27 tense reflex of historical short /<*_>ae-ligature<*/>/ has merged J34 28 with /a:/ (> /<*_>unch<*/>/) from other sources, including earlier J34 29 /a:r/, in path, laugh and so on.

J34 30 Harris' discussion of <*_>oe-ligature<*/>-Tensing suggests that J34 31 sound changes may be phonologized as postlexical rules, but may J34 32 subsequently acquire properties of the lexical syndrome in (2) and J34 33 become (initially non-Structure Preserving) lexical rules, which J34 34 may then also begin to diffuse. Ultimately, the number of lexical J34 35 exceptions may increase, and the rule will be lost, its effects J34 36 being incorporated into the underlying representations.

J34 37 If these suggestions are substantiable, LP gains considerably J34 38 in a number of domains. Labov's two types of sound change can be J34 39 matched with credible synchronic counterparts, and his notion of J34 40 more and less abstract changes linked with the lexical-postlexical J34 41 division, although the case of <*_>oe-ligature<*/>-Tensing shows J34 42 that Labov's correlation of particular features with each type of J34 43 change cannot be maintained. Furthermore, LP can be seen to be free J34 44 of another shortcoming of standard generative phonology, in which J34 45 the incorporation of change into the synchronic grammar was all too J34 46 readily formalisable, with the result that the model cannot J34 47 distinguish naturally-occurring from unattested types of J34 48 change. At the same time, the resultant progressive differentiation J34 49 of dialects and languages was entirely unformalisable, due to the J34 50 static nature of the model and the resistance of early generative J34 51 phonologists to change in the underlying representations. In LP, J34 52 the lexicalization of rules and their eventual loss provides a J34 53 mechanism for change at the underlying level and for the J34 54 introduction of surface and underlying dialect variation.

J34 55 In the following sections, I shall show that the Scottish Vowel J34 56 Length Rule (SVLR) provides further evidence for these proposals, J34 57 and constitutes an arguably even clearer illustration of the 'life J34 58 cycle' of changes and rules outlined above, albeit with some J34 59 interesting differences from Harris' example of J34 60 <*_>oe-ligature<*/>-Tensing. In what follows, I assume a two-level J34 61 lexical phonology/morphology of the type proposed by Booij and J34 62 Rubach (1987). For English, irregular inflection and Class I J34 63 derivation will operate on Level 1, and Class II derivation, J34 64 compounding and regular inflection on Level 2.

J34 65 2. THE SCOTTISH VOWEL LENGTH RULE

J34 66 2.1. The synchronic SVLR: a preliminary discussion

J34 67 A typically English dialect is one which preserves a J34 68 reflex of the West Germanic system of phonemic vowel length, having J34 69 one set of lexically short and one of lexically long stressed vowel J34 70 phonemes ... Scots dialects, on the other hand, are characterized J34 71 by the disruption of this dichotomous pattern, resulting in the J34 72 loss of phonemic length: vowel duration is to a large extent J34 73 conditioned by the phonetic environment (Harris, 1985: J34 74 14).

J34 75 The process generally said to be responsible for this J34 76 historical loss of contrastive vowel length in Scots, and for J34 77 controlling the synchronic distribution of long and short vowel J34 78 allophones, is the Scottish Vowel Length Rule. SVLR was first J34 79 formulated in 1962 by A.J. Aitken (after whom it is also called J34 80 Aitken's Law), although its effects had been observed much earlier, J34 81 in dialect studies like Patterson (1860), Murray (1873), Watson J34 82 (1923) and Zai (1942). A preliminary formulation of the synchronic J34 83 SVLR, which applies in Scots dialects and in Standard Scottish J34 84 English (SSE; the local sociolinguistic equivalent of RP, see J34 85 Abercrombie, 1979) is given in (5), and an illustration of its J34 86 effects for two vowels appears in (6).

J34 87 diagram

J34 88 table

J34 89 SVLR applies to the monophthongs /i u e o/ (found in the J34 90 standard lexical sets FLEECE/NEAR; FOOT/GOOSE/CURE; J34 91 FACE/SQUARE and GOAT/FORCE (Wells, 1982)); to J34 92 /a<*_>unch<*/>/ (lexical sets TRAP/PALM/START and J34 93 LOT/THOUGHT/NORTH (Wells, 1982)) in some varieties, although J34 94 these are consistently long in others, and to the first element of J34 95 the diphthong /<*_>unch<*/>i/ (lexical set PRICE (Wells, J34 96 1982)), the only vowel to be affected both qualitatively and J34 97 quantitatively. Its effects on the two other Scots diphthongs, /au/ J34 98 (MOUTH (Wells, 1982)) and /<*_>unch<*/>i/ (CHOICE (Wells, J34 99 1982)), is unclear; /<*_>unch<*/>i/ is relatively rare in English J34 100 in general, while /au/ is extremely peripheral in the Scots system, J34 101 since Middle English /u:/ did not diphthongize to /au/ by the Great J34 102 Vowel Shift in Northern dialects, including Scots. The result is J34 103 that lexical items like house, out, town have J34 104 <*_>unch<*/> in modern Scots dialects; where /au/ does appear, in J34 105 formal registers and SSE, it may best be regarded as a borrowing J34 106 from Southern English English, and might not then be expected to J34 107 undergo a Scots-specific process like SVLR.

J34 108 The specification [+tense] in the structural description of J34 109 SVLR in (5) is intended to include the input vowels listed above, J34 110 while excluding the non-lengthening lax vowels J34 111 /<*_>unches<*/>/. Lass (1974), Aitken (1981) and Harris (1985) all J34 112 accept that /<*_>unches<*/>/ are exceptions to SVLR, but consider J34 113 /<*_>epsilon<*/>/ to be a lengthenable vowel, although they do not J34 114 discuss it in any detail. However, there is little evidence for its J34 115 inclusion in the lengthening set; in fact, /<*_>epsilon<*/>/ is J34 116 consistently classified as non-lengthening in earlier dialect J34 117 studies (Dieth, 1932; Grant, 1912; Wettstein, 1942). Recent J34 118 experimental studies are inconclusive: Agutter (1988a, b), whose J34 119 work will be discussed further below, did not test J34 120 /<*_>epsilon<*/>/, and McClure (1977), who did, was unable to J34 121 examine it in as full a range of contexts as the other allegedly J34 122 lengthenable vowels. For instance, /<*_>epsilon<*/>/ does not J34 123 appear in stressed open syllables, so that no examples of this J34 124 vowel word-finally or before inflectional [d] or [z] are available. J34 125 /<*_>epsilon<*/> is also rare before /r/ and voiced fricatives, and J34 126 McClure was forced to resort to using the names Kerr J34 127 [k<*_>epsilon<*/>r] and Des [d<*_>epsilon<*/>z].

J34 128 McClure (1977) does claim to have found results broadly in line J34 129 with the length modification expected if SVLR did affect J34 130 /<*_>epsilon<*/>/. However, only one informant, McClure himself, J34 131 was involved in this experiment, and his average vowel duration and J34 132 range of durations were considerably higher than those of any J34 133 speaker tested by Agutter (1988a, b). This makes McClure's findings J34 134 unreliable, since it is at least possible that they reflect J34 135 "an exaggerated differentiation of vowel length in long and J34 136 non-long contexts and extreme carefulness on the part of an J34 137 informant who knew the purpose of the experiment" (Agutter, J34 138 1988b: 15).

J34 139 Since the early dialect evidence is the most conclusive J34 140 currently available, I accept that /<*_>epsilon<*/>/, along with J34 141 /<*_>unches<*/>/, is an exception to SVLR. There is distributional J34 142 and behavioural evidence that these three vowels form a natural J34 143 class of [-tense] vowels in Scots/SSE. All three fail to occur in J34 144 morpheme-final stressed open syllables; furthermore, in casual J34 145 registers or under low stress, /<*_>unch<*/>/ and /<*_>unch<*/>/ J34 146 tend to fall together for Scots speakers, and the entire set J34 147 /<*_>unches<*/><*_>epsilon<*/>/ may also merge, as shown in (7).

J34 148 diagram

J34 149 The inclusion of [+tense] as an input condition for SVLR will, J34 150 then, effect the appropriate exclusions, and is motivated insofar J34 151 as the feature [<*_>unch<*/>tense] itself is motivated. However, as J34 152 Halle (1977: 611) notes, "the feature of tenseness has had J34 153 a long and complicated career in phonetics", and objections J34 154 have been raised against its integrity and validity. Certainly the J34 155 claim that LP improves upon standard generative phonology, for J34 156 instance in that it combats abstractness (McMahon, 1989), can gain J34 157 nothing from avowed support for a "pseudo-feature" (Lass, J34 158 1976). One of the most vocal detractors of [<*_>unch<*/>tense] is J34 159 Lass (see especially Lass, 1976), who bases his case for the J34 160 abandonment of the feature largely on the difficulty of locating J34 161 distinct, measurable phonetic correlates for it. Lass holds that, J34 162 when two vowels differ with respect to a cluster of factors such as J34 163 relative height, backness and degree of rounding, these should be J34 164 considered separately rather than ascribed as a set to "an J34 165 explanatory abstraction" (Lass, 1976: 49) like tenseness. I J34 166 believe, however, that these arguments can be countered, and J34 167 outline five replies below.

J34 168 (i) It is true that tenseness is intimately connected with J34 169 tongue height, frontness/backness and degree of lip-rounding. J34 170 However, the tense-lax dichotomy does not rest on the individual J34 171 contribution of these features, but on their variable conjunction; J34 172 and the weighting of contributory components is not constant in J34 173 distinguishing different tense-lax pairs. So, although tense vowels J34 174 tend to be more peripheral than their lax counterparts, the J34 175 interpretation of peripherality is fluid. A high front tense vowel J34 176 will thus be higher and fronter than its lax equivalent, while a J34 177 low back rounded vowel expresses its peripherality J34 178 vis-<*_>a-grave<*/>-vis its lax partner by being lower, more back J34 179 and more rounded (although some low tense vowels MAY be less J34 180 peripheral than their lax counterparts). It is this variable J34 181 clustering of features, which would be difficult to relate using J34 182 only the contributory elements, that [<*_>unch<*/>tense] is J34 183 intended to encapsulate (for an example of a similar use of J34 184 [<*_>unch<*/>tense], see Allen, 1965).

J34 185 (ii) The use of [<*_>unch<*/>tense] may make otherwise opaque J34 186 processes characterizable (see, for instance, Lieber's (1979) J34 187 account of Middle English Open Syllable Lengthening). This is J34 188 surely one of the major tasks of linguistics, and a primary J34 189 requirement of the formal and theoretical tools it employs.

J34 190 (iii) Lass asserts that tenseness is definable only according J34 191 to its effects rather than "on the basis of a prior J34 192 (historically based) partitioning of the lexicon" (Lass, J34 193 1976: 40). However, a "historically based" J34 194 characterization can be found for the three lax vowels J34 195 /<*_>unches<*/> <*_>epsilon<*/>/ in modern Scots/SSE, which form a J34 196 historically motivated class as the only vowels in the system with J34 197 no tense (or long) Middle English sources. /i e u o/ and the J34 198 diphthongs have only tense sources, and /a <*_>unch<*/>/ result J34 199 from the merger of tense and lax vowels (Lass, 1974). However, J34 200 /<*_>unches<*/>/ have only short, lax sources, as does J34 201 /<*_>epsilon<*/>/, given that all its possible tense source vowels J34 202 collapsed with other long tense vowels during the Great Vowel J34 203 Shift: ME /<*_>epsilon<*/>:/ raised to /e:/ and subsequently, in J34 204 some cases, to /i:/, and although /a:/ in turn raised to J34 205 /<*_>epsilon<*/>:/, it afterwards continued to /e:/, leaving the J34 206 long half-open front slot empty after the completion of the Vowel J34 207 Shift. Synchronically, /<*_>unches<*/><*_>epsilon<*/>/ cannot be J34 208 classified simply as short, since all Scots/SSE vowels are J34 209 underlyingly short, but these alone fail to undergo SVLR.

J34 210 (iv) The property of tenseness may not always be reducible to a J34 211 combination of length plus other phonetic features: for instance, J34 212 it has been claimed that there are cases where both long and short J34 213 vowels may be tense, as in Icelandic (Anderson, 1984: 95-96). J34 214 Furthermore, the association between length and tenseness may vary, J34 215 not only cross-linguistically, but diachronically in one language. J34 216 Thus, whereas in Middle English long vowels are consistently tense J34 217 and vice versa, the advent of SVLR has altered this correlation for J34 218 Scots/SSE, where tense vowels are now those that may become long J34 219 under certain phonological circumstances. Similarly, Harris' J34 220 (1989a) work on <*_>oe-ligature<*/>-Tensing suggests that J34 221 underlying restructuring in varieties like Philadelphia and New J34 222 York City may be producing a distinction of short lax J34 223 /<*_>ae-ligature<*/>/ and short tense /<*_>AE-ligature<*/>/.

J34 224 (v) The above argument is ultimately circular, so long as J34 225 'tense' and 'lax' cannot be independently defined in phonetic terms. J34 226 J35 1 <#FLOB:J35\>These words bear examination.

J35 2 It seems to have been the case that in the parent dialect, CF, J35 3 these words formed part of a lexically-marked stratum of loans from J35 4 Classical Latin. Yet in medieval Latin, "all unstressed J35 5 vowels had become lax" (Halle and Keyser 1971:99). J35 6 Consequently, in the parent dialect, these words are arguably J35 7 stressed not according to the native French paradigm (which would J35 8 have tended to make all such words oxytone), but according to J35 9 Latinate principles: the final syllable is skipped; then the J35 10 penultimate syllable is analysed, and if it is light, that too is J35 11 skipped, and stress is assigned to the antepenult. This principle J35 12 CF seems to gift to English. Notice, however, that in two of the J35 13 examples above (Tydeus and Zepherus), antepenultimate J35 14 stress can also be accounted for by the generalising of the OE Word J35 15 Rule, which will assign strongest stress to the initial syllables J35 16 (where antepenultimate = initial). In other words, we are now J35 17 dealing with two stress principles, but with one result: in many J35 18 cases, the survival of native English principles of stress J35 19 (specifically, the Word Rule) yields identical outputs to J35 20 stress-patterning based on a new principle.

J35 21 This is by no means an unusual situation in language. Giegerich J35 22 (1985:23), for example, notes that the same thing happens in J35 23 German: "a Latinate stress rule has entered into the J35 24 borrowing Germanic language .... One of the reasons for the J35 25 survival of this Latinate stress rule, and indeed for its J35 26 productivity, may be its compatibility with native German J35 27 vocabulary. Note that the [Latinate: McC] rule copes with native J35 28 and nonnative vocabulary alike and no such distinction has to be J35 29 maintained in the phonology of English".

J35 30 In terms, then, of the English of the 14th century and beyond, J35 31 what we are looking at is a very familiar pattern: the central J35 32 stress-assignment rule is in certain words skipping up to two light J35 33 syllables before assigning stress. Halle and Keyser, for example, J35 34 write of this that "The nonnative vocabulary of Chaucer J35 35 consisted of two types of words, namely, learned words largely of J35 36 Latin origin and everyday words borrowed from Old French or J35 37 Anglo-Norman. These two classes had different stress patterns. The J35 38 words of Latin origin were stressed on the antepenultimate vowel if J35 39 the penultimate syllable ended with a weak cluster [= was a light J35 40 syllable; McC]; otherwise, they were stressed on the penultimate J35 41 vowel" (1971:99). Thus what we are dealing with in many of J35 42 the loans from Central French are words derived directly from Latin J35 43 where, as we have already seen, final syllables may be analysed as J35 44 extrametrical. It is from this point on, I think, that we can trace J35 45 the history of Noun Extrametricality in English.

J35 46 Once this happens, something very interesting and, as far as I J35 47 know, unique in English phonology happens. We have a subsequent J35 48 period in English where metrical overlap and ambiguity occurs. J35 49 Arguably, this period is at its height in the 16th century, the J35 50 period when it seems that Latin-based rules begin to be fully J35 51 operative in English - and of course, the period when the J35 52 quantitative metrical experiments take place.

J35 53 But let me return to the concept of metrical overlap. The point J35 54 is made by Dobson. Speaking specifically about late ME and early J35 55 ModE., he notes "the coexistence in English of two modes of J35 56 stressing" (1957, Vol. 2:831). These modes were the earlier J35 57 French (A-N) mode, the mode of right-strength and end-stress, and J35 58 the mode which almost wholly replaced it, the generalising of the J35 59 native Word Rule allied with the new Latinate principles, which J35 60 tended to shift the primary stress to the initial syllable. Once J35 61 this happens, we have, as I have noted, a large group of words J35 62 which had secondary stress at their right edges. But what happened J35 63 to these secondary-stressed syllables? The answer is that J35 64 in both disyllables and trisyllables, they tend to be reduced: J35 65 thus, in disyllables, against empire, and increase (where J35 66 secondary stress is kept), we have e.g. captain and J35 67 pleasure, where secondary stress is lost. As Dobson notes, in J35 68 Romance disyllables, "especially if they were or became J35 69 'popular', shift of stress often led to the second syllable J35 70 becoming totally unstressed" (1957, Vol. 2:831). This also J35 71 occurs in trisyllables, albeit less frequently (see e.g. Dobson J35 72 1957, Vol. 2:827ff.). One bench mark is Levins' Manipulus J35 73 vocabulorum, printed in 1570, which suggestively marked the J35 74 following (primary) stressings: mem<*_>o-acute<*/>rial, J35 75 or<*_>i-acute<*/>ginal, geom<*_>e-acute<*/>trical, as well as J35 76 penultimate stress on ori<*_>e-acute<*/>ntal, J35 77 sacram<*_>e-acute<*/>ntal, accid<*_>e-acute<*/>ntal and final J35 78 stress on e.g. deb<*_>a-acute<*/>te, div<*_>i-acute<*/>ne J35 79 etc. The pattern is familiar: although there is a little way to go J35 80 yet, it is essentially that of today (see further Halle and Keyser J35 81 1971:109ff.).

J35 82 Given Dobson's dating, and the present reconstruction, it would J35 83 seem that secondary stresses at the right edges of words are in J35 84 many cases being reduced at exactly the same time as Latinate J35 85 stress patterning, and with it, the possibility of final syllables J35 86 being extra-metrical, is entering the language. In short, J35 87 we have a unique position in which Latin stress rules are, at a J35 88 certain stage in linguistic history, compatible with native J35 89 vocabulary and stress-principles (see again the quote from J35 90 Giegerich 1985:23, above p. 18). Whether on lexical monosyllables, J35 91 disyllables, or trisyllables, the old rules, and the new, could J35 92 produce similar, and in many cases identical outputs. By J35 93 c.1570 it is possible to see the new lineaments of J35 94 English stress patterning. In (12) below I give a graphic outline J35 95 summary of the chief developments in English stress patterning from J35 96 c.1100 to c.1600.

J35 97 figure&caption

J35 98 3. But what of the poets themselves? If my reconstruction and J35 99 dating are correct, surely our poets should have detected the J35 100 similarities between their English, and the theories of Latin J35 101 prosody they had available to them in the mid- to late 16th J35 102 century? Although the evidence is again difficult to reconstruct, J35 103 this seems to be what happens. A link can be made (and for J35 104 testament to the possibility of linking poetic form to J35 105 (phonological) characteristics of the language, see Allen 1973:12). J35 106 But to understand the link between the Latinate stress phonology J35 107 entering the language and the favoured Latin verse forms of the J35 108 later 16th century, I must first, having reconstructed the J35 109 phonological background, reconstruct the metrical background.

J35 110 While most discussion on the metrical background has focused on J35 111 syllabic quantity (see below), stress itself is not wholly J35 112 irrelevant to the quantitative enterprise. This is so for two J35 113 reasons: first, syllabic quantity in Latin is directly relevant to J35 114 the placement of stress, and so there is a prima facie case that J35 115 stress and quantity may be linked in the verse line. (However that J35 116 relationship may be exploited by poets, it is still a J35 117 relationship.) Second, and consequently, one way of reading and/or J35 118 performing the verse line (specifically, the hexameter) seems to J35 119 have been accentual, i.e. whatever the underlying patterns of J35 120 quantity determining the abstract structure of the line, one J35 121 possible exponent of that pattern is an accentual reading, one J35 122 based on word stress. As Allen puts it (1973:339), "it has J35 123 to be remembered that in ordinary Latin speech ... quantity was J35 124 closely linked with accent; and it is possible that the untaught J35 125 speaker would have been aware of quantitative differences only in J35 126 so far as they were connected with the placement of J35 127 accent". But again, matters are complicated by the fact J35 128 that in the first part of the hexameter, accent, and ictus (= unit J35 129 of the abstract quantitative pattern) did not necessarily coincide, J35 130 and only, in fact, coincided regularly in the last two feet of the J35 131 line. Nevertheless, the link between stress and quantity is J35 132 important: for counterpoint to exist at all in the first part of J35 133 the line does not mean that stress (or accent) is irrelevant, only J35 134 that the link is differently exploited.

J35 135 Allen (1973:345ff.) argues convincingly that "in J35 136 antiquity the scanning pronunciation [i.e. the ictic reading of a J35 137 line: McC] was the less normal way of reading [than an accentual J35 138 one: McC]" and suggests that such a "two-fold J35 139 tradition - pedagogical scanning and accentual reading" J35 140 persists "from the Latin grammarians to the present J35 141 day" (1973:344). Attridge (1974:35) corroborates this: J35 142 "It seems certain ... that the accepted practice in England J35 143 (and elsewhere) was to read Latin verse with normal J35 144 word-accents". Yet the tradition of 'scanning' was not dead J35 145 in England: "We may conclude that the best-educated men in J35 146 England at the end of the sixteenth century would have held that J35 147 the correct way of reading Latin verse was with prose stresses, but J35 148 that even they would be accustomed to using the stressed-ictus J35 149 method for learning by heart or for scanning" (Attridge J35 150 1974:40; see also Pulgram 1975:192). The upshot of this discussion J35 151 is, then, that stress was relevant to the quantitative enterprise J35 152 in England, even if in unexpected ways. The tradition of accentual J35 153 reading, moreover, helps to explain the demise of the movement (see J35 154 section 4 below). It remains, however, for us to look in more J35 155 detail at theories of syllabic quantity.

J35 156 The 16th century had inherited a theory of Latin prosody J35 157 (Attridge 1974:9) where syllables could be long 'by nature': these J35 158 were typically lexical monosyllables (nox, 'night', J35 159 frons, 'foliage'), or syllables containing long vowels or J35 160 diphthongs (e.g. the first syllable of via, 'way'). Syllables J35 161 could also be long 'by position' (see below). In both cases, J35 162 syllabic length (or heaviness) may imply stress. The position rule, J35 163 for example, the more difficult of the two rules, implied that a J35 164 (non-final) syllable was stressed if it was heavy (e.g. first J35 165 syllables of cul.pa, 'blame', fal.sus, 'false'). It also J35 166 claimed a syllable was long (but not necessarily implying stress) J35 167 if it was followed, in the same or different words, by two J35 168 consonants (Attridge 1974:9, based on Raven 1965:23-25); so the J35 169 first syllable of et d<*_>unch<*/>na is 'long' by this J35 170 principle. There were also a group of syllables falling within the J35 171 purview of the position rule which could be scanned as either short J35 172 or long (light or heavy); these were syllables followed by plosive J35 173 + /l, r/. Thus the first syllable of atrox, 'fierce', could be J35 174 scanned as light (a.trox - a later age would call this J35 175 initial-maximal syllabification), or heavy (at.rox) - see J35 176 again Attridge (1974:9). These points are detailed under (13).

J35 177 (13) Latin theories of syllable-structure and versification:

J35 178 a. Syllables long by nature: (i) lexical J35 179 monosyllables (e.g. nox, frons etc.; see also the J35 180 discussion of XVC# syllables below); (ii) syllables containing long J35 181 vowels or diphthongs (e.g. via).

J35 182 b. Syllables long by position: (i) first syllables of J35 183 e.g. cul.pa, fal.sus; (ii) first syllable of e.g. et J35 184 d<*_>unch<*/>na, where is classed as 'long by position' J35 185 since it precedes two consonants.

J35 186 c. Syllables classifiable as long or short: at.rox or J35 187 a.trox, pat.rius or pa.trius etc.

J35 188 Let me try to reinterpret this in terms of non-linear theory. J35 189 What had developed in the history of English was a new kind of J35 190 trading relationship between syllabic heaviness and stress. Of J35 191 course, it is arguable that syllabic heaviness has always traded J35 192 with stress in English - it certainly does in OE, where the J35 193 rightward migration of the OE Stress Rule is quantity-sensitive at J35 194 the left edges of domains: consider here the contrast between J35 195 t<*_>i-acute<*/>mbr<*_>o-grave<*/>de ('he built', with J35 196 secondary stress, since the first syllable is heavy) and J35 197 b<*_>i-acute<*/>fode ('it trembled', with no secondary J35 198 stress since the first syllable is light; McCully and Hogg J35 199 1990).

J35 200 Recall now what the new, Latin-based English Stress Rule (ESR) J35 201 does: it assigns foot-headedness - stress - at the right edges of J35 202 words (not the left edges, as in OE) on the basis of syllable J35 203 weight. This relationship is easily observed in English J35 204 monosyllables, which are all syllabically heavy and all J35 205 "inherently stressed" (Hogg and McCully 1987:37). J35 206 In disyllables, too, heavy initial syllables (as well as light J35 207 ones, a fact which puzzled some 16th-century theorists) are J35 208 primary-stressed. When our 16th-century poets studied Latin J35 209 prosody, they found a similar relationship: heavy syllables are J35 210 equated with 'length', and length implies (or increases the J35 211 expectation of) stress. And although the 16th century was to J35 212 misinterpret part of this implication (by taking it to mean not J35 213 that syllables were 'long by position' but that their vowels J35 214 'lengthened by position'), what is crucial here is the idea of J35 215 syllable structure.

J35 216 J36 1 <#FLOB:J36\>22. International education of the highly J36 2 able

J36 3 Joan Freeman

J36 4 In education, there is no escape from politics, nor for that J36 5 matter from cultural influences. These truths become emphasized in J36 6 the education of the highly able; in the same way that these J36 7 exceptional individuals function at an extreme, so too do attitudes J36 8 towards them. To consider international variations in the education J36 9 of the gifted, enables some conclusions to be drawn about systems J36 10 which appear to work well for them and which may be transferable to J36 11 other countries. However, it is often difficult to extricate a J36 12 nation's cultural outlook from its specific educational J36 13 practices.

J36 14 I am concentrating on three major cultural, rather than J36 15 geographical, divisions in which there are relatively distinct J36 16 forms of provision for the highly able. They overlap considerably, J36 17 in that pockets of all kinds exist within each division, but any J36 18 other division, such as into fluid and rigid societies, is at least J36 19 as value-laden. These divisions are: first, the Western World - J36 20 including Western Europe, North America, and Australasia; second, J36 21 Eastern Europe - the countries that were Communist until 1989/90; J36 22 and third, the Developing World - Africa, South America, and the J36 23 Far East.

J36 24 International concern for the gifted

J36 25 In all societies, the reasons for concern about the highly able J36 26 are dual - to serve the individuals in their personal fulfilment, J36 27 and to serve the community. This special concern for the highly J36 28 able is humane (at the very least) and will eventually increase J36 29 knowledge to help enhance everybody's life, although a high J36 30 intelligence is not necessarily the foundation of good leadership J36 31 or superior morality. For example, a child with an exceptionally J36 32 high IQ may have been brought up in an overly scholarly regimen, J36 33 which could produce either a religious dictator, or an academic J36 34 professor who lacks both social competence and the ability to cope J36 35 with others' demands (Miller 1990). The highly able are neither J36 36 more nor less moral than anyone else.

J36 37 Experimental education for highly able children is increasing J36 38 and the results are becoming more readily available to teachers, J36 39 among whom concern for the gifted is growing steadily. However, as J36 40 teachers in most of the world are only just beginning to accept the J36 41 idea that gifted children do need special provision, this is often J36 42 uneven and dependent on individual efforts.

J36 43 There is a need to develop policies for the highly able on an J36 44 international basis, but the greatest care must be taken in their J36 45 preparation. They should make full use of the research knowledge we J36 46 are accumulating, because opinions are merely opinions, and should J36 47 always be seen as such. What is valued as high ability in one part J36 48 of the world may not be seen as such in another - such as a trance J36 49 state which is valued in Thailand, but may put the child in the J36 50 hands of a clinical psychologist in the West.

J36 51 Some international problem areas

J36 52 There are two major, pervasive influences which often prevent J36 53 the identification and development of gifted potential, and their J36 54 effects are cumulative for each individual. The first is J36 55 socioeconomic status - even in the relatively rich West and in the J36 56 1990s. The second is gender stereotyping, which slots boys and J36 57 girls into areas of study that are not necessarily the most J36 58 appropriate for their abilities. When children are nominated by J36 59 parents or teachers, without testing, as gifted, this usually J36 60 results in two boys being presented for every girl. The physically J36 61 handicapped are often missed too, since giftedness which is not J36 62 obvious or all-round can be difficult for teachers to spot and cope J36 63 with.

J36 64 In multicultural societies, children of non-majority cultures J36 65 have to adapt to both home and school, and consequently may not J36 66 fulfil the expectations for gifted pupils in the educational J36 67 system. For example, the children's spoken language and J36 68 self-expression may be considerably better in their home language, J36 69 but to find this out requires efforts by teachers towards close J36 70 communication with parents, probably including home visits. This J36 71 consideration also applies to children born in inner cities who may J36 72 have limited school language ability but excellent 'street J36 73 language', causing them to be relatively non-communicative J36 74 in the classroom. It applies also to the children's performance on J36 75 nationally standardized tests of ability. The culture in which J36 76 children feel comfortable can even be set to work against J36 77 achievement at school, in the simplest sense of 'them and us'. This J36 78 often occurs in an educational system which is imposed from above, J36 79 without adequate concern for the real needs of the recipients.

J36 80 Acceleration is not uncommon for gifted children worldwide, as J36 81 it is the easiest way a school can manage them. Alternatively, a J36 82 school may just have one fast stream or even one teacher working J36 83 faster with a bright group, most frequently for mathematics. This J36 84 is a form of acceleration 'without tears', avoiding the frequently J36 85 accompanying problems of this move, such as immaturity relative to J36 86 others in the class, difficult peer relations, or small physical J36 87 size, when one child is very much younger in the teaching group. J36 88 Acceleration may also be offered in certain subjects for individual J36 89 children, if the school can cope with its timetabling. In some J36 90 countries, there are specialist teachers to provide guidance for J36 91 the education of the gifted.

J36 92 Research evidence indicates the importance of both material and J36 93 cultural provision if children's abilities are to be developed to J36 94 an exceptionally high level. I have been following up both gifted J36 95 and control children in Britain since 1974: all the 210 children, J36 96 their families and schools were visited and questioned. The J36 97 children's emotional adjustment was found to be independent of J36 98 their measured IQs. However, their scholastic achievements were J36 99 significantly affected by the provision of learning materials and J36 100 tuition, as well as by the family culture. The follow-up results J36 101 have reaffirmed the message from the original study - that gifted J36 102 children need adequate material and teaching provision to realise J36 103 their potential at both school and home. Relationships between J36 104 teachers and pupils were also seen to affect the pupils' J36 105 self-concepts and thus their achievements. Some children's gifts J36 106 appeared to have been abused, in the sense that their education was J36 107 intensely focused on examinations to the detriment of their J36 108 creative side, so that some had taken examination success alone as J36 109 the source of their self-esteem (Freeman 1991).

J36 110 The Western World

J36 111 Most children's physical needs are being met in the Western J36 112 World. Indeed, the common aim of free compulsory education for all J36 113 children, continuing with the availability of further education J36 114 throughout people's lives, has been almost accomplished. This means J36 115 that energy has been freed to refocus on improving educational J36 116 opportunities. Although national resources and outlook are varied, J36 117 there is a considerable commonality of approach to educational J36 118 research and development, built on the scientific and philosophical J36 119 tradition which spread from the 'old world' of Western Europe to J36 120 the 'new worlds' of North America and Australasia. However, more J36 121 subtle differences in outlooks between the thrusting new and the J36 122 complacent old are apparent in what they offer the highly able.

J36 123 In Europe, as the political edifices of many previously J36 124 Communist countries crumbled, thousands of bright young people J36 125 seeking a better life have streamed into the western democracies. J36 126 They bring with them different attitudes to the education of the J36 127 highly able, and a few are even bringing world-class teaching J36 128 expertise in subjects such as sports, gymnastics, singing, ballet, J36 129 circus, and mathematics.

J36 130 The style of interest they bring is different too. Eastern J36 131 Europe has generally been more concerned with the practicalities of J36 132 teaching and outcome, whereas the West has a more prolific output J36 133 of theory. At its extreme, the difference seems to be like that J36 134 between the training of young gymnasts on a trampoline and the J36 135 production of academic research papers which conclude that we need J36 136 more research. Yet, because of these recent political changes, the J36 137 extra money and provision which have been put into the promotion of J36 138 competitive excellence may now be redirected within Eastern J36 139 countries, so that the two halves of Europe may eventually become J36 140 more balanced in top-level achievements.

J36 141 Historically, Europe has had hundreds of years of selection in J36 142 education, which still lives in its collective memory, and to a J36 143 small extent in reality. Although this was originally by social J36 144 status and money, the Middle Ages saw the beginnings of selection J36 145 by ability for some boys. Present-day British grammar schools, J36 146 German gymnasium, and French lyc<*_>e-acute<*/>es, are schools J36 147 which emerged from the 15th century, to provide academic tuition, J36 148 originally in preparation for clerical and church positions.

J36 149 Selection by ability gripped the whole of Britain in 1945, when J36 150 every child in the country was tested at 11 years-old for entry to J36 151 the grammar schools - the Eleven Plus exam. It succeeded in J36 152 providing bright, working-class children with the opportunity to go J36 153 to previously middle-class schools, but failed to make allowance J36 154 for individual development, and has almost entirely been abandoned. J36 155 Among other problems, there was much incorrect placement, as well J36 156 as wide variation in the number of grammar school places between J36 157 education authorities. In the 1990s most British children, like J36 158 their counterparts on the European continent and in the rest of the J36 159 Western World, attend comprehensive, all-ability, J36 160 neighbourhood schools. The difference is that in Britain and the J36 161 USA, about 7 per cent of children go to private schools.

J36 162 Nevertheless, many comprehensive schools throughout the Western J36 163 World are still internally selective, having fast streams or J36 164 tracks, or other forms of division of the children for different J36 165 levels of tuition. Younger children may be sorted in class by more J36 166 subtle and flexible distribution into groups, identified by the J36 167 teacher's estimate of their abilities.

J36 168 Western attitudes to high ability

J36 169 In most Western countries, the training of teachers for guiding J36 170 the learning of gifted children is patchy. It is possible, for J36 171 example, to take a doctorate in teaching gifted children in a few J36 172 parts of the USA, but in Western Europe it would be difficult to J36 173 find even a teaching diploma which provided that kind of J36 174 instruction. However, in Munich, a new college for teachers of the J36 175 gifted is due to open in 1992, and a new postgraduate course is J36 176 about to start in Switzerland. There is a little input on the J36 177 subject at initial training level, but short in-service courses are J36 178 growing everywhere.

J36 179 For most Western teachers, the spectre of J36 180 <*_>e-acute<*/>litism, whether of wealth or ability, is their major J36 181 apprehension. They do not like categorizing children, and have J36 182 serious doubts about the validity of psychological tests of J36 183 ability. Also, they often feel that giving more time and money to J36 184 the already gifted would take resources from those who have more J36 185 obvious needs, such as the handicapped. There is still a long way J36 186 to go before the majority of teachers in the West become convinced J36 187 that gifted children need special attention, and are prepared to J36 188 give it to them.

J36 189 However, in its individualistic way, the West engenders J36 190 voluntary groups who exist to make good what they see as gaps in J36 191 provision. Most countries have at least one association for the J36 192 gifted, run by parents, indeed, many have several (often competing) J36 193 associations. In fact, the amount of interest in a country could be J36 194 judged by counting up the number of these parent associations. J36 195 These usually run out-of-school sessions for the children and their J36 196 siblings, as well as summer schools and weekend conferences in J36 197 special subjects. Although they may also investigate what J36 198 facilities are available and give help to individuals, their J36 199 overall aim is the greater recognition of gifted children and the J36 200 encouragement of appropriate provision for them within normal J36 201 schools. Mensa is an international non-scientific association, J36 202 largely social, to which people can belong by passing a test, set J36 203 by the association. The European Council for High Ability (ECHA), J36 204 based in Bonn, Germany, is the only association of psychologists J36 205 and educators which uses evidence rather than opinion, and which J36 206 crosses all European national boundaries.

J36 207 There are hundreds of privately funded national competitions J36 208 and activity centres for the gifted in the West. Many countries J36 209 have a Young Scientist of the Year, Young Engineer, Young Musician, J36 210 etc. International competitions include the Mathematics Olympiad J36 211 and Foreign Language Competitions. In many countries, local J36 212 education authorities and individual schools may receive financial J36 213 help from industry for specific projects, such as scientific field J36 214 trips or collective creativity on a technical project.

J36 215 J37 1 <#FLOB:J37\>Since contemporary opinion about the Charity J37 2 Commissioners varied so much it is no easy task to form a balanced J37 3 judgement of their work. They thought that they had done well. They J37 4 paid little regard to the arguments of their opponents, though the J37 5 difficulties which they faced must not be underestimated. Once J37 6 again the best method is to study some actual cases. The select J37 7 committee of 1886-7 was set up to examine the administration of the J37 8 Endowed Schools Acts after an earlier committee of 1884 had looked J37 9 at the Charitable Trusts Acts (PP 1884 IX: 3-11). The 1886-7 J37 10 committee heard evidence in particular on six endowments. Of these J37 11 Kendal, Sutton Coldfield, and Norwich, which was a hospital J37 12 endowment, raised matters which have already been fully discussed. J37 13 The fourth case, the removal of the grammar school from Hemsworth J37 14 to Barnsley in the West Riding, had been a keenly fought battle, J37 15 and is the only major example of the transfer of an endowment from J37 16 one place to another (PP 1886 IX: 90: 856-9 (Sir George J37 17 Young)). The other two cases, Scarning in Norfolk and West J37 18 Lavington, deserve fuller study. They were both villages, and thus J37 19 far the impact of educational change on rural communities has not J37 20 been studied. They attracted the attention of Radicals like Jesse J37 21 Collings and Joseph Chamberlain, who were campaigning to raise the J37 22 status of the rural labourer. They throw some interesting light on J37 23 the ideas of the commissioners and on the ways in which they sought J37 24 to carry them out.

J37 25 The commissioners became involved with both Scarning and West J37 26 Lavington in the 1870s. Scarning Free School was an elementary J37 27 school which had remained under the jurisdiction of the J37 28 commissioners because its endowment was above the limit set by the J37 29 act of 1873 (Ed. 21: 12996, 14 December 1872, "land let for J37 30 pounds210"). A scheme for the school was finally confirmed J37 31 in May 1882, though it had met with strong opposition from all J37 32 sections of local opinion (Ed. 21: 12996; Simon 1960: 329-31). The J37 33 scheme provided that the school was to be maintained as a public J37 34 elementary school under the Act of 1870 for children of both sexes. J37 35 Fees "suitable for an elementary school" were to be J37 36 fixed by the governors. Small scholarships were to be held in the J37 37 school, covering the payment of fees and the provision of clothing, J37 38 and there were to be three 'Secker' exhibitions of not less than J37 39 pounds25 a year, each "tenable for three years at any place J37 40 of higher education or professional or technical training or study J37 41 approved by the Governors". These were to be open both to J37 42 boys and to girls with a first preference to those educated in the J37 43 school, followed by children from six neighbouring parishes. The J37 44 scheme bore all the hallmarks of commissioners' orthodoxy. Fees J37 45 were to be charged, whereas the school had previously been free. A J37 46 large part of the endowment was to be spent on scholarships and J37 47 exhibitions for further education. These awards, the trustees had J37 48 been told, represented the appropriate charge on the foundation for J37 49 purposes of higher education (Ed. 49: 5526, 31 May 1880).

J37 50 The scheme had been opposed by the trustees, by the vestry, by J37 51 the rector, and by a meeting of the inhabitants. The trustees J37 52 claimed that, by tradition and in the memory of those who had been J37 53 educated there, Scarning Free School "was originally an J37 54 elementary school and that far beyond the memory of man charges J37 55 were made for teaching Latin and Greek while Reading, Writing and J37 56 Arithmetic were taught free of charge". Because there was J37 57 an urgent need to spend money on the school building and on the J37 58 farm which formed the endowment, money was not available for the J37 59 exhibitions which might be provided when there was a surplus (Ed. J37 60 49: 5526, trustees' memorial 5 July 1880). A petition, signed by J37 61 276 inhabitants, both men and women, concentrated on the fact that J37 62 under the new arrangements free education would be abolished. The J37 63 exhibitions, which had been made the first charge on the endowment, J37 64 would be useless to the village children, who would not be able to J37 65 compete for them successfully (Ed. 49: 5526, 9 July 1880). The J37 66 rector, Augustus Jessopp, who had been headmaster of Norwich School J37 67 and had therefore considerable experience of education, took a J37 68 moderate but critical line. He argued that from the trustees' point J37 69 of view the scheme was acceptable in principle, but that in J37 70 practice it would be impossible both to fund the exhibitions and to J37 71 improve the building. The exhibitions ought to be deferred until J37 72 the property had been improved and its value allowed to increase J37 73 (Ed. 49: 5526, 12 August 1880).

J37 74 Both the trustees and the ratepayers petitioned Parliament that J37 75 the scheme should be laid before Parliament. It duly 'lay on the J37 76 table' for two months, but no member was found to take the matter J37 77 up and it became law. Subsequent events were described by Jessopp J37 78 and by an inhabitant, William Taylor, a ganger on the railway (Ed. J37 79 49: 5526, 20 January 1883; PP 1887 IX: 275-91 (Jessopp); 292-4 J37 80 (Taylor)). When the governors introduced a fee of 1d a week in J37 81 January 1883, a large number of children, accompanied by many J37 82 adults including some substantial farmers, presented themselves at J37 83 the school without their fee money and were not allowed to attend. J37 84 "The men", Jessopp wrote, "have since then J37 85 formed themselves into a League or Union to resist the J37 86 payment of the Fee & to intimidate parents & children who still J37 87 persist in attending the School & paying the Fee." William J37 88 Taylor described to the select committee how the villagers had J37 89 started their own school and maintained it for 12-15 months.

J37 90 When Jessopp gave evidence to the select committee in February J37 91 1887 he was examined by Jesse Collings. Jessopp did not think that J37 92 the penny fee was regarded as a great grievance, but in William J37 93 Taylor's view feelings still ran very strongly over the matter J37 94 (PP 1887 IX: 284: 6612; 292: 6764; 294: 6840). Jessopp's J37 95 general assessment was highly unfavourable to what the J37 96 commissioners had done. He had advised the parishioners not to J37 97 adopt a plan of out-and-out resistance, but he had seen the crucial J37 98 difficulty as being the priority claim on the endowment given to J37 99 the Secker exhibitions. In fact these had never been awarded, and J37 100 there was no school within walking distance of the village at which J37 101 they could be held. The real objection to the commissioners' plans J37 102 was, Jessopp thought, that they had kept to their rules and had not J37 103 made any exceptions for Scarning. When he was asked whether the J37 104 commissioners would listen to requests for change, he replied: J37 105 "Judging from the tenacity of purpose which the J37 106 Commissioners have exhibited in their intercourse with me in the J37 107 past, I should be very sorry indeed to trust to their willingness J37 108 to listen to reason" (PP 1887 IX: 289: 6710).

J37 109 It is difficult to understand why the commissioners should have J37 110 acted as they did in the Scarning case. When Richmond was examined J37 111 on the subject by the select committee, he used the familiar J37 112 argument that, if resources were not devoted to purposes like the J37 113 exhibitions, the money spent on the school would simply save the J37 114 ratepayers from the education rate, which they would otherwise have J37 115 to pay (ibid.: 301: 6095; 302: 6913). The commissioners seem to J37 116 have been obsessed by the argument that the income of an endowed J37 117 school should benefit people of all classes, and not simply those J37 118 who needed an elementary education. The priority given to the J37 119 exhibitions reflects the very reasonable view that able children J37 120 should have the opportunity to advance to higher studies. Yet they J37 121 do not seem to have considered what was practicable under the J37 122 circumstances of a Norfolk village. Jessopp had said that there was J37 123 no suitable higher school near at hand. Richmond himself admitted J37 124 that an exhibition of pounds25 on its own was not "enough J37 125 to take the child of a working man to a boarding school" J37 126 (ibid.: 300: 6888). So even if the money for the exhibitions could J37 127 have been found, there were not likely to be any candidates who J37 128 wanted or could afford higher education.

J37 129 In an earlier age Scarning School, like many others, had J37 130 attracted pupils to read classics. By 1880 it was simply a village J37 131 elementary school. The commissioners seem to have devoted their J37 132 main effort to providing what these children did not want or were J37 133 not able to take advantage of. Even if there had been a local J37 134 demand for the exhibitions, was it possible to find the money for J37 135 them? Informed local opinion thought that, at the time when the J37 136 scheme was made, this was not possible. No more attention seems to J37 137 have been paid to that point than to any of the others made by the J37 138 inhabitants. Well might Jessopp say that he would be sorry to trust J37 139 to the commissioners' willingness to listen to reason.

J37 140 The end of the story is a pure anti-climax. In 1888-9 the J37 141 commissioners considered revising the arrangements, and in March J37 142 1889 a new scheme was made, which made some money available to pay J37 143 the fees of poor children with the best attendance records, set J37 144 aside funds for scholarships, and provided for the instruction of J37 145 boys in practical mechanics and agricultural chemistry and of girls J37 146 in cookery (cl. 35). Eight years later a note in the file written J37 147 to Sir George Young explained that the scholarships were not J37 148 awarded, that the exhibitions "seem to have been swept J37 149 away", that the school was now free, and that nothing J37 150 appeared in the accounts under cl. 35 for technical education. J37 151 "The ratepayers seem to have got the whole, now, for J37 152 maintenance" (Ed. 49: 5526, 22 December 1897). Such were J37 153 the practical results of the commissioners' efforts to impose a J37 154 scheme that no one in the locality had wanted.

J37 155 The story of Dauntsey's School at West Lavington in Wiltshire J37 156 brings up some similar problems of rural life. It shows much more J37 157 determined and long-continued resistance by local people, which J37 158 achieved a large measure of success, largely because their cause J37 159 was taken up by Jesse Collings and Joseph Chamberlain. The position J37 160 of the school was a strange one. There was no independent J37 161 endowment. Under the will of Alderman Dauntsey, who died in 1553, J37 162 the Mercers Company of London as beneficiaries had an obligation to J37 163 make certain payments to the schoolmaster and to the almspeople and J37 164 to keep the school in repair. They were neither trustees nor J37 165 visitors, and they did not appoint the master, who was chosen by a J37 166 local landowner, Lord Churchill (Ed. 27: 5291, 21 June 1878 (from J37 167 the clerk to the company)). The Mercers claimed that they had J37 168 always spent more money on the school and almshouses than they were J37 169 legally obliged to do, but the value of the Dauntsey property had J37 170 greatly increased, and most of that increment went into their J37 171 coffers. All through the lengthy negotiations one central point was J37 172 the amount of money which the company could be persuaded to pay out J37 173 of the Dauntsey funds.

J37 174 In March 1878 a group of twenty inhabitants wrote to the J37 175 Mercers, asking that the schoolmaster, who was also vicar of the J37 176 parish and so had little time to attend to the school, should be J37 177 replaced, and that in future the master should be resident in the J37 178 schoolhouse. Of the twenty signatories to this memorial two J37 179 described themselves as gentlemen; the others were farmers, J37 180 millers, bakers, coopers - what might be called the middle ranks of J37 181 village society (Ed. 27: 5291, 22 March 1878). People of this kind, J37 182 though they were sometimes divided among themselves about the J37 183 proper policies to pursue, formed the core of local opposition to J37 184 any efforts to deprive the village of the benefits of the J37 185 endowment. They were supported from time to time by some of the J37 186 local gentry, and more steadily by the labourers who were keenly J37 187 concerned about the fate of the school.

J37 188 This move by the local groups seems to have led to action by J37 189 the Charity Commissioners, who sent Assistant Commissioner Stanton J37 190 to hold a local enquiry in March 1879. J37 191 J38 1 <#FLOB:J38\>When they did, "they were seldom aware of J38 2 positive and effective response".

J38 3 This paper reports on the results of answers to questions on J38 4 teasing and fighting, which 11-year-olds gave in interviews about J38 5 themselves and school. We consider the extent and nature of teasing J38 6 and fighting in general, and in relation to age, sex and ethnic J38 7 group differences.

J38 8 Method J38 9 The sample consisted of 175 children who had taken part in an J38 10 earlier study, known as the Infant School Project, which looked at J38 11 sex and ethnic group differences in relation to children's progress J38 12 through infant school (Tizard et al., 1988). All the J38 13 children had entered reception classes in 33 ILEA schools, in J38 14 September 1982. The schools were in areas which were predominantly J38 15 multi-ethnic and working class. The two largest ethnic groups were J38 16 white British children and black British children of Afro-Caribbean J38 17 origin, subsequently referred to as white and black children. A J38 18 subsample of children from these two ethnic groups had been J38 19 interviewed about their views on school when aged seven and in the J38 20 last year of infant schooling.

J38 21 As part of a follow-up study of the Infant School Project J38 22 sample at 11 years, these same children who had been interviewed at J38 23 seven and who had not moved out of ILEA during the interim period, J38 24 a total of 114 children, were interviewed again. In order to J38 25 increase the number of interviewees, additional white and black J38 26 children from the sample were selected from the 33 study schools. J38 27 Thus the 175 children comprised of 41 black girls (BG), 38 black J38 28 boys (BB), 46 white girls (WG) and 50 white boys (WB).

J38 29 The interview was structured and explored children's attitudes J38 30 and feelings about a range of topics to do with school. Most J38 31 questions were open-ended and allowed for more than one response. J38 32 Children also used the Smiley Face Scale (see Figure 1), pointing J38 33 to the face which best described how they felt - e.g. "How J38 34 do you feel about fighting?". In order to have J38 35 comparability with the earlier interview, questions remained J38 36 largely unchanged, although wording was occasionally altered to J38 37 take account of the older age range and some additional questions J38 38 included to expand areas of interest.

J38 39 figure&caption

J38 40 Each child was interviewed individually at school and was J38 41 assured that what they told us was confidential. The interviews J38 42 took place in the Spring term of 1989 when the children were in J38 43 their fourth year of junior school, and they were told that J38 44 questions referred to this year in particular. There were three J38 45 white interviewers (two females and one male) and interviews took J38 46 approx. 45 min.

J38 47 Results J38 48 In discussing our results, the children's own comments will be J38 49 used to illustrate the quantitative data. Since one of our aims was J38 50 to look at differences between the two age groups, we compare the J38 51 two data sets where appropriate. It should also be noted that J38 52 because children often gave more than one response to open-ended J38 53 questions, figures in some tables do not total 100 per cent. We J38 54 begin by considering the overall prevalence of teasing and J38 55 fighting.

J38 56 table&caption

J38 57 As can be seen from Table 1, almost all the sample for both J38 58 ages said that children at their school were teased. When asked if J38 59 this happened to them, approx. two-thirds say that it had. Fewer J38 60 children at age 11 agree that racial teasing occurs or that it J38 61 happens to them. It should be remembered that children were J38 62 answering with regard to the current academic year - this J38 63 represents approx. one and a half terms for the 11-year-olds. When J38 64 asked if they were teased more, less or about the same as other J38 65 children, over two-thirds (72 per cent) said they were teased less J38 66 and nearly a third (27 per cent) said about the same.

J38 67 In considering continuity over time, we found that for the 114 J38 68 children who were interviewed twice, there was very little J38 69 association between their responses at seven years and their J38 70 responses at 11. However, there was one exception. Children who at J38 71 seven said they teased other children were more likely to say the J38 72 same again at 11.

J38 73 Prevalence of teasing and fighting: age, sex and ethnic J38 74 differences

J38 75 (a) TEASING

J38 76 Extent and nature: More boys than girls at 11 said J38 77 they had been teased (73 vs 58 per cent). At age seven, it was the J38 78 other way around - more girls than boys - largely due to the fact J38 79 that 84 per cent of the white girls were teased. This figure had J38 80 dropped to 61 per cent at the second interview.

J38 81 We allowed three responses when children were describing how J38 82 children at their school were teased, and two responses when they J38 83 described how they themselves were teased. As is shown in Table 2, J38 84 by far the most frequently mentioned form of teasing was that of J38 85 name-calling and verbal abuse. Very few children mentioned other J38 86 forms of teasing such as physical abuse or being ostracized. More J38 87 black children than white children say that children are teased J38 88 about their clothes and accessories, particularly girls. Half of J38 89 the black girls said this, compared with just under a quarter of J38 90 the white girls. When describing how they themselves are teased, J38 91 more girls than boys mention physical appearance. Thus to give some J38 92 examples of answers to these questions:

J38 93 caption&table

J38 94 "They take the mickey out of their eyes, the way they J38 95 speak, what colour they are and where they come from - take mickey J38 96 out of [their] teeth, when they have gaps like me, and say you have J38 97 big ears." (WG)

J38 98 "Make fun of your nose, even though its<&|>sic! normal J38 99 size. Shape of their body or what they wear, e.g. no name<&|>sic! J38 100 trainers." (BB)

J38 101 This is a similar picture to that found at age seven years, J38 102 with one important difference. Whereas teasing in the form of J38 103 family insults features prominently in the junior school, no J38 104 reference was made to this when the children were infants.

J38 105 Why children think they are teased: Of the children J38 106 who said they were teased, 13 per cent were unable or unwilling to J38 107 say why they thought this was. In Table 3 we list the most common J38 108 responses given as to why children think they themselves are J38 109 teased. These are that the teaser is being provocative, wanting a J38 110 confrontation or even a fight; that they think it is fun to tease J38 111 and enjoy doing so; that the teaser is in some way jealous; that J38 112 their prestige is enhanced; and that what they say is accurate, or J38 113 they dislike the person they are teasing. However, there was no J38 114 significant difference between boys and girls and ethnic groups in J38 115 the responses to this question.

J38 116 table&caption

J38 117 Racial teasing: Although racial name-calling and J38 118 verbal abuse was identified as a means of teasing, as shown in J38 119 Tables 2 and 3, we also asked specific questions to do with this J38 120 type of teasing - e.g. "Do any children get teased because J38 121 of the colour of their skin?". The following answers J38 122 elucidate the situation:

J38 123 "Very much. Call you black this, black that. More J38 124 whites than blacks, so I don't say anything to them. My friend is J38 125 white, but others do it. For example, Paul goes over the limit - J38 126 cusses Hannah's colour and then she gets really angry and then J38 127 families get involved. Paul will get his dad who will have a go at J38 128 Hannah and so it goes on. Sometimes they say they are joking but I J38 129 say it's not a joking matter, shouldn't say things like J38 130 that." (BB)

J38 131 "... it's white people teasing black or other way J38 132 round. All colours tease each other - most of it swearing, some get J38 133 expelled if really bad." (WG)

J38 134 For both age groups, significantly more black children than J38 135 white children say they have been subjected to racist name-calling J38 136 and verbal abuse: for the 11-year-olds, 27 per cent of black J38 137 children, compared with 9 per cent of white children. However, it J38 138 should be remembered that 42 per cent of the sample did not agree J38 139 that children at their school were racially teased, and only 17 per J38 140 cent admitted it happened to them. We shall return to this point in J38 141 the discussion.

J38 142 What children do about being teased: One can see from J38 143 Table 4 that in answer to the question "What do you do when J38 144 you are teased?", retaliation of either a verbal or J38 145 physical nature was the most frequent response. Verbal retaliation J38 146 (e.g. name-calling) was cited more than physical retaliation. J38 147 Reporting the incident to an adult (i.e. "telling the J38 148 teacher") was recorded less often.

J38 149 table&caption

J38 150 Black boys were more likely to say they would retaliate J38 151 verbally and less likely to say they would ignore being teased. J38 152 Although not significant, more girls than boys mentioned physical J38 153 retaliation, but this was often due to an escalation of the J38 154 incident, which had started out with verbal abuse or name-calling, J38 155 for example:

J38 156 "Tell them to shut up. If they keep going on, do it J38 157 back to them and if they still go on, I'll hit them." J38 158 (WG)

J38 159 How children compare themselves with others: J38 160 Significantly more black children than white children say they are J38 161 teased less than others (83 vs 62 per cent). The two most frequent J38 162 responses given as to why they were teased less were that other J38 163 children were afraid of them and what they might do (44 per cent) J38 164 and/or they were popular and well-liked amongst their J38 165 peers, which meant that they wouldn't be teased by them (35 per J38 166 cent). Thus to illustrate:

J38 167 "I'm the tallest in school therefore taken for granted J38 168 that if they say something to me, I'll turn and whack them in the J38 169 head." (BG)

J38 170 "Because the people who do it [tease] like me, they J38 171 mostly do it to people they don't like." (WB)

J38 172 More boys than girls gave fear as the reason (49 vs 39 per J38 173 cent) for being teased less than others, and over half the black J38 174 children in this group compared with a third of white children, J38 175 although this disguises the fact that there were more black girls J38 176 (58 per cent; n = 15) and very few white girls (15 per cent; n = J38 177 3). More white girls in the 'teased less' group gave popularity J38 178 than black girls (60 vs 19 per cent).

J38 179 Being a tormentor: When it comes to being teasers J38 180 themselves, 57 per cent admitted that they sometimes tease other J38 181 children. This was often qualified, usually by saying that they J38 182 only did it when it had been done to them or if they were angry or J38 183 annoyed:

J38 184 "Sometimes if they call me names first, call them same J38 185 names they call me." (BB)

J38 186 "Sometimes when they are really getting on my nerves - J38 187 if they push me and keep on I get annoyed - [say] 'move you spasm' J38 188 - when they stare at you or trip you up, I'll call them a name J38 189 then." (BG)

J38 190 More white boys admitted to teasing than any of the other J38 191 sex/ethnic groups (70 per cent for WB and approx. 50 per cent for J38 192 each of the other three groups).

J38 193 (b) FIGHTING

J38 194 Extent and nature: Of the 104 children (59 per cent) J38 195 who said they were involved in fighting at school, there were J38 196 significantly more boys than girls (74 vs 45 per cent). A similar J38 197 result was reported for the seven-year-olds. White boys (78 per J38 198 cent) got into fights more than black boys (68 per cent). As to the J38 199 reasons why they fought, the most frequently mentioned (75 per J38 200 cent) was that of provocation, either physical or verbal:

J38 201 "They say something about your mum and you don't really J38 202 like it, so it gets into a fight from there. I'm not actually happy J38 203 about fighting, but I'll do it if I have to - sometimes in your J38 204 life you have to do it." (BB)

J38 205 How children feel about fighting: Using the Smiley J38 206 Face Scale, 83 per cent of the sample pointed to faces indicating J38 207 that they hated or disliked fighting. Reasons given for feeling J38 208 like this were that fighting causes distress and/or pain (41 per J38 209 cent); for example:

J38 210 "Because it's horrible - sometimes you end up in tears. J38 211 I'd rather be their friends than fight them. Sometimes it goes on J38 212 for days." (WG)

J38 213 J38 214 J39 1 <#FLOB:J39\>A problem common to all clustering techniques is the J39 2 difficulty in deciding on the most appropriate number of clusters J39 3 to select. The approach used for these data was to plot the numbers J39 4 of clusters against the fusion coefficient to discover a marked J39 5 change in the value of the coefficient (Everitt, 1980). This change J39 6 indicates that two relatively dissimilar clusters have been merged J39 7 and that the number of clusters prior to this merger is likely to J39 8 be the most appropriate. Examination of the coefficients pointed to J39 9 a four or five group cluster solution. This analysis was followed J39 10 by a K-means iterative partitioning technique which, as a cluster J39 11 technique, also produces a partitioning of the cases but differs J39 12 from the hierarchical methods in that it is able to check cluster J39 13 membership and reassign any suboptimally placed children to a J39 14 better fitting cluster. Unlike hierarchical methods, the researcher J39 15 must specify before the analysis the desired number of clusters. In J39 16 this case a four-cluster and a five-cluster solution were J39 17 performed. A split-sample replication of the five-cluster solution J39 18 using the K-means iterative procedure yielded a solution which J39 19 closely approximated that of the whole sample, particularly for the J39 20 children with large discrepancy scores who continued to cluster J39 21 together. Thus the validity of a five-group solution was supported J39 22 on two grounds: similarity of solutions derived from different J39 23 clustering methods and similarity of the solutions obtained for the J39 24 combined sample and for the split-sample.

J39 25 An additional and critical test of the validity of any cluster J39 26 solution can be provided by an examination of variables not J39 27 included in the cluster analysis. In order to examine the external J39 28 validity of the five groups a series of analyses were carried out J39 29 on their performance on a range of cognitive tests. All analyses J39 30 were carried out using SPSSX (Statistical Package for the Social J39 31 Sciences).

J39 32 RESULTS J39 33 From Table 2 it can be seen that Groups 1 and 2 (26 per cent of J39 34 the sample) were reading or spelling above the level predicted by J39 35 both Verbal and Performance IQ. Groups 4 and 5 (63 per cent) were J39 36 achieving considerably below prediction. Table 3 shows the sex, age J39 37 and intellectual performance of groups. Significant differences J39 38 were found between the groups on all these variables. Table 4 J39 39 indicates the performance of the five groups on the reading and J39 40 spelling tests which were similarly found to be significantly J39 41 different.

J39 42 tables&captions

J39 43 Initial identification of the groups was followed by a series J39 44 of analyses comparing performance on the battery of cognitive tests J39 45 with age and IQ (Full-Scale) as covariates, as there were J39 46 significant differences between the groups on these variables. On J39 47 nine of the 20 cognitive measures significant differences were J39 48 found between the groups (see Table 5).

J39 49 A chi-square analysis revealed no significant differences J39 50 between the groups for foot or eye preference nor eye dominance. An J39 51 analysis of the phonetic and visual characteristics of the first 10 J39 52 spelling errors indicated significant differences (see Table 6).

J39 53 Parent and teacher interview

J39 54 The results of a series of chi-square analyses indicated that J39 55 the groups were significantly different on all three variables J39 56 examined in this study (see Table 7). The groups differed on J39 57 whether children had received remedial help and whether there were J39 58 reading or spelling problems reported in the family. In addition J39 59 the number of children who were reading prior to school entry J39 60 differed between the groups.

J39 61 Emotional characteristics

J39 62 Group scores on the anti-social and neurotic subscales of the J39 63 Rutter Children's Behaviour questionnaire (Rutter, 1967) were J39 64 compared by analysis of covariance. Age and IQ were used as J39 65 covariates as the groups differed significantly on these variables. J39 66 No significant differences were found on the anti-social subscales. J39 67 On the Neurotic subscale, cluster Group 3 had a significantly J39 68 higher mean score (ANCOVA, F=3.71, df 4,451; P<0.01), although none J39 69 of the children in this group had a score which would be considered J39 70 diagnostic in a clinical setting according to the cut-offs J39 71 suggested by Rutter.

J39 72 tables&captions

J39 73 Further examination of Group 3

J39 74 It has been argued that dyslexia is not a homogeneous disorder J39 75 and that different groups may be discerned (see, for example, J39 76 Rourke, 1985). In order to look for possible subgroups within Group J39 77 3, the group with the poorest reading and spelling relative to IQ, J39 78 further cluster analyses were carried out using the cognitive and J39 79 emotional variables. This analysis was performed to examine whether J39 80 any further meaningful subdivisions are apparent within this group, J39 81 and whether these subdivisions reflect any particular pattern of J39 82 abilities. Hierarchical cluster analysis suggested a two-group J39 83 solution. A subsequent K-means iterative analysis, with two groups J39 84 being selected, produced one group of 16 and another of 27. Table 8 J39 85 shows the differences between the two groups on the cognitive and J39 86 emotional variables.

J39 87 As can be seen from Table 8, the two groups differed on J39 88 absolute levels of performance on 15 of the 20 cognitive measures J39 89 (75 per cent). The groups were not found to be different on age or J39 90 IQ but differed significantly on all measures of reading and J39 91 spelling as shown in Table 9.

J39 92 table&caption

J39 93 DISCUSSION J39 94 One of the purposes of this study was to examine a large group J39 95 of children to see whether a discrete group(s) with large J39 96 discrepancies between ability and achievement could be identified. J39 97 Using cluster analysis it was possible both to avoid the difficulty J39 98 of having to specify a particular size of discrepancy as being J39 99 significant and to examine concurrently the extent and pattern of J39 100 six discrepancy scores.

J39 101 The issue of heterogeneity in dyslexia was addressed at two J39 102 levels. The first level can be termed an analysis of surface or J39 103 presenting characteristics. This was conducted by examining all the J39 104 groups identified in the analysis of the six discrepancy scores and J39 105 establishing whether different patterns of disability were apparent J39 106 in groups with large discrepancies. The second level was to examine J39 107 whether those identified as a distinct group(s) with large J39 108 discrepancies had different patterns of cognitive abilities from J39 109 the other groups. This approach examines the issue as to whether J39 110 the differences evidenced in discrepancy scores reflect underlying J39 111 cognitive processes.

J39 112 In the initial investigation five distinct groups appeared as a J39 113 result of the cluster analysis: two whose reading achievement was J39 114 at a level higher than that predicted by their intelligence, two J39 115 whose performance was close to prediction and a group whose J39 116 achievement was significantly below their ability. In a similar J39 117 study in the USA Satz and Morris (1981) found two subgroups with J39 118 large discrepancies. These two dyslexic groups were distinguished J39 119 from one another by the extent of their discrepancy scores rather J39 120 than their pattern. In both the Florida Longitudinal Project (FLP) J39 121 and the Linbury study the dyslexic group(s) which emerged showed J39 122 discrepancies in all measures. Thus, in both studies, when cluster J39 123 analysis was applied to reading and spelling in a mixed sample, no J39 124 distinct dyslexic profiles appeared.

J39 125 In the Linbury study 11 per cent of children were identified as J39 126 having large discrepancies while 38 per cent were found by Satz and J39 127 Morris (1981) in the Florida Longitudinal Project. These J39 128 differences are not surprising as in both studies some preselection J39 129 of children had taken place, different variables were placed in the J39 130 cluster analysis, the children in the two studies differed in age J39 131 and both sexes were included in the Linbury study while the FLP J39 132 only considered males. The percentage of participants identified as J39 133 potentially dyslexic in these two studies contrasts with the 4-10 J39 134 per cent found in studies using unselected groups of participants J39 135 (Rutter, 1978).

J39 136 In order to examine whether the group with large discrepancies J39 137 (Group 3) possessed features associated with dyslexia, comparisons J39 138 were made between the groups firstly on reading, spelling and IQ J39 139 tests and then on the other cognitive tests in the study. Group 3 J39 140 produced the lowest performance on all four reading and spelling J39 141 measures, consistent with them being labelled a dyslexic group (see J39 142 Table 4). The groups with the smallest discrepancies, 1 and 2, had J39 143 the highest scores on reading and spelling while Groups 4 and 5 had J39 144 intermediate scores. An examination of age, Verbal and Performance J39 145 IQ revealed significant differences between groups but did not J39 146 indicate any particular tendency for the group with large J39 147 discrepancies (see Table 3).

J39 148 The finding that Group 3 had the poorest performance on reading J39 149 and spelling tests raises the question as to whether this group J39 150 could have been identified simply by selecting those with the J39 151 poorest reading performance. As Group 3 constituted 11 per cent of J39 152 the sample a selection was made of all those children who performed J39 153 at or below the 11th percentile on one measure of achievement, the J39 154 Schonell Reading Test. Sixty per cent of these fell into cluster J39 155 group 3 with the remaining 40 per cent falling into clusters 4 and J39 156 5. A similar analysis was performed on the 'good' readers (Groups 1 J39 157 and 2) to examine whether they represented the top 26 per cent of J39 158 reading. This indicated more than a quarter of the best readers J39 159 were not in Groups 1 and 2 although none was in Group 3. These J39 160 analyses suggest that the cluster analysis applied in this study J39 161 identifies different children from those identified by a simple J39 162 examination of reading performance.

J39 163 A high incidence of reading and spelling difficulties among J39 164 siblings, parents and other family members of dyslexic children has J39 165 long been recognised (Hinshelwood, 1909). Yule and Rutter (1976) J39 166 found that a dyslexic child was three times more likely than a J39 167 non-dyslexic to have a family history of reading problems. The J39 168 reading history of the children showed that Group 3 had the largest J39 169 proportion of family members who reported difficulty with either J39 170 reading or spelling. Groups 4 and 5, however, which had no J39 171 discrepancies between ability and achievement, had a similar J39 172 reported family incidence (see Table 7). These three groups are to J39 173 be contrasted with Groups 1 and 2 which had a lower reported J39 174 incidence. These data were derived from self reports of reading and J39 175 spelling difficulties in the immediate family. Family members were J39 176 not formally assessed and thus the findings must be interpreted J39 177 with caution. In addition to increased reports of familial J39 178 incidence, Group 3 had a higher percentage who had received, or J39 179 were receiving remedial help. Smaller percentages of Groups 4 and 5 J39 180 had received, or were receiving remedial help. None of the children J39 181 in Groups 1 and 2 had received remediation for reading and these J39 182 groups also had the highest percentages of children who were J39 183 reading at school entry (Table 7). These findings are consistent J39 184 with the discrepancies and the overall reading and spelling J39 185 performances apparent in the groups.

J39 186 There is a large body of evidence linking dyslexia with various J39 187 emotional and behavioural problems (see Thomson and Hartley, 1980). J39 188 Teacher ratings and observations suggest that dyslexic children are J39 189 more distractible and dependent than their classmates (McKinney J39 190 et al., 1982). In line with these findings the results J39 191 from the behaviour questionnaire showed that Group 3 had a J39 192 significantly higher Neurotic subscore. However, in view of the J39 193 findings of Rutter (1967) of a higher incidence of neurotic J39 194 disorders in boys, the higher incidence found in Group 3 may J39 195 reflect the sex distribution in the groups.

J39 196 The higher incidence amongst males of reading related problems J39 197 is well established. Critchley (1967), for example, in a review of J39 198 the literature, showed that the percentage of boys in samples of J39 199 children with dyslexia ranged from 66 per cent to 100 per cent. In J39 200 a more recent review of the signs and symptoms linked with J39 201 dyslexia, Wheeler and Watkins (1979) found sex differences of a J39 202 similar order. Other large studies attempting to define dyslexic J39 203 groups have limited their sample to males, presumably in an attempt J39 204 to obtain a higher percentage of reading disabled participants, and J39 205 therefore any direct comparison is precluded (Satz and Morris, J39 206 1981; Van der Vlugt and Satz, 1985). In this study even taking into J39 207 account the number of males in total sample<&|>sic! (58 per cent), J39 208 the percentage of males in Group 3 was extremely high (92 per J39 209 cent).

J39 210 The analyses of group differences on reading, spelling and IQ J39 211 are not independent of the group differentiation as these variables J39 212 were included in the cluster analysis. J39 213 J39 214 J39 215 J39 216 J39 217 J40 1 <#FLOB:J40\>3 Italian Fascism

J40 2 Proto-fascism in Italy

J40 3 If we were concerned simply with reconstructing the evolution J40 4 of Fascism then it might be sensible to start with the inaugural J40 5 meeting of Mussolini's Fasci di combattimento held in J40 6 March 1919 and thus before the full extent of the structural damage J40 7 which the war had inflicted on Italy's liberal institutions had J40 8 become apparent. Our primary concern, however, is to establish how J40 9 forces at work in the political culture of modern Italy exemplify J40 10 the nature of generic fascism. Thus we must go back several years J40 11 earlier to consider the pro-war lobby which formed in the autumn of J40 12 1914 when, despite its conspicuous failings, Italian liberalism J40 13 still seemed unassailable. Some of these 'interventionists' were J40 14 democrats and included Radicals, right-wing Liberals, reformist J40 15 Socialists and several cabinet ministers, as well as the two most J40 16 important members of the government, the prime minister, Salandra, J40 17 and the foreign minister, Sonnino. They all hoped that by fighting J40 18 on the winning side Italy would not only secure territorial gains J40 19 and enhanced international prestige but create a new style of J40 20 dynamic, authoritative parliamentary government, so finally putting J40 21 an end to the rise of revolutionary socialism and the weakness they J40 22 identified with the Giolittian system.

J40 23 What interests us here, however, is the intensive J40 24 extra-parliamentary campaign mounted between August 1914 and the J40 25 'radiant days of May' when Italy formally joined the Entente J40 26 Cordiale. Though highly disparate in their surface ideology J40 27 and in the degree to which they constituted a recognizable J40 28 political grouping, the different elements which joined forces to J40 29 become revolutionary interventionists had one thing in common: the J40 30 belief that entry into the war would inaugurate a new post-liberal J40 31 Italy. In other words, the shared mythic core which made their J40 32 alliance possible was a palingenetic variety of ultra-nationalism, J40 33 so that, according to our ideal type, the pressure group they J40 34 formed can be considered the first significant manifestation of J40 35 fascist politics in Italy.

J40 36 The oldest component had no organizational form as such but was J40 37 essentially a publicistic phenomenon. It nevertheless represented J40 38 an important current in the political counter-culture of pre-war J40 39 Italy. One of its foremost representatives was Papini, who as early J40 40 as 1904 had co-written 'A Nationalist Programme' attacking the J40 41 decadence of liberalism and the divisiveness of socialism. J40 42 Presented as a lecture in several Italian cities that year, the J40 43 programme called for the diffusion of the allegedly aristocratic J40 44 virtues of authority and heroism to revitalize the middle classes J40 45 so as to create a 'world of revived energy' in which not only would J40 46 the arts flourish once more, but patriotic armies and entrepreneurs J40 47 would work together to create a flourishing colonial empire in J40 48 Africa (see Lyttleton, 1973b, p.116-9). Palingenetic myth is not J40 49 only the Leitmotif of Papini's publicistic activity in these J40 50 early years, even dictating the titles of some of his articles, J40 51 'Italy is Reborn', 'Campaign for the Forced Reawakening of Italy' J40 52 (see Papini, 1963), but three decades later it still provided the J40 53 central theme of Italia mia, his most famous work of J40 54 propaganda for the Fascist regime: "Italy's nature is like J40 55 that of the phoenix: cut in two it reconstitutes itself, and hardly J40 56 has it arisen once more than it soars even stronger than J40 57 before" (ibid., p.509). He made important contributions to J40 58 a number of periodicals such as Il Leonardo, Regno, La Voce, J40 59 Lacerba and L'Anima which helped establish the J40 60 respectability of anti-socialist, anti-liberal and J40 61 ultra-nationalist ideas in pre-war Italy.

J40 62 The most influential of these was La Voce, edited by J40 63 Prezzolini who had set up Leonardo with Papini in 1903 and J40 64 co-written with him the Nationalist Programme. The articles and J40 65 editorials of these two men, along with the contributions of the J40 66 poet Soffici, preached an eclectic blend of aesthetic politics J40 67 which drew on Nietzsche and other currents of J40 68 anti-materialist philosophy (Crocean idealism, Bergsonian J40 69 vitalism, as well as conservative nostalgia for a strong unitary J40 70 state). The result was a sustained critique of the mediocrity of J40 71 Italian society and the call for a revolution in Italy, not merely J40 72 cultural but ethical and political which would place the country in J40 73 the hands of a new spiritual elite. As early as 1904 Prezzolini had J40 74 been arguing that the 'old Italy' of corruption and decadence had J40 75 to give way to the 'new' one of energy and heroism (for example J40 76 Prezzolini, 1904). This vision of Italy's rebirth became the J40 77 central creed not only of the Voce circle in Florence J40 78 (1908-16), but of numerous self-appointed Vociani who felt J40 79 they belonged to a new generation destined to complete the J40 80 risorgimento in a political order which was not merely J40 81 post-Giolittian but post-liberal. In 1914 all three men J40 82 threw their publicistic energies into the pro-war campaign. J40 83 Prezzolini turned La Voce into an uncompromisingly J40 84 interventionist magazine, while in the pages of Lacerba Papini J40 85 and Soffici fused avant-garde art with heady visions of an Italy J40 86 regenerated through the war experience (see Gentile, 1972; J40 87 Anderson, 1989).

J40 88 Another component of revolutionary interventionism was closely J40 89 related in its origins to Papinian and Vocian 'cultural J40 90 ultra-nationalism' but operated as a formally constituted political J40 91 pressure group. In December 1910, amid the heightened nationalist J40 92 passions aroused by Austria's expansion into Bosnia in 1908 and the J40 93 imminent prospect of a colonial war with Libya, a Nationalist J40 94 Congress was held in Florence, home of the Voce circle. At J40 95 this congress Corradini, who had co-edited Il Regno with J40 96 Papini in 1904, and his close collaborator Federzoni founded the J40 97 Associazione Nazionale Italiana. The core ideology of the J40 98 ANI's more radical members, spelt out emphatically in its J40 99 periodical, L'Idea Nazionale, perpetuated the Vocian myth J40 100 of the 'new Italy' but also contained two distinctive elements. The J40 101 first, elaborated by Corradini on his return from Latin America in J40 102 1908, was a quasi-Marxist justification of Italian expansionism as J40 103 the act of a 'proletarian' nation asserting its right no longer to J40 104 submit to the hegemony of plutocratic 'capitalist' ones such as J40 105 Britain and Germany. The second was provided by the legal theorist, J40 106 Rocco, who had been recruited by the ANI in 1913. His drive to J40 107 promote a new post-liberal order in Italy was rooted in a J40 108 philosophy of history which focused on the decline of state J40 109 authority allegedly brought about by the diffusion of liberal and J40 110 socialist principles ('demo-socialism') in the wake of the J40 111 French Revolution. The remedy he proposed to the corruption and J40 112 class conflict he saw about him was a peculiar blend of J40 113 technocratic faith with both conservatism and modern imperialism. J40 114 He proposed the replacement of the liberal system by a J40 115 corporativist order in which a powerful industrial class would J40 116 control sectors of the economy under the auspices of a strong state J40 117 in which the authority of the monarchy, the military and the Church J40 118 had been restored. In 1914 L'Idea Nazionale predictably J40 119 turned itself into a major organ of interventionist argument J40 120 (Roberts, 1979, ch.5; De Grand, 1971, 1978).

J40 121 Both Vocian and ANI ultra-nationalism was of a distinctly J40 122 elitist, right-wing complexion and represented a J40 123 transformation of the conservative tradition, albeit in a J40 124 revolutionary rather than a restorationist direction. In this J40 125 respect they can be placed at the opposite end of the spectrum from J40 126 a third component of anti-democratic interventionism, J40 127 neo-syndicalism. This term refers to a number of revolutionary J40 128 socialists who since the turn of the century had, in contrast to J40 129 the 'maximalist' (that is revolutionary Marxist) mainstream, J40 130 increasingly looked to a technologically advanced Italy as the J40 131 precondition for the creation of a heroic proletariat and for the J40 132 subsequent realization of socialism through a network of worker-led J40 133 unions. Though an attempt was made to make this political vision J40 134 the basis of a formal organization with the setting up of the J40 135 Unione Sindacale Italiana in 1912, it remained J40 136 essentially a diffuse current of revolutionary agitation J40 137 represented by some trade-union activists and propagated in J40 138 periodicals (for example Avanguardia socialista and J40 139 Il divenire sociale) and books (for example Labriola, J40 140 1910).

J40 141 The openly palingenetic and increasingly nationalist trend of J40 142 neo-syndicalist thinking predisposed a number of its most J40 143 prominent theorists, notably Lanzillo, De Ambris, Rossoni, J40 144 Corridoni and Panunzio, to campaign alongside their natural J40 145 enemies, the 'right-wing' extra-parliamentary interventionists in J40 146 1914. They believed that "the war would bury for good the J40 147 forms and ideologies of the past and prepare the way for something J40 148 radically new ... Whoever remained passive would be left behind as J40 149 history accelerated and Italy entered a new era" (Roberts, J40 150 1979, p.112). The only vestige of socialism in their belief in the J40 151 revolutionary virtues of national war was the underlying assumption J40 152 that society would one day belong politically and economically to J40 153 the 'producers' rather than to the parasitic ruling, landowning and J40 154 industrial classes who had dominated Italy hitherto. To promote J40 155 their cause they set up a network of pressure groups, the J40 156 Fasci di azione rivoluzionaria (the adjective J40 157 'internazionalista' which originally appeared in their J40 158 name had been quietly dropped), while Corridoni proved his mettle J40 159 as one of the most effective speakers at the mass rallies held in J40 160 Milan to whip up public pressure against neutrality (see Roberts, J40 161 1979, ch.5).

J40 162 A fourth political grouping which promoted interventionism in J40 163 1914 had an even less orthodox pedigree than the neo-syndicalists: J40 164 the Political Futurists. In 1909 their leader, Marinetti, had J40 165 achieved international fame with the publication in Paris of the J40 166 Futurist Manifesto announcing a radical break with all tradition J40 167 (passatismo or 'pastism') in the name of an art which J40 168 would celebrate the dynamism of the modern machine age. In this J40 169 respect 'Futurism' was a child of the same age which gave rise to J40 170 fauvism, expressionism, abstractionism, surrealism and J40 171 constructivism (see Hughes, 1980, pp.40-3). What set it apart from J40 172 such aesthetic movements was that from the beginning the J40 173 inauguration of a 'futurist' age was inextricably bound up with the J40 174 call for national regeneration. The legacy of Italy's imperial, J40 175 religious or cultural past was regarded by futurists as a dead J40 176 weight preventing her from becoming a technologically advanced, J40 177 militarily strong national community. Liberalism, which embodied J40 178 the 'pastist' mentality had to go. The mythic core at the heart of J40 179 such ideas was again unmistakably palingenetic: "Political J40 180 futurism was the irrational and activist commitment to the violent J40 181 destruction of the old world and the creation of a new society J40 182 whose form was as yet ill-defined, in which Marinetti intended to J40 183 have the role of leader and ideologue" (Gentile, 1982, J40 184 p.146).

J40 185 The policies which flowed from this curious form of aesthetic J40 186 politics were distinctly bellicose ("war is the sole J40 187 hygiene of the world" was one of Marinetti's more memorable J40 188 aphorisms). Futurism thus managed to combine chauvinism and J40 189 imperialism with policies in keeping with its J40 190 anti-establishment pose, namely republicanism, J40 191 egalitarianism and anti-clericalism. As a result, Futurists J40 192 such as Boccioni were to be found campaigning against the Church J40 193 and for national expansion in the 1909 elections, and in 1914 were J40 194 eager to support the interventionist cause: indeed, Marinetti J40 195 became one of its most effective spokesmen. In addition, the urge J40 196 to translate avant-garde aesthetics into political change drew J40 197 support from the periodical Lacerba, produced by Papini and J40 198 Soffici in their short-lived Futurist phase (Joll, 1965; Gentile, J40 199 1982, ch.4; Mosse, 1990).

J40 200 Another permutation of aesthetic politics was to be found in J40 201 the interventionist alliance, this time embodied not in a formally J40 202 constituted political group, nor even a current of political J40 203 culture, but in a single ideologue of visionary politics: J40 204 D'Annunzio. His personal discovery of Nietzsche in 1893 had been J40 205 the starting point for a decisive shift from exploring the perverse J40 206 delights of the 'decadent' sensibility (that is in Il J40 207 piacere of 1891) to a self-appointed 'superman'. J40 208 Though initially he felt his role was to resist the rising tide of J40 209 mediocrity unleashed by modern mass society (that is Le J40 210 vergine delle rocce of 1895), the wide-spread food riots of J40 211 1898 left a deep impression on him. From then on he saw himself as J40 212 a seer called upon to use his lyric and dramatic genius to inspire J40 213 patriotic fervour in the masses and bring about "the J40 214 rebirth of Italy" (La Gloria of 1899; Il J40 215 fuoco of 1900) as an heroic, imperialist, modern nation. But J40 216 D'Annunzio was not content to be the poet laureate of nationalism. J40 217 True to his new vision of himself as a synthesis of artist and J40 218 leader, he was drawn irresistibly into the political arena. J40 219 J41 1 <#FLOB:J41\>The only difference between Chicken and Prisoners' J41 2 Dilemma in this format is the transposition of p and s J41 3 for each player, but the games are quite different. Each of them J41 4 contains a puzzle: in the case of Chicken a puzzle about J41 5 precommitment, in the case of Prisoners' Dilemma a puzzle about J41 6 rational actors arriving at a foreseeable but suboptimal J41 7 outcome.

J41 8 These two games are the foundation of modern non-zero-sum game J41 9 theory in social science. This is not the place to review the vast J41 10 literature that has developed from them. It is the place for one J41 11 warning and one encouragement. The warning is that what is true of J41 12 the two-person one-off game is not necessarily true for repeated J41 13 and/or multi-person games, so it is dangerous to make sweeping J41 14 inferences from them. The encouragement is that nevertheless the J41 15 insights to be derived from these very simple cases go a long way J41 16 towards explaining puzzling features of political life in many J41 17 different times and places.

J41 18 The Provision of Public Goods

J41 19 Chicken and Prisoners' Dilemma are examples of collective J41 20 action problems. A collective action problem is any situation in J41 21 which players' individually rational actions lead to an outcome J41 22 which could have been bettered if players had chosen differently. J41 23 To illustrate this most general and hence vague formula, consider a J41 24 problem which has troubled economists for a long time: the optimal J41 25 provision of public goods. A public good is anything which is at J41 26 least to some extent indivisible and/or non-excludable. A J41 27 well-known case is clean air. You cannot take a pound of clean air J41 28 away from the common stock and sell it on a market stall. Nor can J41 29 you exclude anybody from enjoying the benefits of clean air, J41 30 whether or not she has contributed to the cost of producing it.

J41 31 Public goods pose a problem for market economists because they J41 32 cannot be sure that the unregulated market will produce an optimum J41 33 quantity of them. General equilibrium theory builds on the classic J41 34 insight of Adam Smith that the uncoordinated and largely selfish J41 35 choices of individuals may make society, in aggregate, as well off J41 36 as it is possible to be. The modern way of putting this is that J41 37 unregulated perfect competition in all markets at once leads to a J41 38 point on the Pareto frontier. The Pareto frontier is the set of J41 39 points with the property that no further trades can be made. As it J41 40 is assumed that nobody trades unless both (all) parties think they J41 41 would be at least as well off, and some better off, after the trade J41 42 than before, this is an another<&|>sic! way of saying that at the J41 43 Pareto frontier nobody can be made better off without at least one J41 44 person being made worse off.

J41 45 There are two problems with this result. One is that not all J41 46 outcomes which are economically efficient are socially just. The J41 47 other, relevant here, is that markets in which there are public J41 48 goods cannot be perfectly competitive. As you cannot force anybody J41 49 to pay for clean air, nobody will have enough incentive to produce J41 50 it. Producing clean air is the same thing as preventing the J41 51 production of dirty air. Therefore, as in a market equilibrium it J41 52 is impossible to force the polluter to pay, too much pollution will J41 53 be produced.

J41 54 Ensuring that public goods are produced is much the same as J41 55 ensuring that the outcomes players desire in Chicken and Prisoners' J41 56 Dilemma games are produced. Everybody wants clean air but left to J41 57 themselves may fail to produce it. Since the 1920s, economists have J41 58 looked for ways to ensure that public goods are produced, usually J41 59 by recommending that politicians or administrators introduce J41 60 incentives so as to make markets behave as if they were perfect. It J41 61 has taken a long time for academic concern to reach the policy J41 62 arena but current international interest in the 'polluter pays J41 63 principle' provides an example. It is now commonplace for J41 64 economists to argue that permits to pollute, priced at the social J41 65 cost of the pollution, should be sold by governments and tradeable J41 66 amongst would-be polluters, while the proceeds are used to clean up J41 67 the pollution.

J41 68 But there is a catch. If we examine the record of any J41 69 government which professes to believe in market principles, it is J41 70 easy to find actions which are inconsistent with that belief. For J41 71 instance, the Thatcher administrations in the UK rejected available J41 72 moves towards perfect competition in a number of policy arenas. J41 73 Three of them concern motor cars. First, a market-distorting tax J41 74 concession in favour of company cars, inherited from previous J41 75 administrations, was retained (albeit on a reduced scale). This J41 76 imposed not only direct costs arising from the market distortion J41 77 and from lost revenues but also a pollution cost. Motor cars damage J41 78 cities by reason of noise, exhaust pollution, congestion and J41 79 accidents. A system of permits to pollute, in the form of road J41 80 pricing, is technically feasible, works in Singapore and is J41 81 supported by social scientists of a wide range of ideologies, but J41 82 was resolutely rejected by the Thatcher government and also by the J41 83 new model Labour Party under Neil Kinnock. Thirdly, UK car prices J41 84 are held artificially high, even within the supposed free-trade J41 85 area of the European Community, by a combination of requiring 'type J41 86 approval' for car imports and a 'voluntary export restraint' J41 87 agreement with Japanese car makers.

J41 88 Why do politicians disappoint economists so much? Not because J41 89 they are irrational or stupid. I believe that the Thatcher J41 90 administration had perfectly good reasons for rejecting these three J41 91 efficiency-improving changes. Most of the outputs of a political J41 92 system are themselves public goods. A law empowering local J41 93 authorities to charge users for their urban roads is as indivisible J41 94 as clean air. It is not quite non-excludable because some places J41 95 could be exempted from its operation. But in the sense which J41 96 matters, it is not practicably excludable: it is not practicable to J41 97 restrict it to those people who have contributed to the costs of J41 98 its production. Hence the market in government rules is imperfect, J41 99 just as the market in air is imperfect.

J41 100 Another way of putting the same point is as follows. Standard J41 101 economic analysis assumes that most actors in the economy are J41 102 self-interested. In perfect competition, universal self-interest J41 103 would maximize welfare, as if by an invisible hand, no less J41 104 efficiently than universal altruism. But general equilibrium fails J41 105 because the market delivers too few public goods. The same Adam J41 106 Smith who praised the workings of the invisible hand observed:

J41 107 People of the same trade seldom meet together, even for J41 108 merriment and diversion, but the conversation ends in a conspiracy J41 109 against the public, or in some contrivance to raise J41 110 prices.

J41 111 So economists turn to politicians and bureaucrats in the hope J41 112 of getting the market regulated. But it is inconsistent to suppose J41 113 that politicians and bureaucrats are any less self-interested than J41 114 economic actors. In so far as the model presupposes economic actors J41 115 who maximize utility (defined over income, wealth, entitlements and J41 116 status), it must in consistency presuppose political actors who do J41 117 the same thing. In a democracy politicians need enough votes to be J41 118 (re-)elected and enough funds to fight their campaigns and pay for J41 119 the party HQ. Normally, before they can maximize their utility they J41 120 must first be elected. That politicians depend on voters and on J41 121 campaign funds is scarcely news. The distinctive contribution of J41 122 the rational-choice school is to apply game theory to these J41 123 relationships and emerge with predictions about rational J41 124 behaviour.

J41 125 Rational Choice and Voting

J41 126 In the well-known spatial model popularized by Downs, J41 127 politicians who wish to be (re-)elected move to the issue position J41 128 of the median voter. Voters maximize their expected utility by J41 129 voting for the party which maximizes their party differential, J41 130 provided that their party differential discounted by the J41 131 probability that they are pivotal exceeds the cost of voting. Both J41 132 parts of this model have been much discussed. Downs raised an J41 133 obvious problem with the 'rational' voter: on any plausible view, J41 134 the subjective probability that one's own vote is decisive is J41 135 vanishingly small.

J41 136 Therefore it drags the expected utility of voting down to J41 137 fractions of a penny. As the cost of voting and of becoming well J41 138 informed is at least a few pence, it seems that the rational voter J41 139 would never vote. Rational-choice writers have suggested some J41 140 unsatisfactory ways out. Downs suggested that voters might J41 141 nevertheless vote to save democracy from collapse, but this was to J41 142 commit "a common fallacy in social thought which most of J41 143 the book turns on his avoiding": namely, that what was in J41 144 the interests of all was ipso facto in the interests of J41 145 each. Another popular line has been to say that people vote for the J41 146 incidental benefits (meeting their friends at the polls and so on). J41 147 This is a desperate move, invalidated by all empirical evidence. A J41 148 rationally self-interested citizen would almost never J41 149 vote. But Down's analysis is a constructive failure because it J41 150 explains why apparently trivial differences in voting conditions J41 151 (wet weather or dry, Sunday or Thursday, registration at the J41 152 voter's initiative or at the state's) are associated with great J41 153 differences in turnout.

J41 154 Who then are relatively more likely to vote?: those whose party J41 155 differentials and/or subjective probability of decisiveness are J41 156 relatively high and those for whom the cost of voting is relatively J41 157 low. There is no reason for the first two to vary systematically J41 158 with class or income but there is reason to expect the last to do J41 159 so. Voting (where not compulsory) is a form of leisure activity. J41 160 Leisure is probably a luxury good in the economist's sense that the J41 161 demand for it increases more than proportionately with income. J41 162 Therefore there is reason to expect the rich to be more likely to J41 163 vote <}_><-|>that<+|>than<}/> the employed poor. Furthermore, the J41 164 cost of acquiring political information is lower for the educated J41 165 than for the uneducated. As there is a positive correlation between J41 166 education and income, this reinforces the leisure effect. Thus J41 167 there is good reason to expect what empirical studies across time J41 168 and space confirm: that voting turnout rates are higher among the J41 169 rich and educated than the rest of the population. Thus a rational J41 170 politician, seeking to maximize the probability of re-election, J41 171 offers not what the electors want but what the median voter wants. J41 172 The median voter is probably richer and better educated J41 173 <}_><-|>that<+|>than<}/> the median elector. In particular, she J41 174 probably drives a car which in Britain is quite likely to be a J41 175 company one (over half of new car registrations in the UK are of J41 176 company cars). Thus rational choice analysts have an easy J41 177 explanation for the first two of our three observed departures from J41 178 economic rationality, namely the company car tax concession and the J41 179 absence of road pricing.

J41 180 The other half of Downs's model deals with the rational party. J41 181 If there is just one issue dimension, politicians who wish to be J41 182 (re-)elected will converge on the ideological position of the J41 183 median voter. Where politics is truly unidimensional, Downs's J41 184 hypothesis is robust. Every so often, a politician condemns the J41 185 consensus of the existing parties and seeks a "choice, not J41 186 an echo" (the campaign slogan of Barry Goldwater in 1964). J41 187 Goldwater, George McGovern and Michael Foot each led their parties J41 188 to catastrophic defeat after breaking away from the consensus. J41 189 Margaret Thatcher is often cited as a counter-example. However, on J41 190 a number of issues, including race relations, trade union reform J41 191 and nationalistic foreign policy, Thatcherism moved towards, not J41 192 away from, the median voter. On other issues, Thatcherism was J41 193 admittedly unpopular. However, Conservative efforts in 1990 to J41 194 temper the effects of the poll tax and to slow down the J41 195 implementation of NHS reform are evidence for the robustness of the J41 196 Downsian median. So are the strenuous efforts of Neil Kinnock to J41 197 lead the Labour Party to it.

J41 198 When politics is multidimensional, however, the hypothesis gets J41 199 into trouble. Generalizing the median voter hypothesis into J41 200 multidimensional issue space may sound easy. Suppose that one issue J41 201 dimension is 'left-right' judged by voters' attitudes to the J41 202 distribution of income and wealth; another is J41 203 'liberal-conservative' judged by issues such as free speech and the J41 204 death penalty; and a third concerns the regional distribution of J41 205 resources. J41 206 J42 1 <#FLOB:J42\>The growth of investment in plant and machinery J42 2 declined from an annual rate of 23 per cent between 1978 and 1981 J42 3 to 3 per cent between 1981 and 1984. (Singapore, Economic Committee J42 4 1986a: 12). There is, however, evidence of expansion of activities J42 5 involving skilled labour in which Singapore retained a cost J42 6 advantage. This was most clearly the case in electronics, where J42 7 more capital-intensive and sophisticated production has developed. J42 8 In 1984 two chip-making plants were established, one by the J42 9 American-based Texas Instruments and the other by the Italian group J42 10 SGS. This represented a major break in the dominance of the J42 11 electronics sector by assembly work (see chapter 6 for a discussion J42 12 of the significance of these developments in the context of the J42 13 Asian Regional Division of Labour). However, the other major J42 14 development between 1979 and 1982, the rapid growth of computer J42 15 manufacture and the establishment of Singapore as a major exporter J42 16 of disk-drives (Asian Business, December 1981: 46 and J42 17 December 1982: 65), were essentially higher value assembly J42 18 activities. These moves to higher value production characterised J42 19 European and American investment but not Japanese since 1971. J42 20 Japanese investment in Singapore has fallen; it has gone either to J42 21 lower labour cost location or to Europe and the USA. This was a J42 22 major blow to the PAP's strategy. The failure of international J42 23 capital to respond fully to the PAP's 'second industrial J42 24 revolution' reveals the limited extent to which even the most J42 25 powerful Third World state economic apparatus can influence J42 26 international capital.

J42 27 The generally broader base of the economy to a degree insulated J42 28 Singapore from the 1979-82 recession that so seriously affected the J42 29 other NICs. However, sharp falls in commodity prices, including J42 30 oil, slowed the growth of exports (figure 5.2) and made the J42 31 Singapore economy increasingly vulnerable to any major contraction J42 32 in the demand for its manufactured goods and services. This J42 33 vulnerability is underlain by the fact that in 1984 66 per cent of J42 34 goods and services were exported. The other Asian NICs are far less J42 35 dependent on foreign demand - of their total production, Hongkong J42 36 exports 48 per cent, Taiwan 37 per cent and South Korea 26 per cent J42 37 (Singapore, Economic Committee, 1986a: 7).

J42 38 In 1985 all the NICs experienced problems, but Singapore was, J42 39 because of its more exposed position, by far the hardest hit. The J42 40 fall in Singapore's industrial production was much steeper than in J42 41 the other Asian NICs (figure 5.3). The impact of the sharp J42 42 contraction of markets for Singapore's products was reinforced by J42 43 the collapse of the domestic construction industry boom which J42 44 between 1982 and 1985 was a major factor in sustaining growth J42 45 rates. Growth came to a shuddering halt. GDP decreased by 1.8 per J42 46 cent, declines of 8 per cent in manufacturing and 17 per cent in J42 47 construction being partly balanced by a 5 per cent growth in the J42 48 service sector. There were widespread business failures and sharp J42 49 falls in export earnings and, more spectacularly, foreign J42 50 investment (figure 1.1). Unemployment rose from 2.9 per cent in J42 51 1984 to 6 per cent in 1986. In addition the number of Malaysian J42 52 'guest workers' fell from c. 105,000 to c. 60,000 ( Sieh, J42 53 1988: 104).

J42 54 figure&caption

J42 55 The serious impact that the global recession had on Singapore, J42 56 particularly between 1984 and 1986, reveals the vulnerability of J42 57 such an open economy. These experiences resulted in some revision J42 58 of the PAP's development strategies. These centred on the J42 59 recognition that the leading growth area will in future be the J42 60 service sector with manufacturing playing a major but secondary J42 61 role (Singapore, Economic Committee, 1986a). In effect the J42 62 government merely recognised what had been happening since 1980 and J42 63 incorporated it into strategy. Singapore was to be promoted:

J42 64 as a premier Operational Headquarters (OHQS) for J42 65 Manufacturing, Finance and Banking Services, especially risk J42 66 management and fund management, knowledge intensive industries and J42 67 other internationally traded services. (Singapore, Report of the J42 68 Fiscal and Financial Policy Sub-committee, 1986: 1)

J42 69 While the aim of attracting high-value industry was retained, J42 70 the importance of labour-intensive industry was also recognised. To J42 71 this end the continual use of 'guest labour' was accepted. However, J42 72 no major influx of foreign labour is contemplated and given the J42 73 almost zero population growth labour supply is likely to exert a J42 74 major constraint on the economy. Perhaps more significant was the J42 75 realisation that Singapore has, since the late 1970s, lost the J42 76 'competitive edge' in manufacturing.

J42 77 Between 1979 and 1984 total labour costs rose by 10.1 per cent J42 78 while productivity per worker rose by only 4.6 per cent (Cass, J42 79 1985). During this period Singapore's competitive position declined J42 80 by 50 per cent against Hongkong, 15 per cent against Taiwan and 35 J42 81 per cent against South Korea (Singapore, Economic Committee, 1986a: J42 82 31). While rising labour costs were seen as the principal cause of J42 83 this decline, attention has also been drawn to the role of J42 84 increased rents, interest charges, transport costs and the costs of J42 85 power and water in reducing the level of profitability (Singapore, J42 86 Economic Committee, 1986a: 44).

J42 87 A variety of measures has been introduced with the collective J42 88 aims of reducing production costs, raising profitability and J42 89 attracting foreign investment. Taxation on charges for electricity, J42 90 gas and telephone calls was suspended and 50 per cent rebates on J42 91 property taxes granted. Particular attention has focused on the J42 92 reduction of corporation tax from 40 to 33 per cent (Far J42 93 Eastern Economic Review Year Book, 1987: 237). However, J42 94 central to the policy has been the reduction of labour costs, J42 95 through cutting of the employers' contribution to the CPF from 25 J42 96 to 10 per cent, wage freezes and actual reductions. These policies J42 97 are being pursued with the active 'cooperation' of the corporate J42 98 trade unions.

J42 99 It is clear that since 1985 there has been a significant J42 100 retreat from the 'second industrial revolution' strategy (Rodan, J42 101 1987: 172). Whether the revised policies will enable Singapore to J42 102 regain the 'competitive edge' is a moot point. The government's J42 103 ability to reduce labour costs far enough to compete with lower J42 104 cost areas is, in particular, highly debatable. Since 1985 the J42 105 government has attempted to stimulate the economy by increasing J42 106 development expenditure. By 1986 state expenditure as a percentage J42 107 of GDP had risen to 29.3 per cent (table 5.5). Given the reductions J42 108 in tax and CPF contributions this is only sustainable if major J42 109 economies are made in recurrent expenditure and a high level of J42 110 capital inflow is established.

J42 111 caption&table

J42 112 Concern over the level of public expenditure and the need for J42 113 the state to revitalise the economy by promoting new areas of J42 114 activity has, since 1986, led to a systematic policy of J42 115 privatisation. The Public Sector Disinvestment Committee was J42 116 established to formulate a programme of disinvestment in Government J42 117 Linked Companies (GLCs). Some 634 of these concerns together with 7 J42 118 of the 33 Statutory Boards have been considered. However, while J42 119 substantial progress has been made the state has retained a 51 per J42 120 cent controlling interest (Ng Chee Yuen 1989). In addition, J42 121 investment has brought new areas of the economy under central J42 122 government control. Thus assets are being realised in some sectors J42 123 of the economy in order to invest in others. Ng Chee Yuen and J42 124 Wagner (1989: 216) have suggested that there is unlikely to be any J42 125 significant reduction in the involvement of the state in the J42 126 economy; rather the outcome "may be just a change of J42 127 company portfolios and activities; and public enterprise would J42 128 continue to be the instrument of restructuring and J42 129 development".

J42 130 The various policies implemented since 1985 have contributed to J42 131 a restoration of confidence on the part of domestic and, more J42 132 especially, international investors. However, given the highly J42 133 internationalised nature of the Singapore economy - in 1988, 88 per J42 134 cent of total demand was external (Balakrishnan, 1989: 66) - the J42 135 renewed levels of growth that the city state has experienced (table J42 136 1.19) are primarily a reflection of a partial recovery of the J42 137 world-economy. In particular, renewed expansion in the demand for J42 138 electronics resulted in rapid growth in this industry and the J42 139 sectors that underpin it. Indeed it has been the manufacturing J42 140 sector that has been the main element in Singapore's recovery J42 141 rather than finance and business services. However, during 1989 the J42 142 emergence of a world surplus of disk-drives heralded a contraction J42 143 in the electronics sector and a marked slowing of rate of growth of J42 144 Singapore's manufacturing sector and economy as a whole.

J42 145 In addition to Singapore's vulnerability in a period of marked J42 146 instability in the world-economy the totality of government J42 147 policies since 1981 has caused increasingly widespread discontent J42 148 and focused opposition to the PAP. Such opposition may well limit J42 149 government's ability to follow present strategy as well as J42 150 undermining it by reducing the confidence of foreign investors.

J42 151 Thailand J42 152 In many ways Thailand is the regional polar opposite of J42 153 Singapore. While manufactured exports expanded rapidly from the J42 154 early 1970s the country has never made a clean break from ISI. In J42 155 addition until the 1980s the economy was probably the least J42 156 internationalised of the ASEAN groups with a low level of foreign J42 157 investment (tables 1.6 and 1.7) and MNC activity. During the J42 158 immediate post-war period Thailand was probably the most 'open' of J42 159 the South East Asian economies and the one in which capitalist J42 160 production had made least headway. However, these conditions did J42 161 not result in any major influx of foreign activity. The lack of any J42 162 resource not readily available elsewhere in the region, location J42 163 off the main regional and international transport routes, apparent J42 164 instability of the governments and proximity to the former J42 165 Indo-Chinese states combined to discourage foreign undertakings and J42 166 investors. In addition, particularly between 1947 and 1957, the J42 167 nationalistic and often erratic nature of government economy policy J42 168 tended to discourage investors.

J42 169 In the immediate post-World War Two period Thai economic J42 170 development was characterised, like most of South East Asia, by J42 171 intense nationalism. During this period the Chinese population was J42 172 heavily discriminated against. The government of Phibun Songkram J42 173 (1947-57) established a form of state-led capitalist development. J42 174 Under this the interests of Chinese and foreign capital were J42 175 secondary to those of Thai nationals (Skinner, 1958: 317; Ingram, J42 176 1971: 287, 275). Foreign investment and manufacturing activity were J42 177 inhibited by threats of closure, state-promoted competition and J42 178 non-renewal of contracts.

J42 179 During the period 1947-57 development strategy, while never J42 180 clearly stated, had all the characteristics of ISI. However, these J42 181 policies did not centre on tariff protection. Tariffs were treated J42 182 primarily as sources of revenue. In addition, Akrasanee and Juanjai J42 183 (1986: 77) have suggested that the Phibun regime J42 184 "deliberately avoided protecting industries for fear of J42 185 promoting the Chinese community". Rather, industrialisation J42 186 was promoted through the direct participation of the state in J42 187 production.

J42 188 State enterprises were set up with monopolies in such areas as J42 189 brewing, paper manufacture, sugar refining and gunny sack J42 190 production (Marzouk, 1972: 195; Silcock, 1967: 262-3). While some J42 191 were fully state-owned, others were under the control of NEDCOL J42 192 (National Economic Development Corporation Limited). The majority J42 193 of these undertakings were effectively a partnership between the J42 194 state and major Thai and Chinese business interests (Hewison, 1985: J42 195 276; Ingram, 1971: 288). Through NEDCOL the state acted as a J42 196 guarantor of loans enabling comparatively large-scale foreign J42 197 funding to be obtained for many of these enterprises. However, J42 198 direct state investment played a major role in the development of J42 199 manufacturing. The government's share of domestic investment rose J42 200 from 32 per cent in 1952 to over 38 per cent for the period 1953-5 J42 201 (Akrasanee and Juanjai, 1986: 83).

J42 202 Evaluation of the state enterprise programme is difficult. The J42 203 state concerns became bywords for corruption and inefficiency, none J42 204 of them ever reaching target production levels (Ingram, 1971: 287). J42 205 In addition, after the overthrow of Phibun in 1957, the record of J42 206 NEDCOL and the state enterprises became a favourite area of J42 207 propaganda to further discredit his regime and the direct J42 208 involvement of the state in production. However, the state J42 209 enterprises, whatever their failings, did initiate domestic J42 210 production in a number of important areas and set Thailand on the J42 211 ISI road.

J42 212 By 1957 there was mounting opposition to the Phibun development J42 213 strategy. This stemmed from domestic and foreign investors, the J42 214 World Bank and the Bank of Thailand. The protests do not appear to J42 215 have been directed at ISI per se but rather at the J42 216 haphazard way in which the state intervened in the economy. J42 217 J43 1 <#FLOB:J43\>5. International Financial Relations: The Debt J43 2 Crisis

J43 3 Ronald Barston

J43 4 The contemporary international debt crisis which began with the J43 5 Mexican and Brazilian financial collapses in 1982 has proved an J43 6 intractable feature of modern international politics. It is in many J43 7 ways different from earlier debt crises in terms of the number and J43 8 frequency of reschedulings and widespread effects within the J43 9 international system. The current debt crisis has given rise, J43 10 moreover, to complex and innovative multilateral diplomacy J43 11 involving creditor agencies, bank advisory committees, multilateral J43 12 diplomacy involving governments, multilateral institutions and J43 13 non-state actors. The profound complexity and need for mediation as J43 14 well as innovation has also seen the emergence of figures J43 15 previously behind the scenes, such as the Managing Director of the J43 16 IMF and the Director General of the General Agreement on Tariffs J43 17 and Trade (GATT), as players with enhanced and exposed roles in J43 18 international politics.

J43 19 This chapter examines the initial responses to the debt crisis; J43 20 the process of debt negotiation; and the weakening of consensus J43 21 over obligations and management of the crisis. The concluding J43 22 section looks at some of the major international policy issues J43 23 which have arisen in the course of the crisis. The main theme J43 24 explored will be that of showing that, whilst in the short term the J43 25 debt crisis has been successfully contained by innovatory J43 26 procedures and measures, in the longer term consensus has not been J43 27 sustained. The debt crisis has become an intractable and J43 28 increasingly complex feature of the international system.

J43 29 BACKGROUND J43 30 Debt and financial reschedulings of sovereign states have been J43 31 a long-standing feature of the international system. Prior J43 32 to the Second World War external long-term debt consisted primarily J43 33 of bond issues floated abroad by sovereign states and other actors. J43 34 During the nineteenth century there were a number of major defaults J43 35 and rescheduling of foreign bonds. Following the high lending of J43 36 the 1820s for railroads, construction and mining, most Latin J43 37 American governments defaulted to some degree. In the J43 38 mid-nineteenth century Latin America and countries in the Eastern J43 39 Mediterranean (Greece and Turkey) featured as defaulters, whilst J43 40 those after 1880 included Chile (1880-4) and Turkey (1881), whose J43 41 debt was rescheduled after the 1875 default; also notable was the J43 42 major borrowing of Imperial Russia. Brazil, which had largely J43 43 remained in credit, required rescheduling in 1898 and 1914. In the J43 44 period up to 1918, defaults followed in the wake of the Mexican J43 45 revolution and the Russian revolution; Turkey suspended its debt J43 46 payments to the allies in 1914. Subsequently, in the 1930s, most of J43 47 Latin America, Eastern Europe, Turkey and China defaulted, added to J43 48 which there was the breakdown of German reparations in Western J43 49 Europe.

J43 50 In the postwar period governments became more directly involved J43 51 in international debt rescheduling operations. In part this stemmed J43 52 from the growing general governmental involvement in a wide range J43 53 of external economic security policy. The spread of the Cold War J43 54 beyond Europe and the attention paid to the economic component of J43 55 security and military alliance are typified by the US-led Marshall J43 56 Plan, to a lesser degree bilateral Soviet treaties and Council for J43 57 Mutual Economic Association (COMECON) aid, along with regional J43 58 economic arrangements such as the Colombo plan. The decolonization J43 59 process, which intensified in the 1960s, was frequently accompanied J43 60 by the continuation of post-independence economic links J43 61 with the colonial power, at least in the early years, or through J43 62 the formation of aid donor groups for individual countries.

J43 63 The role of multilateral institutions in financial aid and debt J43 64 questions is a particularly notable feature of postwar J43 65 international economic relations. Within the United Nations system, J43 66 the creation of the International Finance Corporation in 1960 J43 67 completed the so-called IMF/World Bank group. The setting up of the J43 68 United Nations Conference on Trade and Development (UNCTAD) in 1964 J43 69 brought increasingly onto the central agenda of international J43 70 politics the broad issues of trade, development assistance and J43 71 technology transfer, raising the level of political attention given J43 72 to less developed country demands for greater trade access, framed J43 73 subsequently in the concept of a New International Economic Order. J43 74 Within the framework of the IMF, developing countries formed the J43 75 Group of 24 in 1971, partly as a counterweight to the Group of 10 J43 76 industrialized countries, with the aim of articulating developing J43 77 country positions on access to and use of IMF resources. The impact J43 78 of the G-24 remained relatively limited until the late 1970s, when J43 79 more cohesive developing country strategies began to emerge in the J43 80 context of the North-South dialogue and second oil crisis.

J43 81 EARLY DEBT RESCHEDULING

J43 82 Before the debt crisis began in 1982, postwar debt J43 83 reschedulings of official debt incurred directly by or guaranteed J43 84 by governments had been relatively few and generally for amounts of J43 85 less than US$ 300m. In the commercial sector no institutional J43 86 machinery as such existed internationally for dealing with J43 87 commercial debt. Debt in the official category was largely J43 88 negotiated within official creditor groups, for example for Poland, J43 89 Yugoslavia and Indonesia. The bulk of official creditor J43 90 transactions have been conducted within the framework of the J43 91 so-called Paris Club of OECD countries, at which the IMF, World J43 92 Bank and other agencies such as UNCTAD are represented along with J43 93 the applicant country and creditor governments. The Paris Club, J43 94 which was initially set up in 1956 to deal with Argentine debt, J43 95 normally meets under the chairmanship of a representative of the J43 96 French Treasury.

J43 97 Debt problems, at least until the late 1970s, had been kept in J43 98 manageable proportions. In the postwar period states have used a J43 99 variety of different strategies to meet their external funding J43 100 requirements. Liberia, for example, has relied heavily on official J43 101 assistance from the United States and Federal Republic of Germany J43 102 and, after joining the IMF in 1962, had eight IMF standby J43 103 arrangements up to 1970. In Ghana's case, almost 80 per cent of J43 104 external debt was in the form of suppliers' credits from Western J43 105 and socialist countries whilst major profit funding, for example J43 106 the Volta River project, came from the World Bank, United Kingdom J43 107 and the United States. The overthrow of Nkrumah in 1966, partly J43 108 because of the dramatic rise in debt repayment to 25 per cent of J43 109 foreign exchange earnings, resulted in a reorientation of strategy, J43 110 including cancellation of trade agreements with the People's J43 111 Republic of China and resumption of relations with the IMF. In J43 112 Europe, Turkey relied heavily on loans and deferments under the US J43 113 Marshall Plan, along with other funding from the World Bank and J43 114 European Investment Bank, while Indonesia offset reductions in J43 115 Western funding by shifting to the Soviet Union and withdrawing J43 116 from the IMF, a policy later reversed under President Suharto. In J43 117 contrast to the above examples, Mexico, with an essentially J43 118 petroleum-based economy, has relied on access to foreign capital J43 119 markets for bank credits and used bond issues to meet its J43 120 financing, largely on the assumption that, given political backing J43 121 from its US patron such access would continue, making it more easy J43 122 to roll over these forms of financial instruments.

J43 123 THE 1982 DEBT CRISIS

J43 124 The 1982 debt crisis broke quite suddenly, against the J43 125 background of Poland's economic difficulties, with first the J43 126 Mexican and then Brazilian collapses. The debt crisis spread to J43 127 take in other parts of Eastern Europe, much of Africa and parts of J43 128 Southeast Asia. While the factors for the payments difficulties of J43 129 individual countries differed, several broad causes can be noted. J43 130 In the first are those associated with public policy failures, such J43 131 as ill-designed high cost-low return development projects. As J43 132 regards external factors, oil importing developing country J43 133 economies were badly hit by the 'shocks' of two sets of oil price J43 134 rises, in 1973 and 1979. Following the oil price rise of 1979, J43 135 developed countries in the main responded with restrictive economic J43 136 policies, which tended to result in reductions in demand for J43 137 developing country exports, already affected by deteriorating terms J43 138 of trade. The more credit-worthy middle income less developed J43 139 countries (LDCs) were able to contain the shock of oil price rises J43 140 only by borrowing against the security of their growth potential as J43 141 newly industrialized countries (NICs). South Korea, for example, J43 142 continued as part of its debt security policy to borrow heavily on J43 143 the commercial capital markets. Indeed the role of commercial bank J43 144 lending in the lead-up to the 1982 crisis and subsequently was J43 145 crucial.

J43 146 Financial innovation facilitated the growth of large-scale J43 147 financial syndication. The development too of Euro-market J43 148 facilities added to the available sources of funds, particularly J43 149 for the offshore banking and parastatal agencies of debtor J43 150 combines, such as Brazil. In banking terms, the business of lending J43 151 to the Third World was highly profitable. The real debt of non-oil J43 152 LDCs rose from US$169 billion in 1973 to US$294.7 billion in J43 153 1979.

J43 154 INITIAL RESPONSES TO THE DEBT CRISIS

J43 155 In the absence of any formal machinery for dealing with the J43 156 debt crisis, the initial responses of necessity took the form of J43 157 ad hocrescue packages, or what has been called the 'fire J43 158 brigade' approach. In the main, such operations involved stop-gap J43 159 financial support, selectively orchestrated to a large extent by J43 160 the United States, together with, in some cases, short-term J43 161 balance of payments finance via bridging loans, arranged mainly J43 162 through the Basle-based central bank organization, the Bank for J43 163 International Settlements (BIS). In October 1982, for example, the J43 164 US Treasury provided a US$1.23 billion 90-day loan to Brazil, J43 165 though this was not formally announced until President Reagan's J43 166 visit to Brazil in December 1982. The US$1.23 billion loan was also J43 167 supported by a trade package, allowing for the relaxation of J43 168 controls on Brazilian sugar exports to the United States and the J43 169 continuation of Brazilian subsidies on steel exports for a further J43 170 two years. In addition, Brazil secured US$600 million bridging J43 171 finance from its six major bank creditors. Further short-term J43 172 finance of US$1.2 billion was provided by the BIS in December 1982, J43 173 pending attempts to agree a financial rescue package involving the J43 174 IMF and commercial banks.

J43 175 The Mexican and Brazilian crises highlighted what were to J43 176 become four central problems in the management of the debt J43 177 question: the need to mobilize internationally very large amounts J43 178 of finance on a recurring basis; the complexity of the negotiating J43 179 process owing to the number of secondary banks and other agencies; J43 180 the need to coordinate the respective involvement of the IMF and J43 181 commercial banks, and the inadequacy of IMF resources to meet the J43 182 financial requirements of debtor countries over and above balance J43 183 of payments financing.

J43 184 The experience of the Mexican and Brazilian debt negotiations J43 185 laid the basis for the subsequent development of the IMF's J43 186 coordinating role. Thus, by mid-December 1982, Brazil had reached J43 187 substantial agreement with the IMF for a Fund-supported programme, J43 188 which was put to the Brazilian bank creditor group meeting in New J43 189 York on 20 December 1982, attended by some 125 bank representatives J43 190 and the IMF's then Managing Director, Jacques de J43 191 Larosi<*_>e-grave<*/>re. Nevertheless the position remained J43 192 precarious, dependent on bridging operations, the maintenance of J43 193 interbank lines and the mobilization of large amounts of commercial J43 194 and international institution funding. Subsequently the ideas J43 195 underlying the IMF's approach to assembling financial packages J43 196 became based on what was known as the 'critical mass' doctrine. In J43 197 essence, the doctrine, which shaped IMF policy until its J43 198 modification in 1986, required commercial bank commitments to have J43 199 reached a critical amount, normally over 90 per cent of that J43 200 required, before IMF funds would be committed.

J43 201 Following the Mexican and Brazilian crises, the debt position J43 202 of a number of other developing countries also worsened J43 203 substantially. In 1983-4, 23 countries sought debt relief within J43 204 the framework of the Paris Club. Normally the Paris Club requires J43 205 an applicant country, if it is an IMF member, to have in place or J43 206 under discussion an economic adjustment programme supported by Fund J43 207 resources subject to upper credit tranche conditionality. The terms J43 208 and conditions of the agreement take the form of an agreed Minute, J43 209 which takes effect when bilateral agreements have subsequently been J43 210 negotiated between the debtor country and each of its creditors. J43 211 The acute nature and frequency of reschedulings in 1983-4 almost J43 212 inevitably meant some relaxation of the Paris Club principle of J43 213 limiting rescheduling to medium- and long-term debt, with the J43 214 inclusion of some short-term, generally trade-related, debt with J43 215 the overall agreement.

J43 216 Apart from the Paris Club agreements referred to above, 32 J43 217 restructuring agreements were reached in principle by 26 countries J43 218 through bank advisory committees, which were an important J43 219 innovative feature of the debt crisis, during 1983-4. J43 220 J44 1 <#FLOB:J44\>As a consequence of this market-driven production J44 2 philosophy, firms must also take on new organizational structures J44 3 such as the 'flat organization' or 'polycentricism'. The case J44 4 studies of Sony and Honda which follow are ample illustrations of J44 5 this, and an even better case would be IBM which, consciously or J44 6 unconsciously, acts as a role model for the Japanese corporations J44 7 (Morris and Imrie 1991).

J44 8 GLOBAL LOCALIZATION: TWO CASE STUDIES

J44 9 One problem with the study of Japanese overseas investment is J44 10 that it tends to treat all Japanese corporations as an amorphous J44 11 mass. As Cusumano (1985) has illustrated, however, even firms in J44 12 the same industrial sector - in this case Toyota and Nissan - can J44 13 display very different business strategies. The two case studies J44 14 described here are in some ways untypical, as they are at the J44 15 leading edge of global localization among Japanese companies. J44 16 Nevertheless, it is arguable that these firms display J44 17 characteristics in their internationalization strategies which J44 18 other Japanese corporations are starting to adopt.

J44 19 Sony J44 20 Sony have been at the forefront of the globalization of the J44 21 Japanese consumer electronics industry; they were the first company J44 22 to start production in North America when they located a plant at J44 23 San Diego in California in the early 1970s and their Bridgend plant J44 24 in South Wales was the first Japanese television plant in the EC J44 25 (Morris 1987; Morris and Imrie 1991).

J44 26 Essentially, Sony have divided their production and market into J44 27 three major supranational trading blocs - Japan and the western J44 28 Pacific rim (Japan, East and South-East Asia, and Australasia), J44 29 North America (including Mexico), and Western Europe. This has also J44 30 been driven by Sony's export orientation. Indeed they are one of J44 31 Japan's most international companies with only 34 per cent of total J44 32 sales in 1988 in Japan (Wagstyl and Buchan 1989). Accordingly, J44 33 production is being reorganized on an integrated scale in these J44 34 blocs:

J44 35 (a) Japan, East and South-East Asia, Australasia. J44 36 There has been a considerable shift of production of consumer J44 37 electronic products and components from Japan to other south and J44 38 east Asian countries. This is especially true of lower value added J44 39 and mature products such as consumer electronics, as opposed to the J44 40 new product markets into which Sony has been diversifying such as J44 41 semiconductors, computer workstations, computer disk drives and J44 42 high-definition televisions. As part of this strategy, for example, J44 43 in 1989 Sony announced the location of a plant in Singapore to J44 44 produce tubes for colour television assembly factories in Malaysia J44 45 and Thailand which will replace production from Japan.

J44 46 (b) North America. The San Diego facility has been J44 47 considerably expanded since its inception in the 1970s to produce J44 48 higher volumes of its initial product, colour televisions, but also J44 49 to produce a diversified product range. Computer workstations, for J44 50 example, are the latest product to be added to the plant. In J44 51 addition the company has set up a R&D centre at San Jose in J44 52 California, to service the North American operation.

J44 53 (c) Western Europe. Sony's extensive European-wide J44 54 complex of plants is rivalled only by that of Matsushita Electric. J44 55 Moreover, this is set to expand further in the 1990s. The hubs of J44 56 the operation are the Bridgend plant in South Wales, the largest J44 57 plant and the only one to produce colour television picture tubes, J44 58 and the Stuttgart operation in West Germany where, in addition to a J44 59 large plant producing colour televisions, the European CTV J44 60 headquarters are located. The company now has eight Western J44 61 European plants spread across the UK, France, West Germany, Spain, J44 62 Italy, and Austria (see Table 1.1).

J44 63 caption&table

J44 64 While there is a degree of duplication between plants, J44 65 production is also strongly integrated. The Bridgend plant, for J44 66 example, supplies picture tubes to the West German and Spanish J44 67 plants. Similarly, the Colmar plant makes components and J44 68 sub-assemblies for other European plants. Sony's expansion in J44 69 Western Europe extends beyond the number of plants. It includes J44 70 diversification of product lines and global localization through J44 71 increased local sourcing. On some products, for example, local J44 72 content levels are close to 100 per cent.

J44 73 This localization policy is being achieved in three ways. J44 74 First, Sony have made extensive use of the growing number of J44 75 Japanese suppliers who have located in the EC and in the UK in J44 76 particular (Morris and Imrie 1991). Morris and Imrie estimate that J44 77 there are at least fifteen such suppliers currently operating. J44 78 Second, through transferring production of key and other components J44 79 from Japan to Europe (CTV tubes would be one such example). J44 80 Moreover in the 1990s Sony plan to transfer the production of J44 81 semiconductors, video cassette recorder heads, optical pick ups and J44 82 magnetic tape coating (de Jonquieres and Dixon 1989). Third, Sony J44 83 are upgrading local suppliers through their supplier development J44 84 programme in the EC and attempting to localize component sourcing J44 85 to areas proximate to their main plants in order to fully operate a J44 86 just-in-time system. In South Wales, for example, the corporation J44 87 is aiming to source 80 per cent of its local content within a five J44 88 mile radius or one hour's drive time (Morris and Imrie 1991).

J44 89 Sony's global localization strategy in Western Europe, J44 90 therefore, comprises transferring production of a wide variety of J44 91 products and sourcing high percentages of content locally. Its J44 92 product range is set to diversify considerably further in the 1990s J44 93 to mobile communications, computers, medical electronics, robotics J44 94 and security systems, plus the key components already mentioned (de J44 95 Jonquieres and Dixon 1989).

J44 96 In addition to transferring production, Sony is transferring J44 97 some of its R&D capacity to the EC. Bridgend is already the R&D J44 98 centre for CTVs, and the company plan two telecommunications R&D J44 99 centres, in the UK and West Germany and a high definition (HDTV) J44 100 research centre in West Germany. Research for HDTV in North America J44 101 will be centred at the San Jose facility. The commitment to this J44 102 policy of global <}_><-|>localizaton<+|>localization<}/> is J44 103 repeated in the autonomy given to international managers. In 1989, J44 104 for example, Sony Corporation appointed a European and an American J44 105 to its main board, the first Japanese company to do so. This is J44 106 part of a strategy of a largely self-sufficient industrial and J44 107 management infrastructure with substantial freedom to run its own J44 108 affairs.

J44 109 Honda J44 110 Honda's place in Japan's motor vehicle industry is very similar J44 111 to that of Sony's in the electronics field; it was a late entrant J44 112 into the market, is considerably smaller than the two major J44 113 companies and has used its internationalization strategy as a J44 114 competitive advantage - given its relative weakness in its domestic J44 115 market.

J44 116 Honda's initial production overseas dates to the late 1960s J44 117 when it took a share in a domestic Malaysian producer. It now has a J44 118 network of production facilities throughout south, south-east and J44 119 east Asia, in Taiwan (1983), Malaysia (1978), Indonesia (1975 and J44 120 1978), Thailand (1984), and India. Honda holds a minority share in J44 121 all seven plants which produce motorcycles, vehicles, and vehicle J44 122 parts. It also has a plant in New Zealand.

J44 123 Despite Honda's entry into the Asian market via local J44 124 production, it is in North America that the company's global J44 125 localization efforts are concentrated, although the company will J44 126 follow this with major production facilities in the UK. Honda's J44 127 first, and main plant at Marysville, Ohio, is part of Honda's North J44 128 American-wide production complex which comprises eight J44 129 manufacturing operations (see Table 1.2).

J44 130 caption&table

J44 131 Car production started at Marysville in 1982, following J44 132 motorcycle production, and as such Honda were the first Japanese J44 133 company to build a manufacturing operation. Moreover, the company J44 134 were also the first Japanese company to build engines in the USA at J44 135 the Anna facility and the first to produce automatic transmissions. J44 136 The company is the most integrated of the Japanese producers in J44 137 North American automobile production; they were the first to export J44 138 vehicles outside of North America, have the largest R&D and J44 139 engineering facilities, and the highest local content of the J44 140 Japanese transplants in the USA. The company is now the fourth J44 141 largest vehicle producer in the USA after the 'Big Three' domestic J44 142 producers (Economist Intelligence Unit 1989).

J44 143 Currently Honda of America is undergoing a five step strategy J44 144 to become self-reliant in the USA, including: boosting exports; J44 145 increasing local content to 75 per cent by 1991; expanding J44 146 production engineering; developing the second US assembly plant at J44 147 East Liberty, and increasing R&D activities. The Ohio and J44 148 California R&D centres have already doubled in size and will J44 149 increase employment further from 100 to 500.

J44 150 Honda, therefore, have gone through a rapid expansion of J44 151 capacity in their various integrated North American facilities. The J44 152 Marysville plant, for example, started with a planned capacity of J44 153 150,000 cars per annum which has subsequently been upgraded to J44 154 500,000 units. In addition, significant investment has taken place J44 155 at the facility to upgrade it from an assembly to a full J44 156 manufacturing operation, with plastic moulding, stamping, welding, J44 157 painting, and sub-assembly facilities. The plant supplies J44 158 components to Honda's plants in Ontario and at East Liberty, Ohio. J44 159 The two main omissions are engines and automatic transmissions, J44 160 which will be built elsewhere. Car engine production started at the J44 161 Anna plant in Ohio in 1985, which has undergone reinvestment to J44 162 increase capacity to 500,000 engines, plus an iron casting facility J44 163 for front and rear suspension components, and aluminium cylinder J44 164 head and aluminium wheel production facilities. Two further car J44 165 assembly units have started production - one at East Liberty, Ohio, J44 166 which will produce a capacity of 150,000 cars per year, and a plant J44 167 at Alliston in Ontario which has a capacity of 80,000 cars per J44 168 annum.

J44 169 As Florida and Kenney illustrate in Chapter 5, Honda, therefore J44 170 have pushed forward a strategy of global localization of car J44 171 production in North America in little over one decade. This has J44 172 included siting R&D and engineering facilities, in transferring J44 173 full manufacturing operations, and in increasing local content J44 174 through transferring the production of components from Japan to J44 175 Honda operations in North America. Local content has also been J44 176 increased, however, by increasing component sourcing from locally J44 177 based suppliers. Local content levels at Honda in North America J44 178 will be 75 per cent by 1991. Indeed, as Florida and Kenney note, it J44 179 is claimed that by the mid 1990s it will be as high as that at J44 180 Chrysler. This has been achieved by being the most 'Americanized' J44 181 of the Japanese transplants. For example, 95 per cent of the steel J44 182 for fabricated components is from the US and 50 per cent of machine J44 183 tools are from US-based plants.

J44 184 The company has three subsidiary component suppliers: Bellemar J44 185 Parts Industries produces seats and exhaust systems from plants J44 186 on-site at Alliston and Marysville; KTH Parts Industries J44 187 produces stamped parts in St. Paris, Ohio, and CALMAC produces air J44 188 conditioning units at Cerritos, California. Other supplier firms J44 189 are spread throughout North America in places as far flung as J44 190 Florida, Massachusetts, and Canada. However, as Florida and J44 191 Kenney's chapter notes, the vast majority of the 220 suppliers are J44 192 found in the states in the auto-corridor from Michigan to J44 193 Tennessee. This is particularly true of the forty Japanese-owned J44 194 suppliers to Honda, all of which are within a one-hundred-mile J44 195 radius of Marysville. Indeed, over a half are in Ohio with a marked J44 196 concentration around the Ohio Interstate Highway 75 which has been J44 197 dubbed 'parts supplier alley' (Economist Intelligence Unit 1989). J44 198 These tend to be dedicated to the Honda plant, while the J44 199 out-of-state suppliers tend to serve more than one transplant. The J44 200 spatial clustering is unsurprising given the just-in-time supply J44 201 system in operation at Marysville, with two hour stocks and certain J44 202 suppliers delivering to the plant on an hourly basis.

J44 203 Honda's global localization strategy is best illustrated, J44 204 therefore, by the North American example. In Western Europe the J44 205 company's activities have been far more limited although this will J44 206 change in the 1990s. In the late 1970s Honda entered into a J44 207 manufacturing and technical agreement with British Leyland (now J44 208 Rover) in the UK. Rover produced Honda cars under licence at one of J44 209 their UK plants which later developed into joint development work. J44 210 This culminated in 1989 with Honda taking a 20 per cent share of J44 211 Rover's car business and Rover taking a 20 per cent share of J44 212 Honda's UK operations. Honda's plans in the UK include an engine J44 213 plant at Swindon which is already operational, and an assembly J44 214 plant, also at Swindon, which will produce 100,000 cars per annum J44 215 by 1994. J44 216 J45 1 <#FLOB:J45\>Proposition 3. Suppose <*_>lambda<*/> J45 2 varies across activities but is constant for any given activity. J45 3 Then, if formula, the equilibrium wage w* is J45 4 a decreasing function of <*_>lambda<*/>. That is, workers with the J45 5 highest damage potential (lowest <*_>lambda<*/>) receive the J45 6 highest wage.

J45 7 Note: This proposition depends formally on <*_>lambda<*/> J45 8 being a constant for any given activity and may not hold if J45 9 <*_>lambda<*/> varies with e and L.

J45 10 Proposition 4. If <*_>lambda<*/> is a constant for a J45 11 given activity, then the production function f{L,e} can be J45 12 expressed as a generalized Solow function of the following type; J45 13 formula for some g.

J45 14 Proof. See Appendix.

J45 15 The intuitive meaning of our result is as follows. Each J45 16 activity has a unique <*_>lambda<*/>. If activities are J45 17 heterogeneous, they may have different values of <*_>lambda<*/>, in J45 18 which case workers with the highest damage potential will always J45 19 receive the highest wages. Activities that are less susceptible to J45 20 damage will always offer lower wages. By restricting ourselves to J45 21 this special class of production functions, we arrive at a positive J45 22 monotonic relationship between the wage rate and the damage J45 23 potential. In other production functions, this relationship may not J45 24 be positive or even monotonic.

J45 25 Let us examine the non-monotonic case briefly. Consider two J45 26 activities, A and B, of which the former is the more susceptible to J45 27 damage from a low supply of effort. A has two strategy options. The J45 28 first is to pay a higher wage than B's to preclude workers from J45 29 reducing effort and causing damage. The second is to offer a wage J45 30 rate lower than B's and put up with some damage due to worker J45 31 demotivation. Under certain conditions on the production function, J45 32 each of these strategies could be equally profitable. By J45 33 restricting ourselves to a specific class of production functions, J45 34 we ensure that lower wages will not fully offset the damage from J45 35 worker demotivation and preclude such a multiplicity of options.

J45 36 Proposition 5 (The Solow Region). Let w>0 and J45 37 e(w) be three times differentiable, such that, e(w), J45 38 e'(w)>0 and e''(w)<0. Then: (i) there is at most one J45 39 wo such that formula; and (ii) J45 40 formula for formula, for some <*_>mu<*/>>0.

J45 41 Proof. We have

J45 42 formula.

J45 43 (i) Suppose that wo is such that J45 44 formula. Then,

J45 45 formula.

J45 46 It follows that wo is unique from Theorem A1 in J45 47 the Appendix.

J45 48 (ii) This follows directly from the continuity of J45 49 E'w. <*_>Square<*/>

J45 50 Proposition 5 implies that, if it exists, the Solow wage is J45 51 unique. In the region of this wage, the elasticity of effort is a J45 52 decreasing function of the wage. Hence within this region it is J45 53 always true that higher wages are paid to workers with higher J45 54 damage potential (provided <*_>lambda<*/> is constant for each J45 55 activity).

J45 56 III. PROPERTIES OF THE WAGE DISTRIBUTION

J45 57 We have just derived the wage rate as an increasing function of J45 58 the damage potential of workers. Let us now try to interpret the J45 59 nature of the wage differentials generated by the model. A J45 60 simplistic way of interpreting the wage differentials in the model J45 61 is in terms of heterogeneous labour. Suppose that skills are J45 62 variable across workers. Suppose also that there are differences in J45 63 the skill requirements of different activities. Assume that firms J45 64 that have more complex equipment require the more skilled workers. J45 65 Then, if skills and jobs do not match, firms with relatively costly J45 66 equipment will be susceptible to a greater degree of damage. In J45 67 this case, firms that are more susceptible to damage will pay a J45 68 higher wage as a strategy to recruit the more skilled workers.

J45 69 However, the existence of wage differentials for heterogeneous J45 70 labour is not very interesting from either a theoretical or J45 71 a<&|>sic! empirical point of view. The main theoretical puzzle, as J45 72 mentioned earlier, is to explain why wage differentials exist even J45 73 when workers have identical skills. In fact, this question becomes J45 74 especially important in the context of a number of empirical J45 75 studies which indicate the existence of wage differentials for J45 76 apparently homogeneous labour. Let us briefly summarize the J45 77 essential aspects of this literature. Dickens and Katz (1987), for J45 78 instance, have examined the nature of wage differences across J45 79 industries for both union and non-union workers using American J45 80 data. Their findings indicate that there are large differences in J45 81 wages for seemingly similar workers. Substantial wage differences J45 82 persist even when union status and standard individual J45 83 characteristics are controlled. Similarly, Krueger and Summers's J45 84 studies (1987, 1988) indicate that workers' industry affiliations J45 85 exert a substantial impact on their wages, even after controlling J45 86 for human capital variables and a variety of job characteristics. J45 87 There is, however, some controversy in the literature on the J45 88 importance of true 'industry effects' as an explanation of the J45 89 observed wage dispersion. Murphy and Topel (1987), for instance, J45 90 argue that actual wage differentials are more likely to arise from J45 91 the unobserved heterogeneity in workers' abilities. However, J45 92 Gibbons and Katz (1989) dispute the excessive emphasis placed by J45 93 Murphy and Topel on unmeasured ability to explain actual wage J45 94 differentials: they argue that the empirical evidence can be J45 95 consistent with both the 'industry effect' and the 'unmeasured J45 96 ability' explanations of wage dispersion. They argue that, in J45 97 practice, the wage changes of workers shifting to other industries J45 98 for exogenous reasons (such as plant closures) seem to be J45 99 consistent with the 'industry effect' explanation of wage J45 100 dispersion. In contrast, the wage changes of workers shifting J45 101 because of endogenous reasons (such as voluntary quits) seem to fit J45 102 in better with the 'unmeasured ability' explanation of wage J45 103 dispersion. In other words, the actually observed wage J45 104 differentials are obviously generated by a variety of different J45 105 factors.

J45 106 Let us now see how our efficiency wage model helps to provide J45 107 an understanding of the existence of wage differentials for labour J45 108 of homogeneous potentiality. We had earlier introduced the concept J45 109 of 'damage potential'. We shall now make use of this concept for J45 110 providing the rationale for various aspects of wage dispersion. J45 111 There are two possible ways of interpreting damage potential. We J45 112 shall call them the 'shirking' and 'performance' interpretations. J45 113 Let us clarify this further by making a conceptual distinction J45 114 between the two. We define shirking to be the conscious act of J45 115 reducing effort. Efficiency wage models normally assume that J45 116 workers shirk when not perfectly monitored. Since perfect J45 117 monitoring is difficult, penalties such as unemployment are J45 118 required to prevent workers from shirking. Performance, in J45 119 contrast, is not completely dependent upon monitoring, even though J45 120 it is definitely influenced by it. We define performance to J45 121 encompass a wide array of attributes which determine the J45 122 effectiveness of work. For instance, performance can depend upon J45 123 how intensely workers concentrate on their jobs; mistakes are bound J45 124 to happen if workers concentrate insufficiently. Performance can J45 125 also depend upon factors such as the willingness of workers to take J45 126 initiatives and function flexibly. We can intuitively visualize the J45 127 difference between shirking and performance as follows. If workers J45 128 are carefully watched over, they cannot idle on the job; that is, J45 129 shirking can be prevented by perfect monitoring. Nevertheless, this J45 130 by itself cannot guarantee good performance. Workers can either J45 131 inadvertently mishandle equipment or be inflexible and not take the J45 132 proper initiatives. That is, there is a discretionary element to J45 133 the effectiveness of work which is difficult to control by pure J45 134 monitoring.

J45 135 The standard rationalization of wage dispersion relies on the J45 136 strike threat or shirking model. If firms are heterogeneous and J45 137 monitoring imperfect, the uneven impact of the damage potential J45 138 from shirking can generate a wage distribution for homogeneous J45 139 labour. Firms more susceptible to damage will offer a higher wage J45 140 rate. We have just proved the formal conditions necessary for this J45 141 result to hold. If monitoring also happens to be relatively more J45 142 difficult in firms that have a greater susceptibility to damage, J45 143 then workers in such firms will receive wages that are J45 144 disproportionately higher in relation to their effort levels. That J45 145 is, in the shirking model of wage dispersion, some workers can get J45 146 higher wages by the fortuitous fact of being in firms particularly J45 147 susceptible to damage. This case corresponds closely to the J45 148 'industry effect' explanation of wage dispersion. Therefore under J45 149 these circumstances all workers will obviously prefer to be in the J45 150 high-paying firms. Suppose that workers acquired perfect J45 151 information about the wage distribution. Then there will be J45 152 queueing for the high-paying jobs. However, workers do not have any J45 153 credible way of offering to work effectively at these jobs for a J45 154 lower wage. Hence under these circumstances the high-paying jobs J45 155 will be rationed to the workers.

J45 156 We shall now provide an interpretation of wage dispersion which J45 157 relies on the concept of performance. We shall abstract from J45 158 shirking as a potential cause of damage. We ignore the shirking J45 159 problem by assuming that workers are costlessly and perfectly J45 160 monitored. Let workers be of homogeneous potentiality: that is, all J45 161 workers can be assumed to have the same intrinsic capabilities. J45 162 However, the effectiveness of their work will vary depending upon J45 163 the effort levels that they are willing to commit. Let contracts J45 164 with regard to effort be fully specified and costlessly and J45 165 perfectly enforced; in other words, assume that there is perfect J45 166 information regarding all aspects of the production process. J45 167 Therefore, the only source of potential damage for firms comes from J45 168 bad performance - as defined earlier. Firms with costly and complex J45 169 equipment require a high level of performance. They cannot tolerate J45 170 mistakes or the lack of proper initiatives from the workers. Hence J45 171 in this context they will specify contracts that require a high J45 172 level of effort. Workers will be paid a proportionately higher wage J45 173 to compensate them for the higher levels of effort. Firms less J45 174 susceptible to damage do not require the same high standards of J45 175 performance. They can tolerate less effective workers. Hence they J45 176 specify contracts paying a lower wage. That is, the heterogeneity J45 177 in the required performance standards of different firms can also J45 178 provide a theoretical rationale for the existence of a wage J45 179 distribution for labour of homogeneous potentiality. However, the J45 180 implications of the performance interpretation for wage J45 181 differentials is different from that of the shirking J45 182 interpretation. The wage dispersion generated by the performance J45 183 interpretation is compatible with a market-clearing equilibrium J45 184 under perfect information, in which neither the employers nor the J45 185 workers have any incentive to deviate from the equilibrium. As long J45 186 as contracts are implemented, employers will be satisfied with the J45 187 given wage distribution. Let us consider the workers' point of J45 188 view. We shall assume that there is no intrinsic job satisfaction J45 189 attached to any activity. Work involves disutility. Since by J45 190 assumption workers possess perfect information and there is J45 191 market-clearing, they will be indifferent between the different J45 192 occupations. To shift to a higher paid job involves costs in the J45 193 form of more effort-intense concentration, willingness to accept J45 194 inconvenience or exercise initiative. The allocation of jobs can J45 195 then be determined on a random basis.

J45 196 The normative implications of the performance model are also J45 197 quite different from those of the shirking model. In the shirking J45 198 model wage differentials can be considered to be normatively J45 199 'unjust', because some workers get higher wages solely as a J45 200 consequence of their industry affiliation. In contrast, wage J45 201 differentials are normatively 'just' in the performance model J45 202 because some workers are putting in more effort than others. In J45 203 this sense, there is obviously a correspondence between the J45 204 performance interpretation of wage differentials and the J45 205 'unmeasured ability' explanation of wage dispersion. The latter J45 206 explanation considers wage differentials to be justified on account J45 207 of the intrinsic differences in the unmeasured abilities of J45 208 workers. The performance interpretation, however, focuses on J45 209 differences in the levels of effort rather than ability. J45 210 Obviously, in any concrete situation all these various factors may J45 211 have some influence in generating wage differentials.

J45 212 The performance-based interpretation of wage dispersion, J45 213 incidentally, may help to explain some empirical aspects of J45 214 inter-industry wage dispersion which are difficult to reconcile J45 215 with the shirking model. In a recent paper, Akerlof and Yellen J45 216 (1988) have argued that certain aspects of inter-industry wage J45 217 dispersion cannot be rationalized by optimizing versions of the J45 218 efficiency wage hypothesis. They concede that the shirking model J45 219 may explain the wage dispersion, for instance, of machine operators J45 220 across industries on the basis of differences in monitoring costs. J45 221 However, they argue that it is difficult to explain the wage J45 222 differences, for instance, of secretaries across industries with J45 223 the same model. J45 224 J46 1 <#FLOB:J46\>Expenditures on health comprise a very small J46 2 proportion of local government expenditure in both France and the J46 3 UK since this is the responsibility of central government. By J46 4 contrast, sub-national governments account for over 50% of total J46 5 health expenditure in Australia, Canada, Denmark and Switzerland. J46 6 The UK and North America stand out for the extent to which control J46 7 over educational spending is shifted towards lower tiers of J46 8 government. Expenditure decentralization in the UK and other J46 9 unitary states in Europe is rather low by comparison with federal J46 10 countries in North America and Europe as well as Scandinavia. In J46 11 Scandinavia total sub-national expenditure is high because of J46 12 responsibility for a substantial fraction of social welfare J46 13 spending.

J46 14 3. Fiscal decentralization in theory

J46 15 Much of the relevant literature on 'fiscal federalism' has J46 16 originated in North America, and reflects the particular government J46 17 structures of the US and Canada. Here we organize the methods and J46 18 results of this literature in the context of the very diverse J46 19 arrangements observed in European countries.

J46 20 Decentralization performs two distinct functions. First, local J46 21 communities' decisions about public services may better reflect J46 22 local residents' preferences. Second, administrative J46 23 decentralization may improve the information needed to implement J46 24 centrally-determined policies.

J46 25 3.1 The decentralization of choice

J46 26 Decentralization allows communities to better adapt to their J46 27 preferences the levels of local services and taxes. Local public J46 28 goods display the 'public good' characteristics of non-rivalry in J46 29 consumption and non-excludability, but benefit only a limited J46 30 geographical area (Cornes and Sandler, 1986). Voting provides the J46 31 most obvious mechanism by which differences in individual J46 32 preferences might be reflected in collective decisions. J46 33 Unfortunately, it turns out to be difficult to devise clear J46 34 criteria for identifying efficient levels of decentralization when J46 35 voting is the mechanism by which preferences are expressed.

J46 36 One rather limited case where decentralization is unambiguously J46 37 desirable is when decisions are made by groups that each contain J46 38 individuals with identical preferences and circumstances. Where J46 39 there is a greater degree of heterogeneity in the population, few J46 40 patterns of decentralization result in an unambiguous improvement J46 41 (in the Pareto sense). The desirability of decentralized J46 42 decision-making hinges on the balance between the interests of J46 43 different individuals, not all of whom will necessarily gain from J46 44 greater decentralization.

J46 45 3.2. Choice through mobility

J46 46 Whilst it is highly unlikely that the population can be J46 47 subdivided geographically into groups consisting of wholly J46 48 identical individuals, circumstances can be envisaged where the J46 49 population might choose to sort themselves in this way. This is the J46 50 so-called Tiebout model (1956). Faced with a range of local J46 51 authorities offering the full spectrum of possible levels of J46 52 provision, individuals choose a particular level of provision by J46 53 moving to the area where it is on offer. Under certain conditions J46 54 this process achieves a Pareto-improvement.

J46 55 Three groups of assumptions are needed for the Tiebout model to J46 56 be achieved in practice: (a) there should be enough jurisdictions J46 57 offering a sufficient range of levels of public services. In J46 58 addition, there should be enough individuals to ensure that public J46 59 goods can be provided efficiently in each of the jurisdictions; (b) J46 60 individual mobility should be costless, and individuals should move J46 61 in response only to fiscal conditions, and not to other J46 62 characteristics of the various localities; and (c) public services J46 63 in each jurisdiction should be financed by lump-sum taxes, J46 64 set equal to the cost of providing public services to the marginal J46 65 group member. This assumption can be relaxed to allow for local J46 66 services provided by a property tax only under very stringent J46 67 conditions, in which zoning regulations specifying the type of J46 68 property that can be built in each area make the property tax in J46 69 effect a lump sum (Hamilton, 1976).

J46 70 The first assumption requires a multiplicity of small J46 71 government units and that the functions assigned to local J46 72 government do not involve substantial economies of scale. Whilst J46 73 many European countries have small local government units in rural J46 74 areas (where fiscal reasons are unlikely to be a significant factor J46 75 in location choices), local government in urban areas tends to J46 76 involve rather larger units offering a correspondingly restricted J46 77 range of choices. In contrast in many US cities a large number of J46 78 small local jurisdictions may indeed provide the required range of J46 79 fiscal choices.

J46 80 When the first two groups of assumptions are not met, the J46 81 Tiebout model can quickly lead to gainers and losers. In J46 82 particular, if not all individuals are freely mobile, an J46 83 improvement in mobility of some would not necessarily result in an J46 84 improvement in the situation of all. Those unable to move may be J46 85 disadvantaged by an influx of residents with preferences sharply J46 86 different from their own. This is perhaps one of the most J46 87 significant limitations of the Tiebout model. Policies designed to J46 88 make the system of local government conform more closely to the J46 89 Tiebout assumptions (for example, by increasing the number of J46 90 government units, and removing obstacles to migration) will result J46 91 in both gainers (those who move to areas more closely approximating J46 92 their preferences) and losers (those who are unable to move, who J46 93 may find that local levels of provision move further away from the J46 94 standard that they would prefer). Piecemeal improvements to bring J46 95 the system closer to the Tiebout assumptions may sometimes reduce J46 96 welfare. Furthermore, population sorting leads to homogeneous J46 97 communities. This outcome is sometimes seen as undesirable, either J46 98 for basic reasons of social philosophy, or because there are J46 99 efficiency benefits from mixed communities.

J46 100 The Tiebout model is of very limited interest in most European J46 101 countries, and proposals for structural reform that aim to move J46 102 closer to the Tiebout assumptions will not necessarily improve J46 103 matters. On the other side, mobility may place constraints on the J46 104 structure and decisions of local government, as we will see J46 105 later.

J46 106 3.3. Decentralization of implementation

J46 107 A fundamentally different perspective is provided by 'agency' J46 108 models of local government, in which local governments administer J46 109 some of the policy functions of central government. Here the J46 110 objective of decentralization is to exploit two main advantages of J46 111 smaller-scale units: better information and better control.

J46 112 3.3.1. Better information about local needs is J46 113 available to local bureaucrats. This allows more precise targeting. J46 114 In principle, of course, information available at a decentralized J46 115 level could be communicated to central decision-makers, but the J46 116 costs of information processing appear to rise sharply with the J46 117 size of organizations.

J46 118 3.3.2. Better control over administrative 'slack' and J46 119 under-performance follows from the information advantages of local J46 120 voters. Delegated decision-making leads to familiar J46 121 'principal-agent' problems of control, since those charged with J46 122 implementing central policies may choose to pursue their own J46 123 objectives (comfortable conditions of service, or 'empire J46 124 building') at the expense of the objectives set by the central J46 125 government. Decentralization to local governments, rather than J46 126 simply to administrative agencies of central government, allows the J46 127 central government to make use of local voters as a control J46 128 mechanism to constrain bureaucratic under-performance.

J46 129 The role of voting is not any more to express individual J46 130 preferences, but to provide control in a situation of common J46 131 preferences and priorities. This presupposes that the interests and J46 132 preferences of local voters broadly coincide with the preferences J46 133 and objectives of central governments. Differences in the J46 134 objectives of central government and local voters can undermine the J46 135 value (from the point of view of central government) of elections J46 136 as a control device to reduce bureaucratic slack.

J46 137 3.4. The assignment of policy functions

J46 138 Taken together, the three views of the role of sub-central J46 139 governments presented above do not provide simple assignment rules J46 140 for policy functions. A range of other considerations are J46 141 involved.

J46 142 3.4.1. Interjurisdictional externalities. As with any J46 143 other externalities, a number of possible approaches could be J46 144 adopted to ensuring that the interests of those affected by J46 145 interjurisdictional externalities are taken into account. J46 146 Bargaining between the affected parties is likely to result in J46 147 efficient outcomes where the numbers of parties is<&|>sic! small. J46 148 Fiscal instruments can be designed to operate like Pigovian taxes. J46 149 A third approach is to move the assignment to a higher level of J46 150 government, so that the interjurisdictional externality is J46 151 internalized (Olson, 1969). In that case, one needs to balance the J46 152 gains with the cost of service provision. Whilst a higher tier J46 153 assignment may allow economies of scale to be exploited through J46 154 specialization and a more efficient division of labour, the costs J46 155 of communication and control rise more than proportionately with J46 156 the size of the government unit (Helm and Smith, 1987).

J46 157 3.4.2. Macroeconomic stabilization. For much of the J46 158 post-war period there has been a general consensus that the J46 159 appropriate location of macro-economic stabilization is the J46 160 national level. Because they have greater spillover effects J46 161 (leakages), uncoordinated local stabilization policies would be J46 162 less effective. Coordination would at the very least introduce new J46 163 costs, and at worst could be thoroughly undermined by the J46 164 incentives to 'free ride'. In recent years, indeed, increasing J46 165 international economic integration has begun to undermine the J46 166 consensus that even the national level of government is J46 167 appropriate. Some of the debate over the transfer of macroeconomic J46 168 'sovereignty' under the European Monetary System and the Delors J46 169 Plan has been conducted precisely with a view to internalize J46 170 spillovers among nation states (MacDougall Report, 1977; J46 171 Padoa-Schioppa, 1987).

J46 172 3.4.3. Income redistribution. Two main reasons argue J46 173 for assigning the distributional functions of government to the J46 174 central level and to confine the role of local governments to those J46 175 policies that do not have a significant redistributive impact J46 176 (Oates, 1972). First, when income levels differ between areas, a J46 177 given reduction in overall inequality can be achieved more J46 178 efficiently if redistribution occurs between individuals than J46 179 between different areas. Second, population mobility restricts the J46 180 ability of local governments to make independent choices about J46 181 redistributional priorities. Any jurisdiction attempting to operate J46 182 a more redistributive policy than its neighbours will tend to J46 183 attract individuals who benefit from it, and to be less attractive J46 184 to those who lose. The ability of the area to sustain any degree of J46 185 redistribution above its neighbours is reduced, or, where mobility J46 186 is high, eliminated altogether.

J46 187 In practice, in most European countries, this does not need to J46 188 be an all-or-nothing choice, and the range of functions operated by J46 189 state and local governments does include many with some J46 190 redistributive effect. Where mobility is low, the fiscal J46 191 externalities from independent redistributive policies may be weak, J46 192 and other considerations may be more relevant in determining J46 193 assignment. In addition, it is possible - albeit at a cost - to J46 194 devise financing arrangements for local government that largely J46 195 offset the budgetary consequences of fiscally-induced migration. J46 196 Economies of scale, informational requirements for efficiently J46 197 targeting redistributional policies towards individuals, and the J46 198 interaction between government unit size and the effectiveness of J46 199 local elections as a control mechanism, are other key issues in J46 200 determining the level of government. Helm and Smith (1987) argue J46 201 that the appropriate assignment for the implementation of J46 202 redistributive policies may contrast sharply with the conventional J46 203 wisdom that redistribution should be left to higher tiers of J46 204 government.

J46 205 4. Assignment of tax instruments

J46 206 4.1. Principles

J46 207 Much as for spending, tax assignment hinges on economies of J46 208 scale, the extent of interjurisdictional spillovers and population J46 209 mobility. Although US evidence suggests that there could be J46 210 substantial economies of scale in tax administration, its J46 211 importance can easily be overstated. It is generally possible to J46 212 assign the administration of taxes to a higher tier of government J46 213 whilst allowing lower levels of government to choose the tax rates J46 214 and other taxation parameters. Separating tax policy and tax J46 215 administration may give rise to certain problems with achieving J46 216 appropriate incentives for enforcement efficiency, but it seems to J46 217 work reasonably well in many countries.

J46 218 Interjurisdictional spillovers in taxation raise the J46 219 possibility of 'tax exporting' - levying taxes on non-residents to J46 220 finance the services supplied to residents. Taxes on business J46 221 activity and sales generally suffer from this deficiency, and are J46 222 particularly inappropriate where government units are small, and J46 223 the degree of economic integration high. In many countries taxes on J46 224 business constitute a significant revenue source for local J46 225 governments. However, as economic integration has increased within J46 226 Western Europe, pressure for the reform of local business taxes has J46 227 grown, with local control over business taxes being effectively J46 228 ended in the UK and Germany, and active debate over the future of J46 229 local business taxation in France and North America.

J46 230 Population mobility places some constraints. J46 231 J47 1 <#FLOB:J47\>VOCATIONAL QUALIFICATIONS IN BRITAIN AND J47 2 EUROPE: THEORY AND PRACTICE

J47 3 by S J Prais

J47 4 This Note considers three questions bearing on the reform of J47 5 vocational qualifications in Britain, against the background of J47 6 changes being introduced by the National Council for Vocational J47 7 Qualifications. First, in what important respects did Britain need J47 8 a reformed and centrally-standardised system of vocational J47 9 qualifications? Secondly, what are the proper criteria for choosing J47 10 between alternative methods of awarding qualifications? Much that J47 11 is at issue hinges on the relative importance of externally-marked J47 12 written tests as compared with practical tasks assessed J47 13 by an instructor; the discussion and conclusions reached here in J47 14 relation to vocational testing apply in large measure also to J47 15 current debates in other contexts, such as the proper role of J47 16 teacher-assessed coursework in school examinations at 16+ (GCSE) J47 17 and the official teacher-assessment of pupils at age 7 (SATs) J47 18 currently being administered in British schools for the first time. J47 19 Our third question is: in what significant ways do Continental J47 20 systems of awarding qualifications differ from those now proposed J47 21 for Britain?

J47 22 Need for a standardised system

J47 23 It is now accepted on all sides that Britain needs more of its J47 24 workforce to be vocationally trained to intermediate levels; that J47 25 is to say, to craft or technician standards as represented, for J47 26 example, by City and Guilds examinations (at part 2) or BTEC J47 27 National Certificates and Diplomas. In engineering, building and J47 28 related trades there has for long been a system for the award of J47 29 qualifications that has worked more or less satisfactorily; indeed, J47 30 the City and Guilds system established at the end of the last J47 31 century was in many ways an internationally admired pioneer, and J47 32 its syllabuses and examinations were followed, and are still J47 33 followed, in many parts of the world. In other occupations, such as J47 34 office work or retailing, a variety of qualifying bodies grew up in J47 35 Britain - such as the Royal Society of Arts, the London Chamber of J47 36 Commerce and Industry, Pitman's, the Institute of Drapers - which J47 37 developed (what has been called) a 'jungle' of qualifications at a J47 38 variety of unco-ordinated levels. In many other occupations in J47 39 Britain there was no system of qualifications at all.

J47 40 On the other hand, in Germany - but also, for example, in J47 41 France, Austria, Switzerland and the Netherlands - vocational J47 42 qualifications and associated part-time or full-time courses were J47 43 developed which covered virtually the whole range of occupations in J47 44 the economy. The qualifications awarded usually at ages 18-20 at J47 45 the end of these vocational courses - the J47 46 Berufsschulabschluss in Germany and the J47 47 Certificat d'aptitude professionelle in France J47 48 - are as widely understood as, say, O-level passes were recently in J47 49 Britain (the narrower and clearer range of attainments encompassed J47 50 by an O-level pass make it a more appropriate standard of J47 51 comparison than the new GCSE, with the very wide range of J47 52 attainments spanned by its awards).

J47 53 What was essentially wrong in Britain with engineering and J47 54 building qualifications was that too few people took them - but I J47 55 believe there was nothing fundamentally wrong with the J47 56 qualification-procedure itself. For the rest of the economy there J47 57 was a serious need (a) to make the system coherent, so that J47 58 equivalent levels could more easily be recognised; and (b) to J47 59 expand the occupational coverage. These two objectives - greater J47 60 recognisability and expansion of coverage - are of course to some J47 61 extent linked. Greater recognisability should lead to greater J47 62 marketability, reduced transaction costs in the labour market, and J47 63 to greater demand for qualifications and skills both by employers J47 64 and by trainees. The benefits to be expected are similar to those J47 65 ensuing from 'hallmarking'. There are also economies of scale in J47 66 organising training programmes, and in specifying standards and J47 67 certification-procedures for a limited number of defined J47 68 training-occupations at defined levels. Something is of course lost J47 69 in standardising and restricting the number of training-occupations J47 70 and levels: just as something is lost in not having a suit made to J47 71 measure; but, it hardly needs saying, manufacturing to standardised J47 72 sizes enables many more to buy a decent suit.

J47 73 Criteria for vocational qualifications

J47 74 There has always been debate on the relative roles of theory J47 75 and practice in general education; that debate has been at least as J47 76 vigorous in relation to vocational education and the award of J47 77 vocational qualifications. The unsatisfactory extremes of relying J47 78 solely on 'time-serving' or solely on 'pencil-and-paper' tests have J47 79 often been contrasted as the basis for the award of vocational J47 80 qualifications.

J47 81 In general, it is clear that all procedures for the award of J47 82 qualifications can provide no more than imperfect indicators of J47 83 future capability. Before describing how qualifications are awarded J47 84 in practice, let us for a moment consider the issues in an entirely J47 85 theoretical way, with the aid of some basic statistical J47 86 mathematics. Suppose we wish to estimate the capability of a J47 87 person, not simply in relation to what he has done so far, but in J47 88 relation to what he is likely to be able to do in the future under J47 89 similar, but not identical, circumstances to those encountered in J47 90 the past; for example, the quality of materials may alter, designs J47 91 may alter, the type of person under whom (or with whom) he will be J47 92 working may alter. Let <*_>xi<*/> denote his true, but yet J47 93 unobserved, capability in the future; and let x denote his J47 94 performance as measured by some test-procedure based on his past J47 95 performance in specimen tasks. Without affecting the argument, J47 96 these can both be considered as multi-dimensional - relating, for J47 97 example, to speed of work, accuracy of work, cleanliness, etc. In J47 98 choosing amongst alternative test-procedures we have to accept, as J47 99 said, that none will be wholly accurate; and we have also to accept J47 100 that testing is an expensive process, and only limited resources J47 101 can be devoted to it.

J47 102 The expected total discrepancy between test and actual J47 103 performance can be divided into two components. In statisticians' J47 104 terms they correspond to bias and variance; in educationists' terms J47 105 they correspond, respectively, to Validity and Reliability. In J47 106 detail - when choosing between alternative estimators, or between J47 107 alternative test-procedures, we wish to minimise:-

J47 108 (a) the bias: that is, in a sufficiently large number of J47 109 repeated applications we would like the expected value to J47 110 correspond to the true value, that is, we wish to minimise

J47 111 E{x}-<*_>xi<*/>;

J47 112 and

J47 113 (b) the variance: as between alternative test-procedures which J47 114 were equally satisfactory from the point of view of their bias, we J47 115 choose the one that has the minimum variability in repeated J47 116 applications (whether by different examiners, or on different J47 117 samples of questions or tasks); that is, we choose the alternative J47 118 which yields the minimum value of

J47 119 E{(x-E{x})}2.

J47 120 These two components contribute to the total discrepancy J47 121 between test and actual performance as follows:-

J47 122 formula

J47 123 i.e. Total mean-square-error =Variance +(Bias)2

J47 124 =Reliability +(Validity)2.

J47 125 In contrasting written and practical testing of vocational J47 126 capability, it is widely agreed that written tests have greater J47 127 Reliability in the sense that different external examiners would J47 128 give much the same marks if they independently examined a group of J47 129 candidates. On the other hand, it is argued by those of 'modern' J47 130 views, written tests have a lower Validity since they are applied J47 131 under 'artificial' examination conditions, and do not test what the J47 132 candidate actually does in the course of his work; on that view, J47 133 the greatest Validity attaches to the assessment of practical tasks J47 134 carried out by the candidate in a workplace environment, preferably J47 135 in the course of his normal work and assessed by his normal J47 136 workplace supervisor. Any lower Reliability of such procedures J47 137 resulting from the supervisor knowing his own trainee or for any J47 138 other reason, it has sometimes incautiously been suggested, is of J47 139 no consequence.

J47 140 The view in favour of giving great weight to written testing J47 141 can perhaps be summarised as follows. First, any argument that J47 142 bases itself on the notion that Validity (ie lack of bias) is all J47 143 that matters, is essentially wrong. We need to be concerned with J47 144 the total expected error associated with a qualification-procedure J47 145 (ie Validity plus Reliability); we are likely to be misled if we J47 146 focus on only one component. Secondly, if in reality there was a J47 147 relation such that procedures of high Reliability had low Validity, J47 148 and vice versa, then that relation needs careful empirical J47 149 research. The relation is likely to vary from one occupation to J47 150 another; for example, it is likely to depend on the relative J47 151 importance in each occupation of applied craft tasks and of J47 152 planning tasks. Thirdly, we have to take into account the costs of J47 153 testing. One simple rule seems to hold very widely, namely, that J47 154 pencil-and-paper tests are quicker and cheaper to administer than J47 155 assessing practical tasks; consequently, written tests method can J47 156 examine a much wider range of activities per unit of resources J47 157 devoted to certification.

J47 158 A simple example may not be out of place. A carpenter or J47 159 mechanical fitter needs to know which type of metal screw to choose J47 160 for each job; screws come in a myriad of different lengths, J47 161 diameters, threads, heads (flat, round, Phillips, etc.); and in J47 162 different metals (brass, steel, chrome, etc.). If a final J47 163 assessment of capabilities had to wait till the candidate had used J47 164 each type in the course of his normal work in front of his J47 165 supervisor, and had done so properly on a sufficient proportion of J47 166 repeated occasions, it would take a very, very long time for him to J47 167 be judged as qualified. On the other hand, a few specimen written J47 168 questions, such as:-

J47 169 Which kind of screw would you use for fitting a mirror to a J47 170 bathroom wall, and why?

J47 171 would only take a few minutes. By testing that the candidate J47 172 knows why he is doing what he is doing, and not merely observing J47 173 that he is doing it correctly, we attain greater confidence that he J47 174 can operate under the variety of different circumstances that arise J47 175 in practice. Notice also that he needs to be tested not only on J47 176 which is the right type of screw but, if that type is not readily J47 177 available, he needs to know which of the available alternatives are J47 178 acceptable, even if less than ideal; and he also needs to know J47 179 which are not acceptable, even if the customer would not J47 180 immediately twig. The case for testing knowledge, and not merely J47 181 observing practice, is thus a strong one even for the simplest of J47 182 tasks.

J47 183 But let us return to the costs of testing. Inevitably only a J47 184 sample of relevant knowledge and skills can be tested. Otherwise J47 185 not only would the direct costs of examination become excessive, J47 186 but so would the indirect costs; as HMI recently noted, the new NVQ J47 187 assessment procedures have already "encroached on the time J47 188 available for teaching and learning". This was said in J47 189 relation to engineering qualifications; but HMI were probably also J47 190 influenced by the example they quoted of the hairdressing NVQ which J47 191 involves a "1000 task checklist".

J47 192 Clearly, the greater the number of test-items, the greater can J47 193 be our confidence in the final verdict - but also the greater is J47 194 the cost. By the familiar statistical rule, a doubling in the J47 195 required precision requires a quadrupling in the number of J47 196 test-items. This rule applies if the observations are independent; J47 197 if they are correlated and, so to speak, to some extent they test J47 198 the same capability in another way, then more than a quadrupling J47 199 will be necessary (it may not even be possible to double the J47 200 precision after a certain point). In other words, the way the first J47 201 ten questions or ten tasks are dealt with by the candidate tells us J47 202 a great deal about him; if we aimed to double our confidence in our J47 203 judgement we are likely to require many more than forty questions J47 204 or tasks. Further, because practical tests are so very much more J47 205 expensive than written tests, both in administration and marking, J47 206 it is efficient when working within a limited budget to allocate J47 207 more questions to written tests than to practical tests. A complex J47 208 balancing exercise is thus involved in the economic design of J47 209 test-procedures; it is not surprising that in reality they are J47 210 developed only slowly over the years, often with step-by-step J47 211 experimentation, and require an intimate knowledge of the details J47 212 of each occupation. J47 213 J48 1 <#FLOB:J48\>Like the council, the house was more than willing to do J48 2 so, and the request was usually granted pro forma (though J48 3 the house would eventually require that the petitioner offer J48 4 evidence that he or she was worth less than pounds10 per annum). In J48 5 most cases, the house simply moved the court in question to waive J48 6 its customary fees and admit the petitioner to writ process and J48 7 representation free of charge.

J48 8 Another common request of this kind was for the house to J48 9 override claims of privilege which had protected defendants from J48 10 legal action. Where the defendant was a member of the house, as J48 11 often happened, the house normally informed the individual of the J48 12 complaint and solicited his response. Almost invariably the member J48 13 agreed to waive privilege and allow the matter to be heard. There J48 14 was a sense in all of these cases that the Lords expected as much J48 15 from fellow members. Where the member was not forthcoming, as J48 16 occurred in one case, the house could override privilege on its own J48 17 initiative. In this instance, Lord Deyncourt had left his J48 18 94-year-old father destitute by refusing to pay rents on properties J48 19 divested to him in trust, forcing his father to sue in Chancery to J48 20 collect the arrears. Lord Deyncourt's subsequent claim of privilege J48 21 had stayed the suit and his father therefore petitioned the house J48 22 for relief. The Lords summarily ordered the suit to proceed, J48 23 notwithstanding the claim of privilege, unless the arrears of rent J48 24 were paid before the beginning of the next legal term. Not all J48 25 claims, however, involved members of Parliament. In one notable J48 26 case, the house was asked to override a general protection issued J48 27 by the Privy Council in favour of the Muscovy Company against suits J48 28 initiated by its creditors. It did so without hesitation.

J48 29 The Lords were also asked on occasion to intercede directly in J48 30 lower court proceedings. In one case they did so to protect a J48 31 plaintiff against what appeared to be prejudicial actions taken by J48 32 the staff of the court of Wards. In another, they were asked to J48 33 prevail upon the court of Requests to expedite a case which had J48 34 been inadvertently suspended by the departure of the then Master, J48 35 Sir John Suckling, who had conducted the initial hearings. More J48 36 often, however, the problems encountered by litigants were J48 37 recurring ones which arose because of procedural or structural J48 38 defects in the system. Frequently cited as problematic were the J48 39 rules governing Bill of Review procedure in the court of Chancery. J48 40 In the absence of any outside appellate recourse from the court, J48 41 disgruntled litigants wishing to challenge a Chancery decree were J48 42 forced to secure a rehearing in that court through a Bill of J48 43 Review. However, the procedure was inherently prejudicial to the J48 44 applicant. In order to be granted a review, the appellant had to J48 45 demonstrate his 'good faith' by first fulfilling all of the J48 46 conditions of the original decree; that is, the judgment had to be J48 47 obeyed in all points before a bill to reverse it would even be J48 48 entertained. This imposed undue hardship in many cases, especially J48 49 where, as often happened, the decree involved a major financial J48 50 settlement between the parties. In addition, the complainant was J48 51 forced to provide bonds and securities to guarantee the payment of J48 52 costs and damages should the case go against him. Perhaps more J48 53 importantly, the bill would only be granted if the petitioner could J48 54 demonstrate some error or errors in the law in the body of the J48 55 decree. Only those errors could be discussed, and then J48 56 "without averment or further examination of matter of J48 57 fact". It was, as Sir Matthew Hale later described it, a J48 58 "somewhat straight-laced" procedure. In special J48 59 cases a bill might be allowed if new matter were discovered which J48 60 could not have been made available for the original hearing, but J48 61 this was much more the exception than the rule. In short, the J48 62 procedure was not designed to accommodate a rehearing on the J48 63 general merits of the cause or on the basis that the proceedings J48 64 had been arbitrary or corrupt. And yet that was very often what was J48 65 needed. In the case of Richard Wright, for example, the masters of J48 66 the court had issued a decree awarding possession of a piece of J48 67 property to Wright's adversaries, contrary to a report issued by J48 68 referees and without a hearing by the Lord Chancellor. As he J48 69 explained to the House of Lords ("his last refuge") J48 70 he was required to turn over the property in question before he J48 71 could be admitted to a Bill of Review. He requested that they J48 72 prevail on the Lord Keeper to waive that condition. The Lords J48 73 agreed and recommended that the case be reheard on a Bill of J48 74 Review, with the sole proviso that Wright put in sufficient J48 75 security to defray any costs or damages that might be awarded. A J48 76 similar situation had occurred in the case of a London merchant. J48 77 George Morgan had actually obtained a decree in Chancery from Lord J48 78 Chancellor Bacon awarding him pounds22,600 in settlement of some J48 79 long-standing commercial accounts. Unfortunately, Bacon's decree J48 80 was abruptly voided by Lord Keeper Williams, solely on the basis of J48 81 a petition from the seven defendants. His decision - that Morgan J48 82 should settle instead for pounds879 - was made out of term and J48 83 without a proper hearing. Again, by the rules of the court once the J48 84 decree had been enrolled Morgan would have had to accept the lesser J48 85 amount and presumably discharge the defendants of their obligation J48 86 before he could be admitted to a Bill of Review. That was clearly J48 87 unreasonable and the Lords ordered instead that the case be reheard J48 88 in its entirety (this time by Lord Keeper Coventry) and that J48 89 neither party was to "take any benefit of any former decree J48 90 or order." The Lord Keeper was additionally requested to J48 91 proceed with "all convenient expedition".

J48 92 Perhaps not surprisingly, the Lords were also called upon to J48 93 intercede in a variety of jurisdictional disputes, typically in the J48 94 context of a complaint made by a petitioner that his or her J48 95 legitimate proceedings in one court had been stayed by an J48 96 injunction issued out of another - the usual result of collateral J48 97 actions frequently undertaken by defendants. The petitioner would J48 98 then request that the house overrule the injunction or, J48 99 alternatively, make a clear determination on where the suit ought J48 100 best to be tried. Ralph Starkey, for example, petitioned the house J48 101 in 1626 asking for a determination on the proper jurisdiction for a J48 102 case involving a disputed will and property in Cheshire. Starkey J48 103 had initiated proceedings in Chancery, but the defendant (his J48 104 younger brother) had then secured a reference to the judges to J48 105 determine whether the Chancery proceedings did not in fact infringe J48 106 on the jurisdiction of the County Palatine of Chester. The judges J48 107 ruled that they did and the Chancery proceedings were stayed. In J48 108 reviewing the case the Lords determined that the suit ought to J48 109 proceed in Chancery, notwithstanding the Chester jurisdiction. J48 110 Their decision was based on the fact that the northern jurisdiction J48 111 would not accept the depositions of witnesses taken in the earlier J48 112 Chancery proceedings and, since most of Ralph Starkey's witnesses J48 113 were now dead, his case would have been wholly and unfairly J48 114 compromised. Equitable considerations, rather than a strict ruling J48 115 on jurisdiction, carried the day. In another case in 1628 Sir J48 116 Humphrey Ferrers petitioned the house protesting an injunction J48 117 issued out of the court of Wards which had stayed proceedings in J48 118 the Prerogative Court of Canterbury touching the estate of his J48 119 father-in-law, Sir John Packington. The injunction had ostensibly J48 120 been issued to protect the interests of Sir John's grandson, a ward J48 121 of the Crown, but had actually been granted on a minor technicality J48 122 at the behest of his widow, to prevent the ecclesiastical court J48 123 from awarding administration of the estate to Ferrers. The Lords J48 124 reviewed the case in full and determined that the Ward's injunction J48 125 amounted to unwarranted and unnecessary interference. They not only J48 126 ordered that the proceedings should continue in the Prerogative J48 127 Court (taking the concerns of the court of Wards into account), J48 128 but, in a rather unusual departure, went on to issue their own J48 129 recommendations as to how the case should actually be settled. As a J48 130 rule, injunctions were seen as a nuisance and an impediment to J48 131 judicial celerity and the courts which had granted them were J48 132 expected to offer clear and reasonable justification for their J48 133 issue or withdraw them.

J48 134 Most of these actions were designed to facilitate proceedings J48 135 elsewhere and did not constitute a formal appeal, at least in a J48 136 conventional sense. Clearly, however, the Lords were required to J48 137 consider the substance of many of these cases in order to make an J48 138 informed decision on how best to resolve them elsewhere. And they J48 139 could (and did) choose in a number of instances simply to void all J48 140 prior proceedings and former decrees in a case in favour of an J48 141 entirely new trial of the central matters in dispute. That was not J48 142 the same as reversing the substance of an earlier decree in favour J48 143 of one party over another, but for the aggrieved plaintiff, it J48 144 amounted to a successful appeal all the same.

J48 145 The house was, of course, willing and able to hear appeals J48 146 fully itself. In the case of complaints from the courts of common J48 147 law it generally did so, according to statutory authority, by way J48 148 of writ of error. Its appellate jurisdiction over courts of equity J48 149 was perhaps less clearly defined, but evolved rapidly (largely out J48 150 of necessity) through the Parliaments of the 1620s. As suggested J48 151 above, the absence of appellate recourse from the courts of equity, J48 152 and most particularly from the court of Chancery, was one of the J48 153 central complaints against the early modern legal system. But it J48 154 was also a difficult problem to resolve. On the face of it the J48 155 limitation made sense. In theory the provision of equitable remedy J48 156 was the responsibility of the king, who acted through J48 157 'intermediaries' appointed from his council. As James I reminded J48 158 his Star Chamber audience in 1616, the Lord Keeper was nothing so J48 159 much as "the dispenser of the king's conscience", J48 160 and the principle undoubtedly applied relative to the J48 161 responsibilities of the Lord Privy Seal in the court of Requests. J48 162 To suggest that an appeal was possible, in any conventional sense, J48 163 from either court, was to imply that royal justice was fallible or J48 164 that it could be (or ought to be) subject to independent review. J48 165 Contemporary political theory would not easily accommodate that J48 166 sort of assertion.

J48 167 However, the theory itself bore little relation to practical J48 168 reality. None of these courts was any longer an adjunct assembly of J48 169 the king's council. They were completely independent and thoroughly J48 170 institutionalized courts with highly sophisticated procedures and J48 171 professional staffs and, in truth, they operated far more J48 172 frequently with reference to principles of common law than to any J48 173 'divinely ordained' system of natural law. Moreover, as suggested J48 174 above, the powers exercised by these courts had steadily expanded J48 175 during the early seventeenth century and their actions - J48 176 particularly in real property disputes - could have a considerable J48 177 and often decisive impact. In view of the increasing delegation of J48 178 real judicial authority within the courts themselves (notably to J48 179 the masters in both Chancery and Requests) and the increasing J48 180 participation of the common law judges in equity proceedings, the J48 181 immunity of those decisions from subsequent review or reversal must J48 182 have seemed both anachronistic and unfair.

J48 183 As it was, once a Chancery decree had been enrolled the J48 184 disgruntled litigant had few options. He or she could proceed by J48 185 Bill of Review, with all of the limitations which that involved, or J48 186 he could petition the king, requesting an entirely new hearing by a J48 187 special commission of review appointed under the Great Seal. The J48 188 latter procedure was inevitably slow and expensive and remained J48 189 singularly unpopular. The only other avenue of redress was to J48 190 proceed by private bill and attempt to reverse the decree by act of J48 191 Parliament. Again, this method involved great cost and delay, J48 192 requiring as it did extensive hearings by both houses. It was, J48 193 however, tried repeatedly in the 1620s, notably without success. Of J48 194 the seven bills presented to reverse decrees in the court of J48 195 Chancery, five never made it past a first reading in the House of J48 196 Commons. J48 197 J49 1 <#FLOB:J49\>Similarly, where the intervener pays on account of J49 2 the debt but not on behalf of the debtor, there is no reason why he J49 3 should have conducted himself as the debtor's agent so long as the J49 4 debtor's assent is not explained by reference to the law of J49 5 agency.

J49 6 There is a further and different reason for preferring the J49 7 second, non-contractual, explanation. The agency solution J49 8 is only necessitated by the assumption that payments in discharge J49 9 generate unexecuted promises. That assumption mistakes the nature J49 10 of discharge even in the simple two party situation, for even there J49 11 a payment in discharge is a self-executing transaction leaving the J49 12 recipient creditor with nothing more to do. That being the case in J49 13 the two party situation, there is no reason why payments in J49 14 discharge by third parties should be differently construed; no J49 15 reason therefore to see the debtor's adoption as an attempt to make J49 16 himself a party to an executory contract rather than as an J49 17 acceptance of a benefit already executed.

J49 18 For these reasons it is thought that the agency explanation of J49 19 the debtor's adoption and consequent discharge is wrong and that J49 20 the non-contractual explanation is to be preferred. J49 21 However, that too is not free from difficulties.

J49 22 Its first challenge arises from a group of cases, usually J49 23 encountered in the context of novation, which may appear to have J49 24 assumed that, where the intervener attempts to discharge a debtor J49 25 by promising the creditor to pay the debt in place of the debtor, J49 26 the debtor can only be discharged if he makes himself a party to a J49 27 tripartite contract with the creditor and intervener. It is J49 28 submitted, however, that that is not the law, since each of these J49 29 cases is explicable on the narrower ground that the intervener did J49 30 not offer his promise rather than its performance, in satisfaction J49 31 of the debt. Where it is clear that the promise is offered and J49 32 accepted in satisfaction, nothing obstructs the application of the J49 33 rule that the debtor can perfect his discharge by simple, J49 34 non-contractual assent.

J49 35 The second difficulty arises from Snelling v. John J49 36 Snelling & Co. In that case each of three brothers, who were J49 37 directors of the company, promised the others that in the event of J49 38 his resigning his directorship he would forfeit the debt owed to J49 39 him by the company. When one brother resigned, he none the less J49 40 claimed the debt. Ormrod J held, following Scruttons Ltd. J49 41 v. Midland Silicones Ltd. and passing somewhat gingerly J49 42 over Hirachand Punamchand v. Temple, that the J49 43 company could not resist the action on its own initiative and would J49 44 have failed had not the promisee-brothers come in to protect it.

J49 45 The case is not easily distinguishable from Hirachand J49 46 unless on one ground. The debtor was the company. The creditor was J49 47 the plaintiff brother who had resigned his directorship. The J49 48 interveners were the brothers who had given good consideration for J49 49 the debtor's discharge. The distinguishing factor was that at the J49 50 time of the transaction the discharge was to operate, not J49 51 immediately, but on the happening of a future event. The creditor's J49 52 action was brought after the critical event had occurred. The J49 53 problem which the case raises in the present context is whether the J49 54 futurity of the discharge can account for the fact that even after J49 55 the event had happened the debtor could not adopt and thus defend J49 56 himself.

J49 57 One of us has argued that the element of futurity was J49 58 sufficient to distinguish Hirachand. That view required that a J49 59 transaction for the future discharge of a debtor should be J49 60 construed as an executory contract which would remain executory J49 61 even after the critical event had occurred. However, that now seems J49 62 wrong, because, once the event has happened, the creditor has J49 63 nothing more to do in order to execute his promise to discharge. If J49 64 he has nothing more to do, his promise cannot still be executory, J49 65 and the debtor should be able to perfect his discharge by J49 66 assenting. If this argument be right, the factor which J49 67 distinguishes Snelling from Hirachand carries no real J49 68 difference, a result which compels us to argue that Snelling J49 69 ought to have followed Hirachand and to have distinguished J49 70 Midland Silicones.

J49 71 At this point a larger problem behind Snelling is J49 72 revealed, for the question now becomes whether a rational J49 73 distinction can be made between the Belshaw v. Bush group J49 74 of cases and Midland Silicones. More particularly the J49 75 question is whether the principle of Belshaw can be prevented J49 76 from operating to allow strangers to take the benefit of exemption J49 77 clauses to which they are not parties.

J49 78 The resolution of that very large problem could not be avoided J49 79 here if the Belshaw cases were so few or equivocal as to be J49 80 threatened by the cases on exemption clauses. Since that is not the J49 81 case, we do not feel justified in digressing at great length from J49 82 our central theme. It is enough to indicate two possible solutions. J49 83 One is that the Belshaw cases ought indeed to be allowed to J49 84 achieve the results which were reached in the Eurymedon only J49 85 by stretching established doctrines of agency and formation of J49 86 contract, should be allowed, that is, to provide the revolutionary J49 87 short cut to a jus quaesitum tertio in this J49 88 kind of case. The other is that the Belshaw cases should be J49 89 dogmatically excluded from operating upon transactions to discharge J49 90 obligations which have not yet arisen, a dogma which might none the J49 91 less turn out to have as little foundation in principle as the rule J49 92 which excludes the creation of a trust of a covenant to convey J49 93 future property.

J49 94 (b) Non-volunteers J49 95 Where the intervener is not a volunteer a distinction must be J49 96 made between cases in which the voluntary character of his payment J49 97 is negatived by reason of a secondary liability, whether personal J49 98 or proprietary, itself not voluntarily incurred, and cases in which J49 99 it is negatived by some other factor.

J49 100 Where the intervener is not a volunteer because of his J49 101 secondary liability to make the payment, the debt is automatically J49 102 discharged by his payment. This proposition cannot be explained by J49 103 reference to the principles governing joint or joint and several J49 104 liability because the secondary liability is generally not of that J49 105 kind. Moreover, although the proposition can be supported from J49 106 dicta which assume its truth, we have not been able to discover a J49 107 case in which it has been fully considered. Its proof is therefore J49 108 somewhat indirect and is as follows.

J49 109 In the case which the proposition supposes, the intervener can J49 110 claim reimbursement from the debtor whether or not the debtor J49 111 adopts the intervention. The success of the claim for reimbursement J49 112 means that the debtor must have received a benefit by the payment. J49 113 If the payment did not discharge the debt, the debtor would have J49 114 received no benefit. The proposition thus comes down to this, that, J49 115 where the intervener can claim reimbursement irrespective of the J49 116 debtor's assent the debt must have been discharged by the J49 117 intervener's payment.

J49 118 This can be further supported by a simple historical argument. J49 119 The majority of claims to reimbursement were brought through the J49 120 count for money paid. That count recited that the money had been J49 121 paid at the request of the defendant. Where the intervener had paid J49 122 by reason of a secondary liability, the courts would 'imply' a J49 123 request on the part of the defendant, the debtor. By 'implying' J49 124 that request the courts assimilated compulsory payments to payments J49 125 expressly authorized by the debtor and thus made it unnecessary to J49 126 advance any other explanation why a compulsory payment did, but a J49 127 voluntary payment did not, automatically discharge the debt.

J49 128 Where the intervener is not a volunteer by reason of a factor J49 129 other than a secondary liability, his payment does not discharge J49 130 the debt. This proposition is subject to one insecure exception.

J49 131 The cases which allow the intervener reimbursement irrespective J49 132 of the debtor's assent to the intervention do not, subject to the J49 133 exception shortly to be mentioned, go beyond non-volunteers who are J49 134 such because of secondary liability. That means that a J49 135 non-volunteer who, for instance, is compelled to pay by duress or J49 136 who pays by mistake cannot move straight to reimbursement from the J49 137 debtor. And, if he cannot, although not a volunteer, the reason can J49 138 only be that the debt is not discharged by his payment.

J49 139 The explanation is that such payments can be recovered from the J49 140 creditor directly. Thus, slightly varying the facts of Astley J49 141 v. Reynolds, if a pawnbroker will not release pledges unless J49 142 the pledgor pays his brother's debt as well as his own, and the J49 143 pledgor is thus constrained to pay in order to obtain his property, J49 144 he may recover from the pawnbroker. It the pledgor chooses to leave J49 145 the money with the pawnbroker, his election to do so, unlike his J49 146 original payment, is a voluntary act and puts him back, J49 147 quoad the debtor, in the position of a voluntary J49 148 intervener. Applying the ordinary rule for volunteers, the debt J49 149 will then be discharged when the debtor assents.

J49 150 The insecure exception relates to cases in which the intervener J49 151 is constrained to act by necessity. It is arguable that he is then J49 152 entitled to claim reimbursement from the debtor on the same basis J49 153 as if he had paid under the constraint of a secondary liability. If J49 154 that is so, it follows that, here too, the payment must J49 155 automatically discharge the debt.

J49 156 A payment by necessity has in common with payments under the J49 157 compulsion of secondary liability the fact that there is no J49 158 question of immediate recovery from the creditor. If it is right J49 159 that a payment by necessity is to be treated on the same ground as J49 160 one made by reason of a secondary liability, there is some basis J49 161 for saying that the true rule is that every non-voluntary payment J49 162 of another's debt discharges the debt, unless the factor negativing J49 163 the voluntary character of the payment is such as to allow the J49 164 intervener an immediate right of recovery against the creditor.

J49 165 Answering the question whether the intervener's payment J49 166 discharges the debtor's debt, we have now considered the case where J49 167 the intervener is a volunteer and, with some inescapable J49 168 anticipation of the discussion of reimbursement, the case where he J49 169 is not. Before considering the intervener's restitutionary rights, J49 170 it is necessary to recall that we have not considered one species J49 171 of voluntary intervener, namely the intervener who pays under a J49 172 secondary liability voluntarily incurred. This is the case of J49 173 Owen v. Tate which will be considered in Part III. It J49 174 need only be said at this point that it may be a case in which the J49 175 intervener's payment, though voluntary, discharges the debt J49 176 irrespective of the debtor's assent.

J49 177 II. Can the Intervener Recover?

J49 178 (a) Volunteers

J49 179 Where the intervener's payment was voluntary and the debtor by J49 180 assenting chooses to perfect his discharge, the intervener cannot J49 181 recover from the creditor but can claim reimbursement from the J49 182 debtor provided that the debtor, when he assents, knows that the J49 183 intervener, when he paid, did not intend to make him a gift of his J49 184 discharge. The cases in which an assenting debtor comes under no J49 185 obligation to reimburse the intervener are: (i) where he assents J49 186 believing that the intervener acted donandi J49 187 animo; (ii) where - if such a case exists - the J49 188 intervener's payment, although voluntary, automatically discharged J49 189 the debt so that the debtor's assent is not the exercise of a J49 190 choice whether or not to accept the benefit.

J49 191 The first proposition in this statement, namely that after the J49 192 debtor's assent no action can be brought by the intervener to J49 193 recover his payment from the creditor, requires no discussion. Once J49 194 the debtor perfects his discharge the intervener has obtained all J49 195 that he paid the creditor for.

J49 196 The proposition that the intervener can recover from the J49 197 assenting debtor, except in the given cases, may seem heterodox at J49 198 first sight, since there are very many cases which lay down that a J49 199 voluntary payment may not be recovered. Those cases are, however, J49 200 concerned with debtors who either have not assented or have J49 201 assented within one of the two excepted cases set out above.

J49 202 Thus, Re Cleadon Trust Ltd. is a complex example of a J49 203 type of case in which the voluntary intervener fails against a J49 204 debtor who has never assented to the intervention and who, by J49 205 resisting the claim for reimbursement, insists that he has not done J49 206 so. J49 207 J50 1 <#FLOB:J50\>LEGISLATION

J50 2 Rights, Restraints and Pragmatism:

J50 3 The Human Fertilisation and Embryology Act 1990

J50 4 Jonathan Montgomery

J50 5 The Human Fertilisation and Embryology Act 1990 represents a J50 6 milestone in biomedical regulation. Not only does it finally bring J50 7 to fruition the long running government discussions about the J50 8 proper limits of reproductive science, it also provides the first J50 9 attempt in English law to provide a comprehensive framework for J50 10 making medical science democratically accountable. Its interest J50 11 therefore arises both from the solutions it adopts for particular J50 12 issues and from the model of regulation on which it builds. If the J50 13 licensing authority which lies at the heart of its provisions J50 14 proves successful, it is likely to be replicated in the oversight J50 15 of many controversial areas of medical progress.

J50 16 The 1990 Act is also significant as a model for establishing a J50 17 workable compromise between incompatible ethical positions. The J50 18 issues underlying the provisions of the Act are not ones on which a J50 19 consensus exists within our society. This resulted in an unusual J50 20 Parliamentary history. The Bill was introduced into the House of J50 21 Lords as a Government Bill but was mostly unwhipped and it J50 22 contained alternative versions of the crucial clause dealing with J50 23 the permissibility or otherwise of research on human embryos. The J50 24 original text of section 11 provided both an absolute prohibition J50 25 of embryo research and a provision permitting the licensing of such J50 26 research together with the proviso that both clauses could not be J50 27 in force at the same time.

J50 28 Thus, while the Government was clear when it introduced the J50 29 Bill that legislative provision had to be made, there was less J50 30 clarity as to the content of that provision. The uncertainty J50 31 extends back beyond the drafting of the Bill itself. The Committee J50 32 of Inquiry into Human Fertilisation and Embryology (the Warnock J50 33 Committee) received representations from the public, deliberated J50 34 the issues and reported in 1984. Yet this was felt to be J50 35 insufficient consultation and a second document to elicit comment J50 36 was published in 1986. Only after this second consultation did the J50 37 Government feel able to publish its own proposals in a White Paper. J50 38 The primary reason for this reticence was the reluctance to deal J50 39 with matters of conscience as Government proposals rather than by J50 40 way of private members' bill. This in turn springs from the J50 41 problems of identifying a consensus on such matters. In the event, J50 42 the 1990 Act can be seen as providing an institutional framework in J50 43 which compromises can be worked out rather than enshrining J50 44 solutions to the disagreements. It is appropriate, therefore, to J50 45 consider briefly why a consensus has proved so difficult to J50 46 achieve.

J50 47 Behind the provisions of the Act lie at least two issues of J50 48 principle on which there is wide disagreement and little scope for J50 49 compromise. The first issue concerns the status of the human J50 50 embryo. <}_><-|>One<+|>On<}/> the one side stand those who hold J50 51 that human embryos have the same moral status as mature adults from J50 52 the time of their conception onwards and that any steps leading to J50 53 their destruction must be analysed as a breach of fundamental human J50 54 rights. From this standpoint, research must be outlawed unless it J50 55 can be carried out without interfering with the embryo's J50 56 development. Abortion can only be permissible where it is carried J50 57 out to protect this same value of human life, that is to save the J50 58 life of the mother. On the other stand those who believe that J50 59 whatever status should be accorded to the human embryo, it is less J50 60 than that of mature adults and may therefore sometimes be J50 61 outweighed by the interests of adults who stand to benefit from J50 62 research or termination. The position of principle taken by the J50 63 first group prevents compromise because any concession will J50 64 necessarily be its abandonment. Concessions can be offered only as J50 65 a tactical ploy to secure a regime as close as possible to the J50 66 ideal.

J50 67 Accepting a less absolutist position on the rights of the human J50 68 embryo does not, however, resolve the issues. It is necessary to J50 69 determine the strength of the claims which are said to outweigh J50 70 those of the embryo. Those who believe that abortion is permissible J50 71 because the interests of the woman outweigh those of the fetus may J50 72 still believe that the need for research cannot justify killing an J50 73 embryo. To carry the debate further, it is necessary to consider J50 74 the strength of the claims to be weighed against those of the J50 75 embryo/fetus. Few would deny that pure research is valuable, but J50 76 whether it outweighs that of the embryo must depend in large part J50 77 on the importance of its application and here there is considerable J50 78 disagreement. Research is largely, although by no means J50 79 exclusively, expected first to improve our understanding of J50 80 infertility and how to overcome it, and second to help reduce the J50 81 incidence of genetic disorders. The latter issue is difficult J50 82 enough, raising the acceptability of eugenic approaches to human J50 83 reproduction. On the former there is even less agreement.

J50 84 The status of infertility raises the second of the major J50 85 background issues to the 1990 Act on which consensus is absent. J50 86 Much rhetoric has emerged under the slogan of 'the right to J50 87 reproduce' but the concept remains an obscure one. It might refer J50 88 to a range of claims. It can refer to no more than the negative J50 89 right no to be robbed of the capacity to procreate without just J50 90 cause. At the other extreme it might be said to entail a duty on J50 91 the state to provide services to help overcome infertility. Between J50 92 these two versions of the right is a claim that the state should J50 93 facilitate assisted reproduction by removing any rules of law which J50 94 prevent individuals taking steps to overcome their inability to J50 95 have children. The Human Fertilisation and Embryology Act 1990 says J50 96 nothing about claims of the first two types, but goes some way J50 97 towards meeting the third type of claim in the new status J50 98 provisions, which permit some of those who use assisted J50 99 reproductive techniques to be defined in law as parents (see below) J50 100 without recourse to a court.

J50 101 The strength of claims based on a right to reproduce depends on J50 102 the way in which infertility is classified. If being infertile is J50 103 to be ill, then overcoming it can more easily be said to be a state J50 104 responsibility. The National Health Service Act 1977 creates a duty J50 105 to provide services for the diagnosis and treatment of illness J50 106 (section 3) and defines "illness" as "any injury or J50 107 disability requiring medical or dental treatment or J50 108 nursing" (section 128). Yet (like kidney dialysis) J50 109 infertility treatment does not cure the underlying problem, but J50 110 rather overcomes some of its effects. The question is: how J50 111 important is it to overcome them? In the view of some it is a low J50 112 priority and cannot outweigh the interests of embryos: J50 113 "Infertility is a grief, not a disease. Those who suffer it J50 114 are not in pain and are not ill." Assessed subjectively, J50 115 however, infertility may be experienced as an extreme deprivation, J50 116 particularly in the face of a culture which regards a childless J50 117 woman as incomplete.

J50 118 These background issues are ones on which society is deeply J50 119 divided. In passing the 1990 Act, Parliament could have sought to J50 120 resolve them, but it will be argued here that the bulk of the J50 121 provisions of the Act do not provide answers to the questions, but J50 122 set up a forum in which they can be debated. Some parts of the Act J50 123 prohibit certain types of research activity outright, the J50 124 amendments to the Abortion Act 1967 alter the protection offered J50 125 for the human fetus and the status provisions do promote assisted J50 126 reproduction in some circumstances. These provisions can be seen as J50 127 resolving debates, but at the heart of the Act is the new statutory J50 128 licensing authority. Here the provisions do promote activities, but J50 129 are facilitative rather than prescriptive. The balance between the J50 130 competing views of the underlying issues of principle is left to be J50 131 determined within the regulatory structure of the authority, it has J50 132 not been established by Parliament itself.

J50 133 A Web of Regulation

J50 134 The nature of the forum in which the debates about infertility J50 135 treatment and embryo research are to be carried out is structured J50 136 by a complex web of discretion, restraints, control and J50 137 accountability. The Human Fertilisation and Embryology Authority is J50 138 to be given powers to oversee the activities of individual health J50 139 practitioners, and will not be limited to applying standards J50 140 established by Parliament or government. It will be able, indeed J50 141 required, to develop its own standards of what is acceptable and J50 142 proper. The authority will thus have considerable autonomy. It will J50 143 not be left entirely to its own devices, however, and a series of J50 144 checks will exist to supervise the exercise of its power.

J50 145 At a statutory level the 1990 Act establishes the ground rules J50 146 on which practitioners and the Authority will be required to J50 147 operate. Section 3 sets out a wide-ranging prohibition on creating, J50 148 keeping or using human embryos. Section 4 prohibits the storage and J50 149 use of human gametes. Exemption from these offences is provided for J50 150 those who have been licensed. Thus, an extensive dispensing power J50 151 is conferred upon by the Authority. It is not unlimited. Licences J50 152 may not authorise keeping an embryo beyond the appearance of the J50 153 primitive streak, placing a human embryo in an animal or replacing J50 154 the nucleus of an embryo (intended to prevent cloning).

J50 155 In addition to these provisions specifying what may not be J50 156 authorised, the Act also sets down the type of activities which can J50 157 be licensed (see Sched 2). These are, however, broadly phrased, and J50 158 include in relation to treatment "using gametes" J50 159 and "placing an embryo in a woman." Licences for J50 160 research must be "necessary or desirable" for J50 161 specified purposes, but these include "promoting advances J50 162 in the treatment of infertility" as well as increasing J50 163 knowledge about the causes of miscarriages and congenital disease. J50 164 These are narrower than the text of the Bill as originally printed, J50 165 which would have allowed research to be licensed "for the J50 166 purpose of increasing knowledge about the creation and development J50 167 of embryos and enabling such knowledge to be applied." J50 168 However, the purposes for which the Authority can grant licences J50 169 remain defined only at an abstract level.

J50 170 Having conferred this considerable flexibility upon the J50 171 licensing authority, the Government has also reserved a number of J50 172 powers to reduce it. The boundaries of the power to license may be J50 173 contracted and expanded by regulations. By requiring annual reports J50 174 to the Secretary of State, detailing both past activities and those J50 175 projected for the following twelve months, the Act seeks to ensure J50 176 that the Authority's practice can be monitored (s 7). The reports J50 177 are to be laid before Parliament by the Minister. Nevertheless, J50 178 pending the exercise of the power to make regulations, considerable J50 179 trust is placed in the members of the Authority.

J50 180 The Act permits the Authority to exercise a high degree of J50 181 control over practitioners. In addition to the power to grant or J50 182 withhold licences, the Authority will be able to issue mandatory J50 183 directions (ss 23 and 24). Failure to comply with these directions J50 184 will sometimes constitute an offence (s 41(2)), but will always J50 185 empower the Authority to revoke a licence (s18(1)(c)). Secondly, a J50 186 code of practice is to be drawn up to provide guidance as to the J50 187 proper conduct of licensees. This is less coercive than the power J50 188 to give directions in that disobedience will not in itself J50 189 constitute an offence, but it may be taken into account by the J50 190 Authority when considering revocation (s 25).

J50 191 In practice, this will make it difficult for practitioners to J50 192 resist the power of the Authority to make policy on controversial J50 193 matters. The difficulties of the Voluntary Licensing Authority in J50 194 the face of practitioners who are sufficiently confident of the J50 195 rightness of their activities to disregard peer pressure should not J50 196 be repeated. This will represent a considerable inroad into such J50 197 people's clinical autonomy. Opportunities to challenge decisions J50 198 are limited. Appeal against refusal of a licence by the Authority's J50 199 licence committee lies to the Authority and will take the form of a J50 200 full rehearing (s 20). However, appeal outside the Authority will J50 201 be available only on a point of law (s 21). J50 202 J51 1 <#FLOB:J51\>Another factor which might be thought to count J51 2 against epiphenomenalism is that the supposition of efficacy plays J51 3 a causal role in the explanation of human behaviour. Such behaviour J51 4 exhibits certain complex regularities which call for explanation J51 5 and which, at present, we explain (at least partly) in J51 6 psychological terms. These psychological explanations, though J51 7 typically of a rational rather than a mechanistic kind, attribute a J51 8 causal efficacy to the mental: they represent the subject's J51 9 behaviour as falling under the control of his beliefs and desires, J51 10 or under the control of his decisions and intentions, which are J51 11 responsive to his beliefs and desires. The claim might then be that J51 12 such explanations gain credibility from the fact that, as well as J51 13 being in their own terms successful, they cannot at present be J51 14 replaced by non-psychological explanations which cover the J51 15 same ground.

J51 16 These common-sense points against the epiphenomenalist are not, J51 17 to my mind, decisive. There is no denying the conflict between J51 18 epiphenomenalism and our ordinary conception of ourselves as J51 19 agents. But the fact that this is our ordinary view of the J51 20 <}_><-|>situaton<+|>situation<}/> does not mean that it is correct. J51 21 Just what grounds do we have for supposing that human behaviour J51 22 is intentional, and does qualify as action, in the J51 23 relevant sense? Again, there is no denying that, for ordinary J51 24 purposes, we need to make use of psychological explanations. But it J51 25 could still be claimed that the ultimately correct J51 26 explanations (maybe only discernible from a God's-eye view) are J51 27 purely physical. And indeed this is something which the defender of J51 28 the science-efficacy argument himself accepts. The most that J51 29 follows from these lines of reasoning is that the onus is on the J51 30 epiphenomenalist to try to justify his unusual position. But J51 31 presumably he is happy to take on this challenge. For he can say J51 32 that what supplies the justification is the scientific evidence in J51 33 favour of the view that the physical world is a closed system, J51 34 together with the already established case against J51 35 token-identity.

J51 36 However, it seems to me that epiphenomenalism is open to a much J51 37 more powerful objection than either of these. For I cannot see how, J51 38 if epiphenomenalism were true, the mind could form a topic for J51 39 overt discussion. Certainly, if mental items have no causal access J51 40 to our speech centres, the notion of an introspective report J51 41 collapses: even if the subject retains an introspective knowledge J51 42 of his mental states, his utterances do not count as J51 43 expressing that knowledge if it contributes nothing to their J51 44 production. But I cannot even see how, on the epiphenomenalistic J51 45 view, our language, as a medium for our utterances, makes semantic J51 46 contact with the mind at all. In what sense, for example, could the J51 47 word 'pain', as overtly used, be said to signify a certain type of J51 48 sensation, if neither the occurrence of the sensations nor our J51 49 introspective conception of their type affects its overt use? Quite J51 50 generally, it seems to me that if the mental contributes nothing to J51 51 the way in which the linguistic practices involving 'psychological' J51 52 terms are developed and sustained in the speech-community, and in J51 53 no other way affects the production of utterances employing these J51 54 terms, then, in respect of their overt use, the terms should be J51 55 analysed in a purely behaviourist or functionalist fashion - which J51 56 would deprive the epiphenomenalist of the linguistic resources to J51 57 enunciate his thesis. It is true, of course, that each J51 58 language-user may mentally interpret each term as signifying a J51 59 certain kind of (dualistically conceived) mental item. But I cannot J51 60 see how such private interpretations could have any bearing on the J51 61 objective meaning of the terms, as employed in speech and writing, J51 62 if, with respect to this employment, they are causally idle. (This J51 63 is not, of course, to endorse Wittgenstein's private language J51 64 argument. The sort of private interpretations which Wittgenstein J51 65 was trying to exclude would not, on my interactionist-dualist view, J51 66 be causally idle.)

J51 67 It does not follow from this that the state of affairs which J51 68 epiphenomenalism postulates is one which could not obtain - that J51 69 there could not be a world in which the mental had no causal J51 70 influence on the physical. Nor does it even follow that we can know J51 71 a priori that the actual world is not of this kind. For J51 72 it cannot be established a priori that the mental actually is a J51 73 topic for overt discussion. None the less, if the point is correct, J51 74 there is a sense in which, as overtly expressed, epiphenomenalism J51 75 becomes self-refuting. For if it is only possible to provide a J51 76 publicly audible or visible formulation of the thesis if the mental J51 77 is causally efficacious, then any attempt to provide such a J51 78 formulation can only succeed if the thesis is false. In effect, J51 79 then, we either have to accept that the thesis is false, or abandon J51 80 the attempt to make its truth or falsity an issue for public J51 81 debate, or even for private but vocal soliloquy. If this does not J51 82 quite refute epiphenomenalism, it at least renders it the sort of J51 83 position which cannot be seriously entertained - the sort of J51 84 position whose falsity (to echo Hume on the existence of body) we J51 85 are forced to take for granted in all our reasonings, even of a J51 86 philosophical kind.

J51 87 Having pressed this objection against epiphenomenalism, I must J51 88 now add a qualification. In claiming that if epiphenomenalism were J51 89 true, the mind could not form a topic for overt discussion, I have J51 90 been assuming that if mental events have no causal influence on the J51 91 physical world, then their occurrence will be, in every way, J51 92 irrelevant to any explanation of physical phenomena. This J51 93 assumption is a very natural one - and something which orthodox J51 94 epiphenomenalists are unlikely to dispute. However, so long as we J51 95 grant the coherence of theism, I think we can envisage two J51 96 situations in which the assumption would be false.

J51 97 The first situation is that in which, while mental items have J51 98 no causal influence on physical phenomena, they serve as J51 99 'occasions' for God to bring about certain physical events. For J51 100 example, it might be that whenever a human subject decides to act J51 101 in a certain way, God, taking note of the decision, causes the J51 102 subject's motor-neurons to fire in the appropriate way. The J51 103 decision itself does not cause the neuronal event; it does not even J51 104 indirectly cause it, by causally influencing God's decision, J51 105 since God is here thought of as entirely active and free, and hence J51 106 as not subject to any kind of external pressure. It is just that J51 107 God chooses to control the physical on-goings in such a way as to J51 108 match the subject's state of mind.

J51 109 The second situation is one in which, instead of controlling J51 110 the physical on-goings in a piecemeal fashion, God deliberately J51 111 arranges things, in advance and globally, so that the biological J51 112 creatures which evolve are constituted in a way which secures a J51 113 match between the functional roles of their neural states and the J51 114 psychological character of the mental items which these states J51 115 causally generate. Everything which occurs in the physical world J51 116 has an efficient cause in the preceding physical conditions, and J51 117 the mental items caused by neural states and events are themselves J51 118 causally idle. But since God has selected the physical and J51 119 psychophysical laws and the initial physical conditions of the J51 120 universe with a view to ensuring that the mental and the physical J51 121 realms harmonize in the appropriate way, there is still a sense in J51 122 which the fact that neural states and events have certain J51 123 psychological effects features in the ultimate explanation of the J51 124 physical phenomena. Thus when a subject makes a decision and his J51 125 muscles contract appropriately, we can say that, although it does J51 126 not contribute anything causally to the muscular movement, this J51 127 decision does have an ultimate explanatory bearing on the movement; J51 128 for it is only because the preceding central-neural event is J51 129 empowered to produce that sort of decision that it is also J51 130 empowered, by the structure of the organism and the physical laws, J51 131 to produce that kind of physical effect.

J51 132 Now of these two situations, it is only the second which has J51 133 any real relevance to the issue which concerns us. For the first, J51 134 in which God takes human mental items as occasions for causing J51 135 physical events, is not really an epiphenomenalist situation in the J51 136 intended sense. It does not deny the causal efficacy of the mental J51 137 in deference to the claims of physical science, and, if J51 138 anything, it looks more like an occasionalistic version of J51 139 interactionism than a form of epiphenomenalism. In any case, it J51 140 would be hard to find any rationale for it except as part of a J51 141 quite general occasionalist theory, such as that of Malebranche, J51 142 which denies the causal efficacy of mental and physical events J51 143 alike. The second situation, however, in which the physical world J51 144 is a closed system and God is the transcendent architect of the J51 145 psychophysical harmony, does yield a form of epiphenomenalism in J51 146 the intended sense. And it might be that, if such a situation J51 147 obtained, the explanatory link between mentality and behaviour J51 148 would be enough to render the mind a topic for overt discussion - J51 149 and, indeed, enough to allow us to retain some conception of J51 150 ourselves as agents. I shall return, briefly, to this point J51 151 presently. But, for the time being, let us continue with the J51 152 assumption that epiphenomenalism represents the mind, not merely as J51 153 having no causal influence on physical events, but as having no J51 154 explanatory role either.

J51 155 On this assumption, as we have seen, epiphenomenalism is a J51 156 position which, even if logically coherent, cannot be seriously J51 157 entertained. And here the crucial point is not that it conflicts J51 158 with common sense (i.e. in respect of our conception of ourselves J51 159 as agents and our employment of psychophysical explanations), but J51 160 that the supposition of its truth would oblige us to suppose that J51 161 the mind was not a topic for overt discussion. In this respect, J51 162 then, things are looking good for the identity-theorist. All he J51 163 needs to do, it seems, is to vindicate premise (a) of the J51 164 science-efficacy argument, and then validate the move from (a) to J51 165 (b), and the identity of mental with physical events will J51 166 automatically follow.

J51 167 However, while it is certainly true that claim (b), together J51 168 with the falsity of epiphenomenalism, would be enough to establish J51 169 the token-identity thesis, it would be wrong to conclude that our J51 170 argument against epiphenomenalism has made the prospects for this J51 171 thesis any better. Quite the reverse. For a little reflection shows J51 172 that the argument in question should be taken as an argument J51 173 against the acceptance of claim (b) rather than as a step in the J51 174 direction of token-identity. Let me explain.

J51 175 Claim (b) asserts that the physical world is a closed system. J51 176 Taken in the context of premise (a), from which it is inferred, J51 177 this implies not just that the only events which cause (or J51 178 contribute to the causation of) physical events are physical, but J51 179 also that the only qualitative factors which are ultimately (i.e. J51 180 metaphysically-fundamentally) operative in the causation of J51 181 physical events are physical. But, thus interpreted, it is easy to J51 182 show that, combined with what we have already established, claim J51 183 (b) itself generates a sort of epiphenomenalism. For although it J51 184 leaves room for the causal efficacy of mental items, it does not J51 185 leave room for the causal efficacy of their psychological J51 186 properties. Thus it allows us to say that a person's decision to J51 187 cross the road caused the subsequent firing of his motor-neurons; J51 188 but it does not allow us to say that its being a decision, or its J51 189 being a decision of that specific psychological type, played any J51 190 causal role in bringing about this effect. In short, it obliges us J51 191 to conclude that, with respect to the causation of the physical by J51 192 the mental, psychological properties are causally idle. Now I am J51 193 not suggesting that this conclusion immediately follows from claim J51 194 (b) taken on its own. After all, someone might accept J51 195 that it is only physical factors which are causally operative, but J51 196 secure the causal efficacy of psychological properties by J51 197 identifying them with physical properties. Or again, he might J51 198 accept that it is only physical factors which are ultimately J51 199 operative, but, by pressing some form of metaphysical reduction of J51 200 mental facts to physical facts, allow psychological properties to J51 201 enjoy a derivative efficacy. J51 202 J52 1 <#FLOB:J52\>Why should the denial of it strike Protagoras as J52 2 'incredible' (170 c 9)? The answer is clear from what was said J52 3 above in explanation of the slogan 'Man is the measure.' The import J52 4 of that slogan is not just that people believe all and only truths, J52 5 but also that we may apply ordinary criteria for deciding who J52 6 believes what. Furthermore Protagoras has all along implicitly J52 7 acknowledged that when we do apply those ordinary criteria we find J52 8 people believing the sorts of thing that Plato observes people to J52 9 believe. For if it were not evident that (to put it at its mildest) J52 10 people hold different and conflicting beliefs, there would have J52 11 been no need for Protagoras to introduce qualifiers and speak of J52 12 things being true for people; plain unqualified truth would have J52 13 done instead. There is thus no doubt that Protagoras will join J52 14 Plato in making the observation with which Plato's argument begins, J52 15 the observation that people think Protagoras wrong.

J52 16 There is a second contrast between Plato's own argument and the J52 17 argument that a belief that there are false beliefs guarantees its J52 18 own truth. When people think that people differ from one another in J52 19 what they know, then they think, not only that some people make J52 20 mistakes which others manage to avoid, but also that some people J52 21 have true beliefs which not all others share. This, I take it, is J52 22 the import of the remarks in 170 b 6-9 about knowledge, ignorance, J52 23 truth and falsehood. The consequence is that if Plato manages to J52 24 vindicate this generally held belief, he will have established that J52 25 Protagoras is wrong on two counts, rather than just on one; he will J52 26 have established not only that one's beliefs can be untrue, but J52 27 also that there can be truths which one does not believe.

J52 28 The third contrast is this. Plato observes not just that J52 29 someone, but that everyone, save perhaps for Protagoras himself, J52 30 thinks Protagoras wrong; there is, Plato observes, an almost J52 31 universal consensus of opinion against Protagoras. There was no J52 32 parallel to this in our argument concerning the belief that at J52 33 least one belief is not true. If that argument works at all, then J52 34 it works even if only one person believes that there are untrue J52 35 beliefs; and it certainly does not work any better if that belief J52 36 is shared by lots of people. Plato's argument does by contrast need J52 37 the premiss that, with the possible exception of Protagoras J52 38 himself, everyone thinks that man is not the measure. It needs this J52 39 premiss for two reasons, one slight and one serious. The slight J52 40 reason is that only if Protagoras is outnumbered can Plato infer, J52 41 as he does in the aside at 171 a 1-3, that Protagoras' doctrine is J52 42 more untrue than true. The serious reason is that unless it is J52 43 already accepted that no one else agrees with Protagoras' doctrine, J52 44 getting Protagoras to disagree with it will not mean that the J52 45 consensus against Protagoras' doctrine is completely unanimous. Yet J52 46 complete unanimity is what Plato wants: he means to prove that J52 47 Protagoras' doctrine is not true for anyone whatsoever (171 c 6-7; J52 48 compare 170 e 9 - 171 a 1). Only thus can Plato drop all J52 49 qualifications and call Protagoras' doctrine just plain untrue, J52 50 without any fear of the riposte: "Some people think it J52 51 untrue; but others think otherwise; and might not both sides be J52 52 right?"

J52 53 How can Protagoras be encouraged to join the consensus that man J52 54 is not the measure? Perhaps he does not need to; perhaps he is J52 55 already party to that consensus anyway. If so, we can proceed at J52 56 once to the conclusion that Protagoras' doctrine is not true for J52 57 anyone, and thus is not true at all (170 e 7 - 171 a 1). If not, J52 58 there is a little more work to do. Plato undertakes that extra work J52 59 at 171 a 6 - c 3. Protagoras must, says Plato, agree that those who J52 60 disagree with him think the truth when they think him mistaken; and J52 61 so he must himself agree he is mistaken. The converse, Plato points J52 62 out, does not hold: people other than Protagoras are under no J52 63 corresponding pressure to agree with him and think themselves wrong J52 64 (171 b 4-5). Why does Plato bother to point this out? Perhaps J52 65 Protagoras might otherwise try to declare a draw on the grounds J52 66 that his opponents are in a position exactly like his own; and J52 67 might then try to urge that this draw is yet another illustration J52 68 of how man is the measure. If it was intended to stop Protagoras J52 69 defending himself in this way, then Plato's remark is a success. J52 70 Those who disagree with Protagoras can hardly be expected to accept J52 71 also that his doctrine is true, or true for them, or even true for J52 72 Protagoras himself; after all, qualified truth, truth-for, is J52 73 Protagoras' notion, not theirs. Thus while the opposition need not J52 74 in any way accept the doctrine of Protagoras, Protagoras must agree J52 75 with the opposition that his doctrine is false. And thus the J52 76 doctrine of Protagoras is not true for anyone, not even for J52 77 Protagoras himself.

J52 78 Few readers of the Theaetetus feel entirely satisfied by J52 79 this. Protagoras is made to concede that the views of his opponents J52 80 are true. That, readers rightly feel, is a big concession. Need J52 81 Protagoras make it? Can he not simply say that what others think on J52 82 this topic is true for them, and for them alone? Why need J52 83 Protagoras himself accept as true what other people think on this J52 84 topic? Is Plato being fair when he has Protagoras accept the other J52 85 side's belief as true, instead of merely true for the other side J52 86 (171 a 8, b 2, b 6, b 11; compare b 1)? Yes, Plato is being fair. J52 87 Protagoras cannot here concede only that the opinion of those who J52 88 disagree with him is true for them. He must go further, and J52 89 speaking for himself say that their opinion is true. This can be J52 90 shown even if for the moment we allow that some beliefs can be true J52 91 for one person without being true for every other person; for even J52 92 if some beliefs can do that, the belief that man is not the measure J52 93 could not be among them. Consider an analogy; 'Traffic keeps left' J52 94 is, if you care to put it that way, true in Wales, and not true in J52 95 Canada; but such things as 'Traffic keeps left everywhere', J52 96 'Traffic keeps left somewhere', or 'Traffic keeps left in Canada', J52 97 purport to be true no matter what the country in which they are J52 98 said; such a thing therefore could not be true in only some J52 99 countries; such a thing would have to be either true everywhere or J52 100 else true nowhere whatsoever. As with truth in places, so too with J52 101 truth for people. If a belief purports to be true for every person, J52 102 it cannot be true for only some; such a belief would have to be J52 103 either true for everyone, or else true for no one whatsoever. This J52 104 applies to the widespread consensus that Protagoras is wrong. Those J52 105 party to the consensus hold, among other things, that if the best J52 106 that can be said for a belief is that it is their belief, then that J52 107 belief is not true. Now you cannot think such a thing, and think J52 108 also that the best that can be said for it is that you think it. It J52 109 would be absurd for me to say 'If something is only my opinion, J52 110 then it is not true; but mind you, this is only my opinion.' The J52 111 belief that Protagoras is wrong therefore purports to be something J52 112 more than just true for those who happen to accept it. That is why J52 113 Protagoras cannot simply say that the opinion of his opponents is J52 114 true for them, and stop there; that is why he must go on to say for J52 115 himself that their opinion is true. But once Protagoras does that, J52 116 there is a universal consensus that it is, without qualification, J52 117 false to say that man is the measure. And so we can conclude that J52 118 man is not the measure, full stop.

J52 119 Why not everything flows

J52 120 Protagoras' theory of relative truth may therefore be rejected. J52 121 When however the Secret Doctrine was first expounded, Protagoras' J52 122 theory of truth was connected with a metaphysic summarised in the J52 123 slogan 'all things flow'. That metaphysic implies that all our J52 124 thoughts, even those that apparently contradict one another, are J52 125 nevertheless equally true. That metaphysic therefore cannot be J52 126 allowed to stand. For if we allow that metaphysic to stand, then J52 127 our refutation of Protagoras will mean only that the truth which J52 128 equally characterises all our thoughts is not a relative truth, but J52 129 an absolute one, unqualified by any reference to their thinkers. J52 130 And if any thought is guaranteed to be unqualifiedly true just J52 131 because it is thought, then that threatens objectivity no less than J52 132 a guarantee that any thought is true for its thinker.

J52 133 Plato has a fairly brisk way with the idea that 'all things J52 134 flow'. Those who accept that slogan should accept, says Socrates, J52 135 that everything is changing in every way at every time J52 136 (Theaetetus 182 a 1). The 'every' here is to be taken with the J52 137 utmost strictness. It is not just that we have, for example, J52 138 Socrates change from seeing to not seeing the whitewashed wall as J52 139 white, and the whitewashed wall change from being white for J52 140 Socrates to not being white for him. It is not only objects, J52 141 whether perceiving or perceived, that change; perceptions and J52 142 perceived qualities must change too. Thus Socrates' glimpse of the J52 143 wall too must be changing, from being a glimpse to not being a J52 144 glimpse; and likewise the whiteness which makes the wall white for J52 145 Socrates cannot be stably a whiteness, but must itself be changing J52 146 (182 d 1 - e 5).

J52 147 We will later explore the consequences of this completely J52 148 universal change. In the meantime, let us ask why Plato insists J52 149 that those who accept the slogan 'all things flow' must believe the J52 150 particularly extreme version of that doctrine which he expounds. J52 151 Plato does in fact offer us some argument for his insistence. At J52 152 181 e 5-8 he points out that if any less extreme version were true, J52 153 then saying 'everything moves' would be no more correct than saying J52 154 'everything is still'. The trouble is that this argument sounds J52 155 less than conclusive. It invites the reply: 'all things flow' is J52 156 only a slogan; the doctrine which we mean by that slogan is the one J52 157 summarised so well at Theaetetus 156 a 2 - 157 c 2; according J52 158 to this doctrine of ours, the 'things' which 'flow' are perceiving J52 159 subjects, such as Socrates, and perceived objects, such as walls; J52 160 the sense in which a perceiving subject 'flows' is that Socrates, J52 161 who perceives the wall as white now, can nevertheless perceive J52 162 other things differently, and can even perceive the wall J52 163 differently at some other time; likewise, the sense in which a J52 164 perceived object 'flows' is that the wall, which now is white for J52 165 Socrates, can nevertheless have other colours for other people, and J52 166 can even have other colours for Socrates himself at some other J52 167 time; perceived qualities and perceptions of those qualities are J52 168 not themselves among the 'things' which, according to our doctrine, J52 169 'flow'; so perhaps 'all things flow' is therefore a poor slogan for J52 170 our own modest doctrine, and a better slogan for a doctrine far J52 171 more extreme; but at most that requires us to abandon the slogan; J52 172 it hardly requires us to accept the extreme doctrine; and in that J52 173 case, Plato's refutation of the doctrine that 'all things flow' is J52 174 a refutation of a doctrine that nobody either does or need hold; J52 175 Plato has just invented an easy target for himself. Such at any J52 176 rate is the accusation which Plato's arguments invite. But is Plato J52 177 guilty as charged?

J52 178 Plato is innocent; there are sound reasons for him to formulate J52 179 and refute the extreme doctrine that 'everything is changing in J52 180 every way at every time'. The first reason is that only the extreme J52 181 doctrine coheres with Protagoras' claim that man is the measure of J52 182 all things. J52 183 J53 1 <#FLOB:J53\>I aim to make these inadequacies clear, and then to add J53 2 some brief speculations on why they are present.

J53 3 There is one significant feature of Aristotle's whole J53 4 discussion in Book VI, which I would not regard as an 'inadequacy', J53 5 but which deserves a mention here at the outset. That is the J53 6 tension between Aristotle's reasoning in this book and his J53 7 discussion of infinity in Book III. To put it very roughly, his J53 8 position in Book III is that there is no 'actual' or 'completed' J53 9 infinite, and all infinity is merely 'potential'. But in Book VI he J53 10 shows no such tendency to be suspicious of the infinite, and J53 11 apparently accepts without qualms that a line does ('actually') J53 12 contain infinitely many points, that a stretch of time does J53 13 ('actually') contain infinitely many instants, and so on. Thus his J53 14 answer in this book to Zeno's most famous paradox on motion J53 15 apparently accepts that one who traverses a finite distance has J53 16 thereby completed a series of infinitely many distinct tasks, in J53 17 traversing the infinitely many 'half-distances' contained J53 18 within the original distance. This is possible, he here says, J53 19 because the finite time available may equally be divided into a J53 20 corresponding and infinite series of 'half-times' J53 21 (233a13-31, 239b9-29). Later in Book VIII he J53 22 will reconsider this paradox, and with the doctrine of Book III in J53 23 mind he will deny that a moving body ever does pass over infinitely J53 24 many 'actual' points (263a4-b9). But here in J53 25 Book VI we find no such complications. A simple hypothesis J53 26 evidently suggests itself: Book VI was written before the doctrine J53 27 of Book III was worked out. (In my concluding section I shall note J53 28 some other pointers to the early date of Book VI.)

J53 29 I The Definition of a Continuum

J53 30 In chapter 3 of Book V Aristotle defines what it is for one J53 31 thing to be continuous with another. He begins by defining J53 32 'succession' (t<*_>o-grave<*/> veFexes): J53 33 one thing succeeds another when it comes after it, in position or J53 34 in some other way, and there is nothing between them that is of the J53 35 same kind (226b34-227a6). A special case of J53 36 succession is 'being next to' (t<*_>o-grave<*/> J53 37 vech<*_>o-acute<*/>menon): one thing is next to another J53 38 when it succeeds it and is also in contact with it J53 39 (227a6), contact being defined as occurring when (some J53 40 part of) the 'limits' of the two things are in exactly the same J53 41 place (226b21-3). A special case of being next to is J53 42 'continuity' (t<*_>o-grave<*/> J53 43 sunech<*_>e-acute<*/>s): one thing is continuous with J53 44 another when they are next to one another and in addition the J53 45 limits at which they touch are the same thing, or 'have become J53 46 one', being indeed 'held together' as the word J53 47 sun-ech<*_>e-acute<*/>s implies. J53 48 Aristotle's thought evidently is that things which are merely next J53 49 to one another need not hold together - one may move one of them J53 50 leaving the other where it is - whereas if they are continuous then J53 51 they move together (227a10-17).

J53 52 At the opening of Book VI Aristotle recalls this definition, J53 53 and appears to suppose that it has also explained to us what it is J53 54 for a single thing to be a continuous thing - or, as I shall say, a J53 55 continuum - though clearly it has not explained this at all J53 56 (231a21-6). But if we follow through the argument that he J53 57 at once plunges into, we can I think reconstruct the definition J53 58 that he has failed to state but must have had in mind. For he J53 59 claims that it follows from the definition that a continuum (such J53 60 as a line) cannot be made up from indivisible things (such as J53 61 points). The first argument is simply that an indivisible thing has J53 62 no limits, for a limit J53 63 (v<*_>e-grave<*/>schaton) must be different J53 64 from what it limits, and thus what has limits must have parts, and J53 65 must therefore be divisible (231a25-9). As a second J53 66 argument Aristotle adds that indivisible things such as points J53 67 cannot form a continuum by being in contact with one another, even J53 68 if we waive the point about their lack of limits. For, he says, if J53 69 a point could be in contact with a point then they would have to be J53 70 in contact 'as wholes', and this (I think) he takes to imply that J53 71 they would have to be in exactly the same place. At any rate he J53 72 objects that contact of this kind will not produce anything J53 73 continuous, because a continuous thing will be divisible into parts J53 74 that occupy different places (231a29-b6). J53 75 (The thought here seems to be this. When one (perfect) sphere J53 76 touches another there will be a point on the surface of the one J53 77 that touches a point on the surface of the other, and these are J53 78 different points, just as the face of one cube in contact with the J53 79 face of another are different faces (see n. 5 above). Hence one J53 80 point may be said to be in contact with another, but such contact J53 81 yields spatial coincidence, so no matter how many points we may put J53 82 together that are in contact with one another in this way we shall J53 83 still not form anything bigger than a single point. But this J53 84 subtlety is generally ignored in what follows, where it is simply J53 85 claimed that indivisible things cannot be in contact (e.g. J53 86 231b17).) Finally he adds as a third argument that points J53 87 cannot even be successive, since between any two points there will J53 88 be a line (and hence, he presumably means to add, a further point) J53 89 (231b6-10). Summing up this reasoning, Aristotle J53 90 concludes that what a thing is made up from it may also be divided J53 91 into, and so we have shown that a continuum cannot be divided into J53 92 indivisibles. Anything whatever must be either divisible or J53 93 indivisible, and if divisible then either divisible into J53 94 indivisibles or divisible only into what may always be divided J53 95 further; this last is a continuum. And conversely every continuum J53 96 is divisible into what may always be divided further, for if it J53 97 could be divided into indivisibles then one indivisible could touch J53 98 another, "for the limits of (two) continuous things touch, J53 99 and are one" (231b10-18).

J53 100 Now, if this argument is to be intelligible, what must we take J53 101 the definition of a continuum to be? First, it must evidently be J53 102 given as a premiss that a continuum has at least two parts which do J53 103 not coincide (213b4-6). For Aristotle clearly does not J53 104 count a point as itself a continuum, and this seems to be the J53 105 minimal premiss needed to rule out that suggestion. To avoid some J53 106 complications I should like to strengthen this premiss a little to J53 107 the following: a continuum may always be divided (without J53 108 remainder) into two parts that do not coincide. It is true that, J53 109 given a 'logic of parts' that is nowadays orthodox, the supposedly J53 110 stronger version would follow at once from the original. For the J53 111 now orthodox logic supposes that, given any one proper part of a J53 112 thing, there will always be some one further part of the J53 113 thing, which comprises all the rest of it excluding the given part. J53 114 Thus the whole may be said to be divided into these two parts in J53 115 the sense that the two do not overlap one another - i.e. they have J53 116 no common part - and together they exhaust the original whole, in J53 117 so far as every part of it must overlap at least one of these two. J53 118 But it seems to me doubtful that Aristotle would accept the J53 119 assumption on which this reasoning is based. For example, it J53 120 implies that if we begin with a line three inches long, and J53 121 subtract a part one inch long from the middle of it, then the two J53 122 separated line-segments remaining may be counted as together J53 123 forming one part of the original. But this is not a very J53 124 natural use of the notion of one part, and equally it is not very J53 125 natural to say that the operation just envisaged would divide our J53 126 three-inch line into just two parts. I shall come back to this J53 127 problem shortly, but for the moment let us simply side-step it, by J53 128 adopting the stronger premiss: any continuum may always be divided J53 129 into two.

J53 130 Then it is perfectly clear from the whole run of Aristotle's J53 131 discussion that we must add a further premiss: any division of a J53 132 continuum into two (non-coincident) parts must divide it into parts J53 133 that are continuous with one another in the sense defined in Book J53 134 V, i.e. the two parts must touch, and the limits where they touch J53 135 must 'be one'. These two premisses are, I imagine, the premisses J53 136 that Aristotle regards his argument in this passage as depending J53 137 upon, and as given by the (unstated) definition of what a continuum J53 138 is. We may therefore re-construct the definition by J53 139 designing it to yield just these premisses: a continuum is anything J53 140 which (i) can be divided into two parts, and (ii) is such that any J53 141 two parts into which it is divided must share a limit. But if this J53 142 is indeed the definition Aristotle has in mind, then either it is J53 143 incorrect or the conclusion that he attempts to deduce from it will J53 144 not follow. That is, it will not follow that the parts into which a J53 145 continuum may be divided must themselves be further divisible.

J53 146 Suppose we take a finite line, and suggest 'dividing' it into J53 147 these two parts: one part is to be an end-point of the line, and J53 148 the other part is to be all the rest of the line, excluding this J53 149 end-point. (Thus the second part is what we call a 'half-open' J53 150 interval, containing all the points of the line except this one J53 151 end-point.) Now we may surely assume that a finite line is a J53 152 continuum, if anything is, so if Aristotle's definition is correct J53 153 then the two parts into which we have divided it must share a J53 154 limit. According to one of Aristotle's lines of thought, they do J53 155 not, since he claims that a point has no limit. But in that case we J53 156 must simply reject the definition as incorrect. However, if my J53 157 interpretation is right, then he also has another line of thought J53 158 which allows points to be in contact with one another ('as wholes') J53 159 despite their alleged lack of limits. This is most easily J53 160 harmonized with his general position by allowing that a point may J53 161 be said to be its own limit. If we do allow this, then J53 162 the suggested definition of a continuum may be retained, even in J53 163 the face of this example. For the end-point of the line, and the J53 164 rest of the line, now do share a limit, namely the end-point J53 165 itself. It is its own limit, and it must also be reckoned to be one J53 166 of the limits of the rest of the line. For we cannot say that the J53 167 rest of the line has no limit in this direction without once more J53 168 subverting the suggested definition, and if it is to have a limit J53 169 then clearly there is no other candidate than this same end-point. J53 170 Thus, if the definition is to be retained, while admitting this J53 171 example as a 'division', then we must conclude that it is possible J53 172 to divide a line into two parts, where one of these parts is a J53 173 mere point, and hence not further divisible. (And we may add, J53 174 incidentally, that any point of the line may thus be exhibited J53 175 as one of the parts into which it may be divided, by considering a J53 176 division into three parts, one of which is the given point, one the J53 177 rest of the line (if any) that is to the left of it, and one the J53 178 rest of the line (if any) that is to the right of it.)

J53 179 There are two possible avenues one might explore, in seeking to J53 180 rescue Aristotle from this objection. One would be to note that I J53 181 have merely conjectured his definition of a continuum (since he J53 182 fails to state any definition himself), and to seek for an J53 183 alternative conjecture. Here one may remark that he quite often J53 184 appears to take, as definitive of a continuum, the thesis that J53 185 in this passage he is attempting to prove. J53 186 J54 1 <#FLOB:J54\>We will suppose that at the time of utterance both J54 2 parties are alive. What is required for the truth of this J54 3 conditional? Following Stalnaker's idea, it is true iff in the J54 4 situation as close as possible to ours but in which John dies J54 5 before Joan, Joan inherits the lot. Thus if, for example, John's J54 6 will actually leaves everything to Joan, a situation as close to J54 7 ours as possible is one in which his will is unchanged, and hence a J54 8 situation in which, given the normal flow of events, Joan will J54 9 inherit the lot. (If John is going to die before Joan, then the J54 10 situation to consider is the actual future course of events: J54 11 nothing is more similar to the actual world than itself.) In this J54 12 case, Stalnaker's theory seems to give precisely the right J54 13 result.

J54 14 It also delivers the right result in the case of J54 15 (1.21),

J54 16 If someone broke in, they must have repaired the damage before J54 17 they left,

J54 18 which, arguably, has to be treated as undefined by the J54 19 probability theory. Consider a world in which someone did break in, J54 20 but which is otherwise as similar as possible to ours. In ours, J54 21 there is no trace of violent entry. How could this be, unless the J54 22 intruder had repaired the damage behind him? So the closest (that J54 23 is, most similar) world in which the antecedent is true is one in J54 24 which the consequent is true, so the conditional is true (and hence J54 25 can be used to infer that no one broke in, given that no intruder J54 26 would stop to repair the damage behind him).

J54 27 Stalnaker explicitly considers, and rejects, the theory J54 28 mentioned in chapter 2.4 that a conditional requires for its J54 29 truth some connection between what would make its antecedent true J54 30 and what would make its consequent true. He offers the following J54 31 counterexample: suppose that at the time of the Vietnam conflict J54 32 you are convinced that the United States will use nuclear weapons, J54 33 come what may. You therefore also think that the Chinese entering J54 34 the conflict would have no effect whatsoever on the US decision to J54 35 use nuclear weapons. Yet you will confidently affirm the J54 36 conditional:

J54 37 1) If the Chinese enter the Vietnam conflict, the United States J54 38 will use nuclear weapons.

J54 39 Notice that on Stalnaker's theory this conditional is, if your J54 40 political views are correct, true, because any smallest revision of J54 41 the actual world which verifies the antecedent will of course J54 42 verify the consequent. But this is not to say that the truth of the J54 43 consequent is in general sufficient for the truth of the J54 44 conditional. On the contrary, this will not always hold: not, for J54 45 example, in conditionals of the form 'if not-A then J54 46 A'.

J54 47 What impact does Stalnaker's account of the truth conditions of J54 48 conditionals have upon validity? It turns out that his account of J54 49 truth conditions validates (in the standard sense) just the J54 50 arguments that are valid in the probabilistic sense.

J54 51 He agrees with the probabilistic theory in finding (2.8.1) J54 52 invalid, and he offers a sophisticated explanation of the J54 53 appearance of validity (Stalnaker [1980]). His idea is that J54 54 (2.8.1) corresponds to a 'reasonable' inference: one such J54 55 that, in any context in which the premises are acceptable, so is J54 56 the conclusion. But reasonable inferences can be invalid, and J54 57 (2.8.1) is an example. If we think it is valid, it is because J54 58 we mistake reasonableness for validity.

J54 59 Everything hinges, of course, on the account of reasonableness. J54 60 Without giving Stalnaker's whole theory, let us simply see how it J54 61 is intended to apply to the example. A context in which the premise J54 62 of the argument is acceptable is one which leaves open whether it J54 63 was the butler or the gardener who did it. (If you knew who did it, J54 64 you shouldn't assert the premise, but simply 'the gardener did it', J54 65 or 'the butler did it', as the case may be.) When you have asserted J54 66 or accepted the premise, the context includes the premise: it is J54 67 now part of what is taken for granted. Hence when you evaluate the J54 68 conclusion, you should consider a world in which the gardener J54 69 didn't do it, but in which the premise holds true. Such a world J54 70 must be one in which the butler did it. Reasonableness puts J54 71 restrictions on what alternative worlds to consider in evaluating a J54 72 conditional.

J54 73 Even if reasonable, the argument is invalid, by Stalnaker's J54 74 account. For there is no guarantee whatsoever that in the world J54 75 most like ours, assuming that ours is one which verifies the J54 76 premise, but in which the gardener did not do it, the butler did. J54 77 One possibility is that our world is one in which the butler and J54 78 the gardener were in league: they both did it. The most similar J54 79 world to this verifying 'The gardener did not do it' could well be J54 80 one in which the victim was not murdered at all: a world in which, J54 81 say, the victim fell (rather than being pushed). The relevant world J54 82 cannot be exactly like our world, since in our world (we are J54 83 supposing) the butler and gardener both did it. A world in which J54 84 the butler-gardener conspiracy broke down might well be less J54 85 similar to ours than one in which the conspiracy is in force, but J54 86 the conspirators were thwarted at the last moment by their victim's J54 87 accidental death. The truth of the premise of the argument thus J54 88 does not guarantee the truth of the conclusion, upon Stalnaker's J54 89 semantics.

J54 90 Stalnaker intends his account to apply to subjunctive as well J54 91 as indicative conditionals. But how will it specify the difference J54 92 between matched pairs of these different kinds of conditional? For J54 93 example

J54 94 2.4.34) If Oswald didn't shoot Kennedy, someone else J54 95 did.

J54 96 2.4.35) If Oswald hadn't shot Kennedy, someone else would J54 97 have.

J54 98 Let us apply Stalnaker's recipe to these successively, J54 99 retaining the background assumptions we relied upon in chapter J54 100 2.4 to justify the reasonableness of accepting the first while J54 101 rejecting the second.

J54 102 For (2.4.34) we consider a world in which Oswald didn't J54 103 shoot Kennedy. Remember that Oswald's guilt was a disputed J54 104 question, but Kennedy's death from shooting was not. So a similar J54 105 world will be one in which Kennedy died from a shooting, while not J54 106 having been shot by Oswald. Such a world, obviously, verifies the J54 107 consequent. Hence Stalnaker's account has it that (2.4.34) is J54 108 true, and this is as it should be.

J54 109 What should we consider for (2.4.35)? Although Stalnaker J54 110 is not explicit, he clearly intends us to consider a world in which J54 111 Oswald did not shoot Kennedy. (In other words, we revise the J54 112 antecedent from 'hadn't' to 'didn't'.) So it would appear that we J54 113 have to consider just the same world as we did before. And just as J54 114 we must revise the antecedent, to apply Stalnaker's recipe, so we J54 115 have to revise the consequent. What we need to consider, then, is J54 116 whether in a world in which Oswald did not shoot Kennedy, but which J54 117 is otherwise as like as possible to this, someone else shot J54 118 Kennedy. In short, we seem to have the same question before us as J54 119 we did in the case of (2.4.34), in which case we should give J54 120 the same answer. But this answer, that (2.3.35) is true, J54 121 conflicts with our original intuition. (We imagined that we thought J54 122 Oswald acted alone, there was no conspiracy, so if he hadn't shot J54 123 Kennedy no one else would have.)

J54 124 Stalnaker recognizes that there will be considerable J54 125 context-sensitivity in point of what counts as the world most J54 126 similar to ours in which the antecedent holds, and he suggests that J54 127 this can account for the difference between (2.4.34) and J54 128 (2.4.35). What remains unclear, however, is the relationship, J54 129 which presumably ought to be systematic, between the contexts we J54 130 envisage for (2.4.34) and (2.4.35), and the different J54 131 dimensions of similarity relevant to the truth conditions. For J54 132 (2.4.35) we want to count a world in which Oswald did not J54 133 shoot Kennedy and Kennedy was not shot as the most similar to ours; J54 134 for (2.4.34) we want to count a world in which Oswald did not J54 135 shoot Kennedy, and he was shot, as the most similar to ours. But J54 136 what features of the context make for this difference? Clearly a J54 137 vital difference is that in (2.4.34) it is taken for granted, J54 138 as part of the background, that Kennedy was shot. In (2.4.35) J54 139 this background assumption somehow gets suspended in selecting the J54 140 relevant world. But what is the mechanism? It is certainly relevant J54 141 that the envisaged use of (2.4.35), unlike (2.4.34), J54 142 takes for granted that Oswald in fact did shoot Kennedy, yet any J54 143 world which will satisfy Stalnaker's test will be one in which the J54 144 presupposition is abandoned. Perhaps abandoning taken-for-granted J54 145 facts induces much more relaxed standards of similarity. A world as J54 146 similar as possible to this one but in which Oswald didn't shoot J54 147 Kennedy is certainly not this world, in the context of J54 148 (2.4.35), but is not certainly not this world in the context J54 149 of (2.4.34).

J54 150 The more obvious account of 'most similar' in this case brings J54 151 out both the indicative and the subjunctive as true. This is not J54 152 invariable. Gibbard [1980] has suggested the following case, J54 153 designed to show that Stalnaker's account works best for J54 154 subjunctive conditionals, and the probability theory for indicative J54 155 ones.

J54 156 Jack and Stone are playing poker and Stone has just bet the J54 157 limit in the final round. Zack has seen Stone's hand, seen that it J54 158 is a good hand, and has signalled its contents to Jack. Zack J54 159 remains unaware of the contents of Jack's hand. Stone, suspecting J54 160 mischief, orders the room to be cleared. Not knowing the outcome of J54 161 the game, Zack can confidently assert later:

J54 162 2) If Jack called, he won.

J54 163 His grounds are that Jack is an experienced player, wants to J54 164 win, and, knowing Stone's hand, knows whether or not he will win by J54 165 calling. However, Zack can reasonably be doubtful about

J54 166 3) If Jack had called, he would have won.

J54 167 After all, Stone's hand was good, and it's not likely that Jack J54 168 had a better. Let us first confirm Gibbard's claim that the J54 169 probability theory gives an acceptable account of (2), and J54 170 then see how Stalnaker's account deals with the two sentences.

J54 171 On the probability theory, (2) is assertible iff it has J54 172 high probability, which by (1.4), means that Pr (Jack won|Jack J54 173 called) is high. The result is that (2) is ruled highly J54 174 assertible, as it seems to be, since, given that Jack called, it is J54 175 extremely likely that he won.

J54 176 Turning now to (3), Gibbard argues that our intuitions J54 177 harmonize with Stalnaker's account. We in fact regard the truth or J54 178 falsehood of (3) as determined by whether or not Jack's hand J54 179 was better than Stone's. On Stalnaker's account, we must find a J54 180 world as similar as possible to ours, but in which Jack called. J54 181 According to Gibbard, such a world is determined by the following J54 182 criteria:

J54 183 it is exactly like the actual world until it is time J54 184 for Jack to call or fold; then it is like the actual world apart J54 185 from whatever it is that constitutes Jack's decision to call or to J54 186 fold, and from then on it develops in accordance with natural J54 187 laws. (pp. 227-8: I have changed Gibbard's Pete to Jack)

J54 188 However, it seems that a world in which Jack calls, and which J54 189 is maximally similar to ours, is a world in which Jack knows J54 190 Stone's hand and calls. Preserving similarity with our world, this J54 191 can only be because Jack knows that he has a winning hand. In this J54 192 world, therefore, the consequent of the conditional is true, rather J54 193 than false, and so, contrary to Gibbard's claim, Stalnaker's J54 194 account seems not to deliver the desired result for (3).

J54 195 Gibbard envisages Zack's response to (3) as occurring when J54 196 Zack is still ignorant of what Stone did. The most natural setting J54 197 for (3), however, is one in which it is known that Jack J54 198 folded. Given that he folded because he knew Stone had the better J54 199 hand, Stalnaker's account clearly works well. The alternative world J54 200 is like our world up to the time of Jack's decision. Before then, J54 201 he knows he has a losing hand. J54 202 J55 1 <#FLOB:J55\>In effect, the sub-commissions, instead of expediting J55 2 valuation, had actively engaged in collusion with local vested J55 3 interests to restrict the amount of teind available for J55 4 redistribution: a situation that can be traced directly to the J55 5 hostile reception accorded to Charles's legal decreet of 18 J55 6 September 1629.

J55 7 Class Collusion

J55 8 The legal decreet can be separated into five major components J55 9 affecting the future disposal of superiorities and teinds under the J55 10 general headings of compensation for surrenders, costing of teinds, J55 11 limitations on purchase, quantification of teinds and J55 12 redistribution for pious and other uses. Other than costing and J55 13 quantification where the Commission for Surrenders and Teinds was J55 14 empowered to make empirical revisions, Charles was content to J55 15 ratify recommendations already made by the Commission.

J55 16 Compensation to the temporal lords for loss of their J55 17 superiorities was confirmed as 1000 merks for every chalder of J55 18 victual or 100 merks of free rent: that is, ten times the J55 19 feu-duties hitherto paid by their feuars less the feu-duty paid by J55 20 the temporal lords themselves to the Crown. Superiorities were to J55 21 be purchased at a price specified as ten years' rental (that is, J55 22 free rent), an offer initially made to John Maitland, first earl of J55 23 Lauderdale, but regarded as prejudicial to the nobility when upheld J55 24 by the Commission on 29 June 1627. The actual mechanism of sale, J55 25 suggested by the militant vanguard on 10 November 1628, was not J55 26 approved formally until 17 July 1630. Feuars who advanced the J55 27 purchase price for superiorities to temporal lords were to hold J55 28 their kirklands directly of the Crown as freeholders. Their J55 29 feu-duties, which were now owed to the Crown, were to be retained J55 30 for up to ten years to recoup monies advanced to temporal lords. J55 31 Thus, by mortgaging the kirklands to existing feuars, temporal J55 32 lords were to be compensated for their superiorities at no cost to J55 33 the Crown. Not every feuar, however, had a sum equivalent to ten J55 34 years' rental readily available. Nor were all feuars prepared to J55 35 combine to strip temporal lords of their superiorities. Moreover, J55 36 no change was proposed in apportioning taxation. Feuars who bought J55 37 out superiority were still to pay reliefs to temporal lords who J55 38 remained responsible for uplifting taxes from kirklands.

J55 39 The price at which heritors could acquire control of their own J55 40 teinds was costed at nine years' purchase for every 100 merks paid J55 41 annually in money: that is, a purchase price equivalent to nine J55 42 years' valued rent. Where teinds were paid in kind, the same J55 43 costing applied once the valuations were commuted into monetary J55 44 sums - allowance having been made for regional variations in the J55 45 quality and quantity of crops designated as teind. However, not J55 46 every heritor was entitled to purchase control over his own teinds. J55 47 Charles ratified his agreement of the previous summer with the J55 48 royal burghs that teinds assigned to ministers' stipends, J55 49 educational provision and social welfare could be led by but not J55 50 sold to heritors - a prohibition on purchase extended to the teinds J55 51 of universities and hospitals in 1633. Moreover, heritors had to J55 52 await the expiry of existing tacks before purchasing their teinds J55 53 from clerical titulars. When Charles eventually abrogated this J55 54 provision on 20 May 1634, churchmen not heritors were given the J55 55 first priority to buy out tacksmen and secure the whole teinds of J55 56 parishes for the permanent use of the Kirk. Heritors were also J55 57 required to pay teinds to tacksmen of temporal lords and lay J55 58 patrons with no immediate prospect of leading, far less of J55 59 purchasing, their own teinds. For valuations could not be concluded J55 60 until current tacks - usually long leases - had expired.

J55 61 While the right to purchase their own teinds was neither J55 62 guaranteed nor immediate for some heritors, it was not practicable J55 63 for others. The right to purchase was not so much a standing J55 64 concession as a singular option. The coronation parliament of 1633 J55 65 went on to impose specific time-limits on this option. All heritors J55 66 whose teinds had already been valued were obliged to make an offer J55 67 to purchase by the autumn of 1635. Those heritors awaiting J55 68 valuation of their teinds were given no more than two years to make J55 69 an offer on the completion of valuation. Because of a lack of ready J55 70 cash, few heritors outwith the ranks of the nobility and leading J55 71 gentry were able to exercise their option to purchase within two J55 72 years. Heritors were not entirely quit of superiority if they J55 73 purchased their own teinds. The formal contract of sale issued from J55 74 1631 required heritors to provide the titular of every parish with J55 75 a written undertaking not only to pay reliefs for taxation, but to J55 76 meet their share of parochial teinds apportioned to ministers' J55 77 stipends, other pious uses and the king's annuity.

J55 78 Where the teinds were intermingled with other landholding dues J55 79 and were thus to be valued conjointly, Charles determined that the J55 80 quantity to be assessed as teind should be the fifth part of the J55 81 valued or constant rent: that is, a fifth of composite sums paid as J55 82 feu-duty was to be accounted teind when determining the valued rent J55 83 of each estate. Where the teinds were assessed separately, Charles J55 84 allowed a deduction of one-fifth of their estimated value to be J55 85 retained as a personal allowance for the heritor. Basically, this J55 86 personal allowance from the valued rent was afforded because teind J55 87 quantified separately tended to carry a higher valuation than that J55 88 where teind was intermingled with other landholding dues. The J55 89 personal allowance was thus a practical corrective to titulars J55 90 profiteering from the loading of valuation in their favour. This J55 91 personal allowance primarily benefited barons and freeholders who J55 92 were now guaranteed a minimum share of their own teinds whether or J55 93 not they exercised their option to purchase.

J55 94 Indeed, the personal allowance was the only guaranteed portion J55 95 of the heritors's teinds not liable to redistribution. The legal J55 96 decreet upheld the revised minimum for ministers' stipends raised J55 97 by the Commission from 500 merks or 5 chalders of victual to 800 J55 98 merks or 8 chalders on 30 May 1627. While the amount requiring J55 99 redistribution within each parish varied, the national minimum was J55 100 an identified target. However, the contribution towards pious uses J55 101 was not limited by national ratings or current demands, but was J55 102 elastically geared to future local aspirations for schooling and J55 103 social welfare. Contributions to the king's annuity were fixed J55 104 nationally but variably in accordance with the Commission's J55 105 recommendation of 29 May 1627. The rating of 6 per cent to be J55 106 exacted as annuity applied only to payments of teind in money, not J55 107 in kind where exactions varied according to the type and quality of J55 108 victual and fluctuated annually according to local rates of J55 109 commutation. Given that a significant if indeterminate proportion J55 110 of teind was to be redistributed, the heritors' inducement to J55 111 purchase was not so much financial as managerial.

J55 112 Reflecting on the legal decreet on 24 October 1629, Charles J55 113 affirmed that he had provided the necessary guidelines for the J55 114 surrender of superiorities and teind redistribution "that J55 115 non of our subjectis interested can have just caus to J55 116 complain". The presentation of the legal decreet before a J55 117 Convention of Estates in July 1630 provided no ringing endorsement J55 118 from the political nation, however. The resistance encountered over J55 119 teind redistribution was such that leading officials were initially J55 120 averse to moving its ratification "for feir of J55 121 repulse". Only adroit presentation of the legal decreet as J55 122 a composite package by the lord president, William Graham, seventh J55 123 earl of Menteith, ensured acceptance. Nonetheless, the gentry, as J55 124 spokesmen not just for the shires but for the political nation as a J55 125 whole, called upon the Convention to petition the king to take J55 126 account of "the great feare" aroused by his efforts J55 127 to compel compliance with the Revocation Scheme.

J55 128 While Charles remained unresponsive to such petitioning, he had J55 129 already provided the touchstone for the furtherance of class J55 130 collusion. His quantification of teind as a fifth of the valued J55 131 rent, when intermingled with other landholding dues, was open to J55 132 manipulation by titulars and lay patrons with the active support of J55 133 heritors. Taking advantage of ties of social deference and the J55 134 loading of proof against heritors, titulars and lay patrons J55 135 concocted valued rents that devalued the amounts apportioned as J55 136 teind. The coronation parliament pointed out that valued rents were J55 137 undervalued by as much as a third in relation to feu-duties where J55 138 teinds and other landholding dues were paid as composite sums. J55 139 Undervaluing could not be accomplished, however, without the J55 140 connivance of the sub-commissions who were empowered to scrutinise J55 141 all landed titles and contractual conditions. Thus, collusion was a J55 142 community enterprise designed to diminish the amount of teind J55 143 available for redistribution for ministers' stipends, pious uses J55 144 and, above all, the king's annuity. By 1 July 1635, collusion had J55 145 become so prevalent that the Commission for Surrenders and Teinds J55 146 was itself censured by Charles for its routine acceptance of J55 147 valuations from sub-commissions in which valued rents had been J55 148 diminished by a third in relation to composite feu-duties. An J55 149 official committee of inquiry into the operation of the Commission J55 150 reported to Court by the end of December 1636 that "the J55 151 great ill of undervaluing" was the foremost impediment to J55 152 comprehensive teind redistribution. Moreover,"the farr J55 153 greater sort" of the teinds were "not yet J55 154 valued".

J55 155 Limited Accomplishment

J55 156 No more than seven temporal lords had commenced negotiations J55 157 with the Commission prior to the Convention of Estates in July J55 158 1630, which served to expose the critical weakness of the J55 159 Revocation Scheme - notably, the king's lack of financial resources J55 160 to implement the surrender of kirklands, regalities and heritable J55 161 offices. Prior to the implementation of his legal decreet, Charles J55 162 was already committed to expenditure of around pounds250,000 for J55 163 the compulsory purchase of four regalities and twelve heritable J55 164 offices. Yet the annual income ordinarily available to the Crown J55 165 during the first four years of his reign had dropped by J55 166 pounds27,321-14/4 (from pounds223,930-7/4 to pounds196,608-13/-) J55 167 and his expenditure had moved from a modest surplus of J55 168 pounds64,838-15/8 to an accumulating deficit of pounds137,640-7/6. J55 169 Furthermore, the revenues raised from the land tax of pounds400,000 J55 170 and the 5% tax on annualrents<&|>sic! voted by the Convention of J55 171 Estates in 1625, were earmarked for British expeditionary forces on J55 172 the continent, national defence and the maintenance of the Scottish J55 173 establishment at Court and in Edinburgh. Although both taxes were J55 174 renewed by the Convention of 1630, military and establishment J55 175 expenditure was still accorded priority.

J55 176 Charles's continuing financial embarrassment obliged him to J55 177 rely on feuars to buy out their own superiorities, a policy impeded J55 178 partly by legal loopholes arising from transfer of title, but J55 179 primarily because feuars were unprepared to advance the requisite J55 180 lump sums. Charles took four years to clear up the loopholes J55 181 arising from his legal decreet. As well as prohibiting temporal J55 182 lords questioning their feuars' titles at law, feuars were obliged J55 183 to produce documentary evidence that compensation, including any J55 184 arrears of feu-duty, had been paid in full before they received a J55 185 charter for their kirklands direct from the Crown. Actual, as J55 186 against threatened, legal action between temporal lords and their J55 187 feuars was rare. Conversely, bands of suspension raised between J55 188 August 1632 and January 1634 reveal that only the most prominent J55 189 gentry among feuars of kirklands had sufficient local standing or J55 190 independent resources to enforce their right to buy out J55 191 superiority. The tendency among most feuars was to collude actively J55 192 with temporal lords in exploiting legal loopholes. Thus, feuars J55 193 resigned their kirklands in favour of their temporal lords who then J55 194 registered the resigned lands as their property with the Commission J55 195 for Surrenders and Teinds. On the lands being declared exempt from J55 196 revocation, they were granted back to the feuars.

J55 197 The financial embarrassment of the Crown - which led Charles to J55 198 suspend negotiated surrenders of regalities and heritable offices J55 199 in October 1634 - was compounded by temporal lords withholding J55 200 payments of their feu-duties for over seven years. Only eleven J55 201 temporal lords, less than a third of the total number, had made any J55 202 meaningful effort to negotiate with the Crown. Although no temporal J55 203 lord was exempted from the scope of revocation, Charles had not J55 204 enhanced prompt compliance by his blatant favouring of courtiers. J55 205 No more than two temporal lords, both anglicised absentees - James J55 206 Stewart, fourth duke of Lennox and James Colville, Lord Colville - J55 207 were prepared to make unconditional surrenders. J55 208 J56 1 <#FLOB:J56\>In 1935 the historian G. M. Trevelyan called Edmonds a J56 2 historian. Edmonds took umbrage at that description. He insisted J56 3 that he was "but a G.S.O. writing a military account of a J56 4 modern campaign with the assistance of friends". The J56 5 friends he referred to were the thousands of correspondents, all J56 6 former participants in the events he was trying to describe, upon J56 7 whose co-operation he depended to supplement the official written J56 8 records.

J56 9 Edmonds produced several drafts of each book before he J56 10 dispatched the final typescript to the publisher. He began by J56 11 reading through all the pertinent documentary evidence, war J56 12 diaries, surviving operational orders, and any private papers which J56 13 participants in the events he was analysing had made available to J56 14 him, he made notes on them, and then produced a first draft. He J56 15 then checked through his notes once again to ensure that he had J56 16 omitted nothing of importance. But he was far too good a historian J56 17 to believe that all the evidence could be found in the documents or J56 18 for that matter that the documents were necessarily accurate. War J56 19 Diaries were often written up well after the events they purported J56 20 to describe and by officers who were not present at the time. J56 21 "Reports of operations written immediately after an J56 22 action", he wrote in 1931, "are of little value J56 23 except as a general guide".

J56 24 There were also occasions when he discovered that deliberate J56 25 attempts had been made to tamper with or suppress evidence. During J56 26 his period as Sir John French's Chief of Staff in 1914, Sir J56 27 Archibald Murray sometimes falsified the times and dates of orders. J56 28 But Edmonds was able to detect this because of the date and time J56 29 stamp which his confidential clerk had placed on the documents in J56 30 question. One of Edmonds's assistants discovered that Sir Henry J56 31 Wilson removed copies of operations orders from the GHQ files for J56 32 August 1914 because they reflected badly on the shaken state of J56 33 mind of the headquarters staff. In June 1917 Sir Henry Rawlinson's J56 34 Chief of Staff, Sir Archibald Montgomery-Massingberd, destroyed 4 J56 35 Army's War Diary for the opening weeks of the Battle of the Somme, J56 36 probably in the hope of concealing the fact that Rawlinson had been J56 37 reluctant to follow Haig's order and to seek a quick breakthrough J56 38 on 1 July. He substituted for it his own narrative of events. But J56 39 Edmonds detected this fraud by checking the narrative against the J56 40 evidence presented in the war diaries of subordinate formations. J56 41 Efforts were also made to omit evidence which might reflect badly J56 42 on the commanders in question. Thus the 3 Army war diary contained J56 43 no mention of a warning that the army had received about an J56 44 imminent German counter-attack at Cambrai on 30 November 1917. J56 45 Edmonds also had access to typescript copies of Haig's personal J56 46 diary and noted that not every entry had been written at the time J56 47 it purported to have been and that some passages had been inserted J56 48 later.

J56 49 The oral and written reminiscences of survivors were vital J56 50 because they enabled Edmonds to amplify and cross-check the written J56 51 record. He therefore had copies of his first draft made and J56 52 circulated them to as many participants as possible. Drafts of the J56 53 Loos volume were sent to about 700 officers, more than half of whom J56 54 made suggestions which Edmonds thought it worth while to consider. J56 55 Drafts of the volume which examined the opening of the German March J56 56 offensive in 1918, an action which was pre-eminently a soldier's J56 57 battle, seem to have been sent to just about every commander down J56 58 to battalion and battery level. Edmonds himself questioned J56 59 surviving senior officers over lunch. Some followed Haig's example J56 60 and lent him copies of their diaries or other private records. But J56 61 Edmonds did not treat the mass of information he gathered in this J56 62 way uncritically. He understood that memories were apt to lapse J56 63 after the passage of years and that his respondents might try to J56 64 put the best possible gloss on their actions. "It is the J56 65 duty of the historian to make head against these difficulties with J56 66 the aid of the documents and the evidence of other J56 67 witnesses." Edmonds did any necessary rewriting and J56 68 produced a second draft incorporating any relevant French, German, J56 69 or Italian material. That was then sent to the Dominion general J56 70 staffs and to the Australian official historian, C. E. W. Bean, for J56 71 their comments. Only then did he prepare a final draft.

J56 72 The danger of Edmonds's determination to tap every available J56 73 source of information was that it did lay him open to pressure to J56 74 doctor his conclusions to suit the wishes of his informants. If he J56 75 did not do so they might simply refuse further co-operation and the J56 76 geese which lay the golden eggs might drop dead. Many of his J56 77 informants were ambivalent about helping him. "Yes: we all J56 78 want to know the TRUTH, as you say", Sir Ivor Maxse, a J56 79 former divisional and corps commander, wrote to him in 1927, J56 80 "But who knows where to discover it? and, if discovered, J56 81 how to reproduce it? or whether it should be told?" J56 82 Some witnesses were undoubtedly concerned about their personal J56 83 reputations. In lending Edmonds some of his private papers Lawrence J56 84 asked in return that "if you are using anything which might J56 85 be controversial I should like to know what it is". Others J56 86 were less concerned about their personal reputations and were more J56 87 concerned about the effect criticisms of the high command might J56 88 have upon the willingness of the next generations to submit to J56 89 military discipline. For example, one former divisional commander, J56 90 Major-General A. Solly Flood, sent Edmonds a press cutting critical J56 91 of Haig together with a note saying:

J56 92 But setting aside all questions of loyalty to our J56 93 revered chief, articles couched in these terms cause one much J56 94 anxiety on account of the effect they must have on the minds of the J56 95 uninformed public and of posterity, which may be called upon in J56 96 their turn to undertake service on behalf of their King and J56 97 country.

J56 98 A third category were worried lest criticisms of the army's J56 99 conduct might harm Britain's international prestige. In 1929 Sir J56 100 Arthur Conan Doyle raised the issue of Edmonds's description of the J56 101 precipitous retreat of 21 Division at Loos in September 1915: J56 102 "Very frankly. Could it be toned down into more general J56 103 terms?" he asked, "You have to remember that there J56 104 are fellows of the baser sort in America & elsewhere who will leave J56 105 out all your compensating paragraphs and simply quote the passages J56 106 which describe the panic. Then there may be a row." J56 107 Similarly, Sir George Macdonogh warned him that it might be unwise J56 108 to include criticisms of General P<*_>e-acute<*/>tain's conduct in J56 109 1918. "It has got to be remembered that not merely is he J56 110 still alive but he is also at present the French Minister of J56 111 War." Edmonds generally met his critics at least half-way. J56 112 He dealt with pressures like these in one of two ways. He confined J56 113 the text to a bald narrative of events and omitted criticisms or he J56 114 placed awkward evidence in footnotes or appendices, perhaps on the J56 115 assumption that only those already in the know would bother to read J56 116 them. In the volume published in 1927 examining the second Battle J56 117 of Ypres in 1915 he agreed to Sir Arthur Currie's request that he J56 118 should suppress the fact that three times during the battle he had J56 119 ordered his Canadians to retreat. When he exposed one of Murray's J56 120 attempts to change the date and time of a GHQ operations order in J56 121 August 1914 he did so in a footnote to an appendix. Similarly, when J56 122 Montgomery-Massingberd complained that the first draft of the J56 123 volume on the battle of the Somme had been written in a J56 124 "captious spirit" he agreed to revise it and tone J56 125 down the criticisms. Only by turning to the volume of appendices J56 126 could a reader discover that during the planning stage of the Somme J56 127 Haig and Rawlinson had disagreed fundamentally about whether 4 Army J56 128 should attempt to break clean through the German defences on the J56 129 first day of the battle. In 1937 he informed Liddell Hart that J56 130 "In my first draft [of OH, ii. (1918)] I took the same J56 131 view as you do to the 'backs to the wall' order (? where was the J56 132 wall), but found that hundreds of my correspondents in the fighting J56 133 units were against me and I felt bound to alter it."

J56 134 Just occasionally he let himself go when his targets could not J56 135 retaliate or when the need to stay silent had passed. In the first J56 136 edition of the volume examining the retreat from Mons, published in J56 137 1922, he made no mention of the panic which for a time seems to J56 138 have gripped the headquarters of Haig's 1 Corps and led to it J56 139 diverging from Smith-Dorrien's 2 Corps. But in 1933, five years J56 140 after Haig's death, he published a revised edition in which he J56 141 described this incident and added that "Haig momentarily J56 142 lost his head - a remarkable lapse for so stout hearted a J56 143 fighter." There was, as he laconically noted "a gap J56 144 in the records" about this incident. Similarly, by 1947 J56 145 enough had happened to remove any inhibitions he might have felt J56 146 about criticizing P<*_>e-acute<*/>tain. In the 'Retrospect' to the J56 147 final volume in the series he felt free to write that J56 148 "After General P<*_>e-acute<*/>tain was given command of J56 149 the French Armies on 15th. May 1917, our Allies did very J56 150 little", and that his refusal to reinforce Gough's 5 Army J56 151 in March 1918 "very nearly brought about a J56 152 disaster".

J56 153 Edmonds's work received a mixed reception both from J56 154 participants and from later historians. Some readers liked what he J56 155 had written. Haig read and commented upon all the volumes up to and J56 156 including Loos before he died. "They are all J56 157 excellent" he wrote of some draft chapters on Neuve J56 158 Chapelle, "and I congratulate you on the way in which you J56 159 have told the story so accurately, and yet without attaching J56 160 blame to anyone...". More surprisingly this camp also J56 161 included Field Marshal Lord Wavell, who reviewed several volumes in J56 162 the Times Literary Supplement. In 1949 he described the J56 163 series as "the best official military history yet J56 164 written" possibly because he shared Edmonds's view that J56 165 many British generals in France and Flanders had believed that they J56 166 were engaged in "Open Warfare at the Halt" but in J56 167 reality "warfare on the Western Front after 1914 was Siege J56 168 Warfare and should have been treated as such".

J56 169 However, some readers thought that he had gone too far to meet J56 170 the wishes of his informants and that his books were anodyne. Sir J56 171 William Robertson read one draft and wrote that "It leaves J56 172 the same taste behind it as when one drinks skimmed milk. I suppose J56 173 he wished to avoid having trouble with anyone as to what he says, J56 174 and therefore leaves out most of what is worth saying." And J56 175 thirdly, there were those like Liddell Hart, whose view of the high J56 176 command's respect for the truth was very largely soured by the fact J56 177 that Edmonds often told him what he really believed in private but J56 178 would not say the same thing in print. Liddell Hart believed that J56 179 Edmonds's practice of covering up the deficiencies of the high J56 180 command meant that he was only storing up trouble for the future. J56 181 "Any writer of history" he informed Edmonds in June J56 182 1934, "who helps to flatter the 'brasshat's' self-delusion J56 183 as to the respect in which he is held, is preparing the country for J56 184 a greater disaster".

J56 185 His books were a qualified success when measured against the J56 186 self-imposed task of instructing young officers but J56 187 concealing the worst mistakes of their seniors. Some of the J56 188 evidence which Edmonds had concealed remained hidden for a long J56 189 time. As late as 1952 the publication of Robert Blake's The J56 190 Private Papers of Douglas Haig could still evoke a sense of J56 191 shock amongst those not already privy to what one reader of them J56 192 called "the petty jealousies that pervaded high J56 193 places". In 1932 Lord Gort told Edmonds that if only the J56 194 next generation of officers read his books they would avoid some of J56 195 the worst mistakes of their predecessors. J56 196 J57 1 <#FLOB:J57\>France, Russia and America concluded versions of it. J57 2 The resumption of war in 1859 had hardened Palmerston's outlook; he J57 3 dismissed apprehensions that the fall of Peking might subvert the J57 4 Chinese empire, and disrupt trade. The aftermath produced his J57 5 public conversion, in language greeted with incredulity: Britain J57 6 was morally bound to rehabilitate the central authority she had J57 7 undermined; hence the deployment of her troops to counter the J57 8 Taiping rebels. It should not be forgotten that Lewis's argument J57 9 against the coercion of both China and Japan in 1857 impressed J57 10 Palmerston. The de facto protectorate that Lewis feared J57 11 became a reality, while the Indian Mutiny inspired profound J57 12 misgivings at the thought of responsibility for another huge J57 13 Oriental dependency. The Chinese empire was appreciably weaker than J57 14 the Ottoman, but as in the latter, a network of interlocking J57 15 agreements spread the burden among Western nations, and restrained J57 16 others from too much interference harmful to Britain's local J57 17 economic superiority. Although Japan's external trade was a J57 18 fraction of China's, Palmerston attached similar importance to J57 19 international co-operation there, when it came to bombarding the J57 20 Japanese into compliance with the letter and spirit of agreements J57 21 which granted extraterritoriality and designated treaty ports on J57 22 the Chinese pattern. The Anglo-Japanese treaty of Yedo could not be J57 23 directly attributed to force, unlike that of Tientsin in the same J57 24 year, "unless", observed Lewis drily, "the Emperor J57 25 of Japan is influenced by the example of his brother of J57 26 China".

J57 27 The fiercer Japanese reaction to Europeans coming among them J57 28 resulted in a succession of Western naval operations, the largest J57 29 of which assembled seventeen warships - eight British, the rest J57 30 French, American and Dutch - to silence the batteries guarding the J57 31 Straits of Shimonoseki at the entrance to the Inland Sea. The J57 32 United States Navy was represented, owing to the civil war in J57 33 America, by a small vessel under the command of a junior officer, J57 34 for whom Palmerston wanted the Order of the Bath that was to be J57 35 conferred on senior allied officers in the action. A beneficiary of J57 36 China's defeat at the hands of Britain and France, the United J57 37 States had not, after all, been a belligerent power in the J57 38 conflict; although units of her navy did come under and return J57 39 Chinese fire. Her insignificant naval participation at Shimonoseki J57 40 should be marked, the prime minister argued, to draw her into more J57 41 active promotion of "the common cause". The civil J57 42 war growth of the American fleet, and speculation about its future J57 43 employment, reinforced the case for partnership with her in the Far J57 44 East, where her traders and missionaries were busy. Co-operation J57 45 helped to disarm critics at home. Cobden and Bright found it J57 46 difficult to attack the United States, while Gladstone hoped their J57 47 joint efforts in China would offset Anglo-French discord in Europe. J57 48 Interrogated by a friend, he confessed to the "misery ... J57 49 it is not too strong a word" of having to endorse decisions J57 50 that flowed from a policy and the commencement of war which he J57 51 still condemned: "Often have I tried to cure or mitigate J57 52 the system of parading force which I conceive to be the root of the J57 53 evil." Later, when the question of coercing Japan arose he J57 54 asked only that the resort to force should be "short". J57 55 The British bombardment of Kagoshima, which unintentionally burnt J57 56 down the town, revived moral unease in Parliament. Ministers, said J57 57 Layard in his successful explanation, conformed to the wishes of J57 58 backbenchers, "who were forever telling the Government that J57 59 it was their duty to open fresh markets all over the J57 60 world". The Kagoshima debate "went well", J57 61 wrote Gladstone in his diary, reconciled, politically, at any rate, J57 62 to the element of violence in commercial empire.

J57 63 Lord Elgin, who recorded his distaste for much that he had to J57 64 do in the Far East, put it to Gladstone, doubtfully, that J57 65 compulsion on the Chinese to follow their strong trading instincts J57 66 without hindrance amounted to liberation of a sort: "This J57 67 may hardly be a justification of such force, but it affords some J57 68 grounds for a new and better order of things." Trade J57 69 improved; but little else. If legalizing the import of opium into J57 70 China from India, to whose finances it was so important, ended J57 71 abuses connected with smuggling the drug before 1858, that legality J57 72 did not still the prickings of conscience, even in mercantile J57 73 circles handling the commodity. The value of opium was still about J57 74 twice that of imports from Britain in 1865, evidence of China's J57 75 intractability to Westernization, as of India's in a lesser degree. J57 76 The illusion of trade as civilizing in itself had begun to fade by J57 77 then. Palmerston was conscious of the subtle change and its J57 78 damaging implications. To call British policy in the Far East J57 79 'selfish' was to misunderstand its main concern, fundamentally the J57 80 same anywhere, the accessibility of markets. Government had an J57 81 absolute duty to expand the nation's trade "by every means J57 82 in its power ... and so to render our own people happy and J57 83 prosperous at home". That great purpose was not to be J57 84 confused with the sectional interests of British merchants J57 85 overseas. Those who challenged his definition of the national J57 86 interest in global commerce were "doing their best to take J57 87 the bread out of the mouths of our J57 88 working-classes"; a statement that wrung a cry of J57 89 protest from Bright. "I do not admit", said J57 90 Palmerston, "... that our policy has been a selfish one in J57 91 the sense in which the word is sometimes employed." The J57 92 welfare of the country and the classes could not be sacrificed to J57 93 international altruism which had no place in the real world.

J57 94 Morocco, unlike China and Japan, produced talk of resignation J57 95 in the cabinet. The Anglo-Moroccan commercial treaty of 1856, which J57 96 broke down the country's isolation, had a strategic motive. Unable J57 97 to prevent Spain's invasion in 1859-62, warranted by the attacks of J57 98 Riffian tribesmen on her enclaves in the North and supported as it J57 99 was by the European Great Powers, Britain came to the Sultan's J57 100 rescue with diplomacy and finance. She objected to any significant J57 101 cession of territory and arranged a loan for him in London towards J57 102 the indemnity on which the Spanish evacuation of the fortress city J57 103 of Tetuan was conditional. The cabinet twice considered J57 104 guaranteeing the loan to attract investors, and decided against it. J57 105 Russell, with some encouragement from the prime minister, proposed J57 106 to ignore the collective opinion after his colleagues had J57 107 dispersed. He or Russell, Gladstone believed, might have to resign; J57 108 but the foreign secretary dropped the idea for the time being. J57 109 "So that is over", the chancellor told his wife in J57 110 January 1861. Later that year he had second thoughts, and ideally J57 111 wanted a joint Anglo-French guarantee: "Thus ... mutual J57 112 jealousies would be neutralized." There was reason to fear J57 113 France behind Spain. Morocco's trade, which Britain dominated, J57 114 could not compare with that of Turkey. The case for a loan was far J57 115 more political and military than economic. The cabinet refused to J57 116 see it in November 1860 as "preventive", afraid of J57 117 taking on, at a difficult juncture in the affairs of Europe, a J57 118 tangible commitment to the preservation of Moroccan independence. J57 119 Gladstone did not contest Palmerston's insistence on the value of J57 120 her independence to the British position in Gibraltar, and J57 121 therefore in the Mediterranean. Neither man thought it probable J57 122 that Parliament would agree to a loan on its security, which it J57 123 regarded jealously.

J57 124 Layard, who arrived at the Foreign Office in the summer, acted J57 125 as intermediary between the government and the City, relaying Baron J57 126 Rothschild's advice that "a virtual guarantee" J57 127 would be needed to get the money on the right terms. Palmerston J57 128 wrote a long memorandum on the arrangement that Rothschild wanted. J57 129 According to the prime minister, it involved only "a shadow J57 130 of responsibility" and avoided recourse to Parliament. In J57 131 fact, it involved a substantial responsibility without J57 132 Parliamentary sanction. Under the convention with Morocco, a J57 133 British commissioner, whose duties the minister resident and J57 134 consuls discharged, received half the total customs revenue on J57 135 which the loan was secured; subtracted interest and sinking fund J57 136 payments; and returned the surplus to the Moroccans. Gladstone felt J57 137 the additional functions given to Britain's representatives might J57 138 very well lead to further interference in the country and J57 139 Anglo-French dissension. Lewis disliked the evasion of J57 140 Parliamentary sanction by the absence of a formal guarantee, when J57 141 investors supposed, as they were meant to do, that government would J57 142 not see them lose their money. In his memorandum, Palmerston J57 143 anticipated that if the Sultan failed to meet these obligations, J57 144 "we as a maritime power ... have ample means without J57 145 trouble or exertion ... to compel him to stand by his J57 146 bargain". The cabinet reluctantly agreed to the convention; J57 147 "an impolitic measure", Lewis called it. J57 148 Parliament, apparently, had few questions. The City was distinctly J57 149 pleased with the loan, four times oversubscribed, "both as J57 150 a political and a financial operation". With official J57 151 British sponsorship, the Sultan paid 5 per cent on his bonds; J57 152 without it, the Pasha of Egypt had recently had to pay 10 per J57 153 cent.

J57 154 Palmerston's personal crusade against the slave trade finally J57 155 succeeded when it did only because states too powerful to coerce J57 156 abandoned policies that would have ensured its defeat. Secession, J57 157 civil war and Lincoln dispersed the Southern lobby in Congress. J57 158 France's revival of the traffic under the guise of 'free J57 159 emigration' lent it a veneer of respectability. Slaves purchased on J57 160 the coasts of Africa landed in French sugar colonies as indentured J57 161 labour, with the prospect of freedom to go home, in theory, on the J57 162 expiry of their contracts. France wanted to be able to tap the J57 163 supply of coolies available from India to British territories as J57 164 voluntary substitutes for freed negroes. Britain's objections J57 165 rested on her regulations governing the supply of indentured labour J57 166 and its treatment, and on the partial diversion of the flow that J57 167 was none too large from her own sugar colonies. The French were, J57 168 they made it clear, driven to the expedient of 'free emigration' J57 169 from Africa. Talking to Palmerston in November 1857, Napoleon's J57 170 minister of state, Fould, maintained, not altogether J57 171 disingenuously, that their captives or subjects were all that J57 172 African chiefs had to sell in exchange for European goods. They J57 173 certainly had nothing else so valuable. Fould did not deny, J57 174 speaking privately, that his country was dealing in slaves. The J57 175 French government, throughout, contended otherwise; this official J57 176 pretence exasperated Palmerston: "it is not a question of J57 177 law but of common sense", he remarked. Disturbing in J57 178 themselves, the open activities of the French emboldened J57 179 clandestine slavers of various nationalities. French vessels J57 180 engaged in recruiting 'free emigrants' were fitted out as slavers. J57 181 Two earned notoriety. The human cargo of the misnamed Regina J57 182 Coeli, lying off the West African shore, staged a bloody J57 183 revolt, known to the world after the ship was found drifting and J57 184 taken into Freetown harbour. When Palmerston, out of office, saw J57 185 the Emperor and Walewski in Paris some months later, they talked J57 186 about the revived French slave trade, exposed for what it was.

J57 187 Within weeks, two French battleships entered the Tagus and J57 188 secured the release of the Charles et Georges, arrested J57 189 in Mozambique waters and brought to Lisbon, whose captain alleged J57 190 that he had been picking up "colonists". The British were J57 191 severely embarrassed by their inability to help Portugal, whom they J57 192 stood treaty-bound to aid; but they left her to suffer for having J57 193 carried out the general anti-slavery policy adopted at Britain's J57 194 instance and overseen by her. Malmesbury, the foreign secretary of J57 195 the moment, consulted Palmerston, and instructed Cowley at Paris to J57 196 transmit what he had said to the ex-premier, soon to visit the J57 197 Emperor at Compi<*_>e-grave,<*/>gne about the danger of war and J57 198 France's neglect of arbitration over the Charles et J57 199 Georges. The British ambassador was enjoined to tread J57 200 delicately, "but we cannot pass over the event in J57 201 silence". Napoleon lost no time in moving towards the J57 202 British, and exploiting the nervousness of the Tory government. J57 203 Negotiations on the availability of coolie labour had been J57 204 initiated in January 1858 when Palmerston was still prime minister J57 205 and following the candid discussion with Fould. "The French J57 206 are, I believe, excessively anxious not only to get the thing done, J57 207 but to get it done immediately", wrote the civil J57 208 servant who had talks with Persigny, the ambassador in London, that J57 209 were cut short by the change of ministry. J57 210 J58 1 <#FLOB:J58\>It was the second total war within a generation J58 2 that finished off for the then-foreseeable future the particular J58 3 form of politico-economic dispensation which the Conservatives had J58 4 come to control so firmly by the 1930s. The Conservatives of the J58 5 time could hardly be said to have been entirely in the grip of J58 6 economic orthodoxy of the kind preached by Alfred Marshall, J58 7 authority in economics before Keynes. While they had little choice J58 8 about going off the gold standard, the Tories had dispensed with J58 9 free trade, and behind the shield of protection they showed that J58 10 they were no slaves to free-market ideology by engaging in J58 11 substantial intervention in industrial organization in order to J58 12 ameliorate unemployment. The National Government was adventurous in J58 13 the economic sphere by previous peace-time standards, but what J58 14 restrained it was the only tenet of Gladstonian Liberalism which it J58 15 did, if sometimes only emotionally, adhere to - the balanced J58 16 budget, balanced preferably at low levels of taxation and J58 17 expenditure. In the 1920s, Neville Chamberlain had developed J58 18 Beveridge-style social security plans, but had drawn back at the J58 19 cost. The means test was more in accord with the maxims of 'sound J58 20 finance' to which few contemporaries did not subscribe. To borrow J58 21 the Dickensian allusions used in Chamberlain's 1934 Budget speech, J58 22 the Conservatives could prosper politically in what was later J58 23 depicted as being Bleak House (although many, perhaps most, did not J58 24 see it at the time as being bleak at all) because the electorate J58 25 did not have great expectations from the state. The Second World J58 26 War changed that. To the neo-classical question, 'Where is the J58 27 money going to come from?' the answer became: 'They always seem to J58 28 find it for wars, so they can find it during peace.' Hence, funding J58 29 the Beveridge plan was supposed to be no problem. Moreover, Keynes J58 30 seemed to be saying that high levels of public spending were a J58 31 guarantee of peace-time full employment. It was not until the war J58 32 that Keynes's ideas were acted upon - in the 1941 Budget to be J58 33 precise. There is now no compelling reason to believe that adopting J58 34 his ideas - which kept changing and which were not fully set out J58 35 until The General Theory was published in 1936 - would J58 36 have meant a cure for the unemployment of the 1930s. As for the J58 37 foreign examples that the National Government was supposed to have J58 38 foolishly ignored, the evidence about the effectiveness of the J58 39 various New Deals in bringing down unemployment now seems, at best, J58 40 to be inconclusive. Nevertheless, in the period immediately J58 41 afterwards, those who either believed, or affected to believe, that J58 42 'something could have been done' about the large-scale unemployment J58 43 were able to rewrite recent British economic history to the J58 44 detriment of the Conservative Party. As one historian later J58 45 remarked, "that Jarrow, rather than Coventry, should still J58 46 be the symbol for a whole decade points to a propaganda victory the J58 47 Left won not in the thirties but in the forties."

J58 48 This propaganda victory was made possible by the fact that J58 49 Britain, while not suffering disproportionately during the J58 50 Depression, did subsequently have a different experience from most J58 51 other European countries in that she was one of the few which did J58 52 not suffer the greater hardships of occupation and/or military J58 53 defeat. For those searching for grievances in British politics, the J58 54 hardships of the 1930s were the best available. The peculiar J58 55 war-time atmosphere in Britain may have been a particularly fertile J58 56 one for the development of grievances. For there was all the J58 57 disruption of total war, and, yet, from Britain's expulsion from J58 58 continental Europe in 1940 down to her return four years later on J58 59 D-Day, it was the minority of her armed forces that had to fight. J58 60 The North African campaign was important to Britain's morale, and J58 61 El Alamein was a signal victory, but it was a sideshow compared J58 62 with the Eastern Front. Hence, except for the brave and the J58 63 unlucky, there was plenty of time for the Army Bureau of Current J58 64 Affairs to do its work, and for events like the Cairo Parliament. J58 65 After all, Nazi Germany was not stopped by Churchill's speeches J58 66 ("some chicken, some neck" and so on). The Russians J58 67 bled Nazi Germany white. Evelyn Waugh, in the guise of Guy J58 68 Crouchback, was dismissive of the Soviet contribution and of King J58 69 George VI's gift of the Sword of Stalingrad. He was able, too, to J58 70 resist the image of Stalin as 'Uncle Joe'. Others perhaps were less J58 71 resistant, especially as once Nazi Germany had attacked her former J58 72 Soviet ally, the British official propaganda machine actively J58 73 promoted a favourable image of Stalin's tyranny. A Red Army Day was J58 74 held in 1943. The Ministry of Information went so far as to advise J58 75 George Orwell's publishers against bringing out Animal J58 76 Farm. It seems likely that the Russian war achievement J58 77 reflected favourably on Labour: the Soviet Union was the J58 78 socialist country. Moreover, when it came to Britain's evident J58 79 unpreparedness for war in 1939, as the title of one infamous J58 80 pamphlet would have it, the Guilty Men were all supposed J58 81 to be the Conservative ministers of the 1930s, and if the pamphlet J58 82 was even one-quarter as effective as propaganda as it is J58 83 commonly assumed to have been, it may well have helped to pave the J58 84 way for 1945, in deflecting attention from Labour's undistinguished J58 85 record on defence and pinning the blame on the 'Old Gang'. If so, J58 86 it helped to bring down Churchill, one of the few who were 'not J58 87 guilty', almost all of them Tories.

J58 88 Even the Conservatives' expectation that they would benefit, as J58 89 they had in 1918, by being associated with a war-winning leader J58 90 proved false. The situation was different. In 1918, Lloyd George J58 91 led the war-time Coalition against the rest. In 1945, Churchill led J58 92 the Conservatives against a Labour Party whose major leaders had J58 93 been prominent members of Churchill's war-time administration: J58 94 Ernest Bevin, Herbert Morrison, Stafford Cripps and Attlee himself. J58 95 Apart from Churchill, Labour had held the key positions: Bevin as J58 96 Minister of Labour, Morrison as Home Secretary and Attlee often J58 97 effectively the prime minister while Churchill ran the war. The J58 98 Foreign Secretary had been a Tory, Eden: but Churchill was often J58 99 his own Foreign Secretary. A Conservative in name or spirit was J58 100 always Chancellor of the Exchequer: but concern about paying for J58 101 wars is greatest once they are over. If the Conservatives secured J58 102 some marginal advantage from parading as a National Government in J58 103 1935 - and remembering the scale of their 1924 victory, for J58 104 example, it need only have been marginal - it may have been very J58 105 much less than Labour derived in 1945 from its association with the J58 106 war-time National Government. The charge of inexperience and J58 107 incapacity could no longer stick. Labour also possibly derived some J58 108 advantage from believing that they would lose the post-war election J58 109 anyway. It made them less inhibited about promising the New J58 110 Jerusalem. When the Conservatives held back in 1943 from a J58 111 commitment to the Beveridge plan, the Labour backbenchers pressed J58 112 for a division. James Griffiths said to Beveridge: "This J58 113 debate, and the division, make the return of a Labour Government to J58 114 power at the next Election a certainty." Possibly it did, J58 115 but Griffiths was one of the few Labour politicians who saw J58 116 this.

J58 117 The 1945 general election would have been even more remarkable J58 118 than it was if the Conservative Party had displayed a monopoly of J58 119 virtue in conducting its campaign. Churchill, though, combined J58 120 promises about executing his 'four year's<&|>sic! plan' to extend J58 121 state social provision with a commitment "to make an early J58 122 reduction in taxation". The Liberals similarly combined J58 123 pledges to extend public expenditure with "a progressive J58 124 reduction in the burden of taxation", and Labour promised J58 125 "social provision against rainy days, coupled with economic J58 126 policies calculated to reduce rainy days to a minimum", and J58 127 "taxation which bears less heavily on the lower income J58 128 groups". Churchill, however, not entirely consistently, J58 129 also reminded the electors that:

J58 130 the State has no resources of its own. It can only J58 131 spend what it takes from the people in taxes or borrowing. Britain J58 132 is now a nation of taxpayers ... The willingness of this generation J58 133 to bear their fair share of sacrifices must, though we hope for J58 134 relief, be continued ... the plans for social progress which we are J58 135 determined to carry out, cause and require a much higher J58 136 expenditure than before the war. The burden must be borne by all J58 137 citizens as taxpayers. There is no easy way of one section getting J58 138 great benefits from the State at the expense of another. The nation J58 139 can have the services it is prepared to pay for. Where all benefit, J58 140 all will have to contribute. The revenue is not created by waving a J58 141 magic wand.

J58 142 These economic liberal sentiments were always unlikely to be J58 143 music to the ears of an electorate then being told that a largely J58 144 cost-free New Jerusalem was on offer as the alternative. J58 145 Retrospectively, it is not unreasonable for a historian to write J58 146 that:

J58 147 Britain came out of the Second World War as an J58 148 obsolescent industrial economy with grievous weaknesses. Instead of J58 149 first devoting all possible resources and effort to remedying this, J58 150 she chose to load this economy with the vast and potentially J58 151 limitless cost of the Welfare State, placed current expenditure J58 152 before capital investment, and thus set the pattern of the next J58 153 thirty years.

J58 154 It may well be that "by the time they took the bunting J58 155 down from the street after VE-Day and turned from the war to the J58 156 future, the British in their dreams and illusions and their J58 157 flinching from reality had already written the broad scenario for J58 158 Britain's post-war descent." In a liberal democracy, J58 159 though, the electorate had to be persuaded of the necessity for J58 160 some form of programme of 'national efficiency' involving, as it J58 161 seemed very likely to do, high social costs with widespread J58 162 unemployment in what it would have to be hoped would be the interim J58 163 period, while awaiting problematical long-term benefits for the J58 164 economy. Even dressed up, this would have been an unlikely message J58 165 to win sufficient support in 1945 in competition with the more J58 166 comfortable attractions of the Keynesian Welfare State, which do J58 167 not lack electoral appeal even now.

J58 168 "They have had a very hard time" was J58 169 Churchill's magnanimous verdict on the electorate which had turned J58 170 him out of office, and the story has been told many times of how, J58 171 under his leadership, the Conservative Party then modernized itself J58 172 to become between 1945 and 1951 "the most effective J58 173 Parliamentary Opposition of the century". Much has tended J58 174 to be made, not least by that politician's official biographer, of J58 175 R. A. Butler's work at the Research Department in restoring Tory J58 176 fortunes:

J58 177 It was he ... who cut the Party afloat from its 1930s J58 178 unemployment moorings; it was he who had brought it to accept the J58 179 concept of the Welfare State; it was he perhaps above all, who made J58 180 sure that, if the voice of the Tory turtles was still heard in the J58 181 land, they were henceforth regarded as mere historical J58 182 curiosities.

J58 183 Easily perceived electoral realities in fact dictated the J58 184 Conservatives' behaviour after the 1945 defeat. Labour were J58 185 "the masters of the moment and not only of the moment but J58 186 for a very long time to come", Sir Hartley Shawcross told J58 187 them from the government benches. Churchill was not alone in J58 188 feeling that there was "some disgrace in the size of the J58 189 majority" that Labour had obtained. It took little insight J58 190 to appreciate that, in immediate retrospect and for many years J58 191 afterwards, the popular conception of the interwar era seemed to be J58 192 less of The Lost Weekend than a mood of Goodbye To J58 193 All That. There was certainly no 'back to 1939' sentiment to J58 194 parallel that of 'back to 1914' in the 1920s. With an electoral J58 195 base of 39.8 per cent of the vote in 1945, in reality the J58 196 Conservative position always was far from being hopeless. Given J58 197 what the war-time Coalition Government had done and planned for J58 198 with its concurrence and its related promises in 1945, the J58 199 difficulty of the Conservative Party was not 'to accept the concept J58 200 of the Welfare State', or, come to that, Keynesian economics, but J58 201 to reassure the electorate that the clock would not be turned back. J58 202 J59 1 <#FLOB:J59\>The Lambeth Freehold Land Society enjoyed as its patron J59 2 one of the most persistent campaigners for electoral reform, Locke J59 3 King.

J59 4 The brightest constellation of Liberal patrons was to be found J59 5 on the board of the National Freehold Land Society. It was J59 6 originally conceived in 1849 as an offshoot of the Metropolitan J59 7 (later National) Parliamentary and Financial Reform Association, a J59 8 body which had involved both Francis Place and Feargus O'Connor in J59 9 its early months. In practice, the council of the Association J59 10 became dominated by former Anti-Corn Law League activists. It was J59 11 this body which delegated twelve of its members to form the initial J59 12 steering committee of what, following the style of the parent J59 13 organization, was at first titled the Metropolitan and Home J59 14 Counties Freehold Land Society. Among the twelve were Sir Joshua J59 15 Walmsley (president of both the Association and the Society); J59 16 Cobden; Joseph Hume, the promoter of 'the Little Charter'; Samuel J59 17 Morley; and Gilpin. John Bright joined as director a year later, J59 18 having initially devoted his energies to an abortive 'Commons' J59 19 League' (whose objectives likewise included the exploitation of the J59 20 forty-shilling freehold franchise).

J59 21 The freehold land movement thus gave the impression, as George J59 22 Thompson remarked, of being "a confederacy mainly of J59 23 Liberal politicians". By any standards the contingent of J59 24 Liberal parliamentarians involved was both sizeable and important, J59 25 and it provides perhaps the fullest index of the movement's initial J59 26 political seriousness. For Cobden especially, freehold land J59 27 societies were more than merely a vehicle for stiffening the J59 28 Liberal spine of county electorates: "In proportion as this J59 29 40s. freehold movement made progress, in the same proportion would J59 30 they find the votes of the House of Commons on all liberal J59 31 questions would make progress." Their presence, and J59 32 especially that of Cobden, should be seen in the context of that J59 33 sense of isolation and frustration experienced by Anti-Corn Law J59 34 Leaguers in the decade following repeal. As McCord some time ago J59 35 observed, "the disappearance of Corn as an immediate J59 36 political issue put the initiative in British politics back firmly J59 37 in the hands of the established parties." It was this cadre J59 38 which pushed for the formation of an umbrella body, the Freehold J59 39 Land Union, and for a newspaper, The Freeholder, which J59 40 like The League before it was distributed gratis to the J59 41 influential as well as sold to the faithful.

J59 42 Yet these moves to place the freehold land movement on a J59 43 professional footing similar to the Anti-Corn Law League were also J59 44 an indication of a failure of political imagination. Cobden J59 45 especially was incapable of thinking in terms of a wider reforming J59 46 movement: his lukewarm response to the Commons' League, with its J59 47 broad parcel of reforming policies, suggests a loss of political J59 48 nerve, whilst his fixation that the freehold land movement could J59 49 reinstate him and his followers at the pivotal point of political J59 50 action reveals a misreading of both the mood of the working class J59 51 and its capacity for action on the issue. "The J59 52 forty-shilling scheme alone will not do the work, and J59 53 alone it will not work extensively," wrote Bright to J59 54 Cobden, "I think you exaggerate the extent to which people J59 55 will adopt the system, especially as no definite object or battle J59 56 is before them." In a private letter, the Sheffield MP John J59 57 Arthur Roebuck was more forthright: "Cobden is a poor J59 58 creature, with one idea -the making of county voters. He is daunted J59 59 by the country squires, and hopes to conquer them by means of these J59 60 votes."

J59 61 If Cobden intended the freehold land movement to absorb J59 62 political energies largely baffled since Repeal, he also meant it J59 63 to serve the same function as the League in clothing an attack on J59 64 the landed interest:

J59 65 I want it as a means to all that we require, and upon J59 66 my conscience it is, I believe, the only stepping-stone to any J59 67 material change. The citadel of privilege in this country is so J59 68 terribly strong, owing to the concentrated masses of property in J59 69 the hands of the comparatively few, that we cannot hope to assail J59 70 it with success unless with the help of the propertied classes in J59 71 the middle ranks of society, and by raising up a portion of the J59 72 working class to become members of the propertied order; and I know J59 73 of no other mode of enlisting such co-operation but that which I J59 74 have suggested.

J59 75 The requisite "portion of the working class" J59 76 Cobden defined as "teetotallers, non-conformists and J59 77 rational Radicals". Whatever its shortcomings as a grand J59 78 political strategy, the participation of prominent advanced J59 79 Liberals in the movement visibly set a seal of approval on the J59 80 working-class ethic of self-help, whilst avoiding any J59 81 offensive suggestion of charity or patronage. Much of the J59 82 movement's significance lay in its extending that code of J59 83 self-improvement, which had hitherto been largely confined to the J59 84 labour movement, into the new context of liberalism.

J59 85 "The citadel of privilege" was assailed both at J59 86 the polls and through the routine functioning of the societies. J59 87 Their very modus operandi advertised the virtues of free J59 88 trade in land. Supporters argued for the social necessity of the J59 89 sale and dispersal of landed property unhindered by the laws of J59 90 settlement, primogeniture, mortmain or entail. In the counties, J59 91 though few electoral victories could positively be ascribed to the J59 92 movement, societies formed the nucleus for constituency J59 93 organization and broadened awareness of Liberal policies and J59 94 personalities.

J59 95 There are also grounds for suggesting that the movement to some J59 96 extent did undermine the confidence of landed interests. At its J59 97 simplest, evidence of this can be found in proceedings before the J59 98 revising barristers' courts, where the franchises created through J59 99 theses societies were vigorously contested by protectionist J59 100 interests. The quality of land, building construction, and services J59 101 such as mains drainage, were all disputable when a franchise was J59 102 initially registered, and most societies took care to retain the J59 103 services of both builders and surveyors to attest on such issues. J59 104 Still more indicative of Tory disquiet was the formation in 1852 of J59 105 a "defence Society (for it really was a defence movement on J59 106 our part, and we were driven to it)". The Conservative Land J59 107 Society copied the very procedures of the societies it opposed, and J59 108 like the National and other municipal societies developed large J59 109 suburban estates, similarly evolving into a permanent building J59 110 society:

J59 111 FREEHOLD LAND SOCIETIES have done good service to the J59 112 Radical cause at the late election. Lancashire has long since been J59 113 secured through their means, and the West Riding all but converted J59 114 into the pocket borough of the Cobden family. In East Surrey and J59 115 Middlesex the Conservative candidates were defeated solely by their J59 116 means, and in many other counties there would not have been a J59 117 contest, much less a hard battle, had it not been for the exertions J59 118 of radical Freehold Land Societies ... Why should not the J59 119 'CONSTITUTIONAL PARTY' have their LAND SOCIETIES, and raise up a J59 120 counteracting body of intelligent freeholders to poll against those J59 121 of their destructive rivals? This course ought to have been pursued J59 122 years back.

J59 123 The most realistic attempt to counteract the impact of the J59 124 freehold land movement was made in the form of proposals for J59 125 legislative intervention to curtail the scope of its political J59 126 activity. Disraeli discussed such a project, "which will J59 127 probably be the key to future power", as early as 1850. It J59 128 was the end of the decade, however, before he presented a package J59 129 of reforms to Parliament which would, if passed, have nullified the J59 130 movement's additions to the electorate. The 1859 Reform Bill was J59 131 the most eloquent testimony to the impact of the freehold land J59 132 movement. It is now largely remembered as the occasion of the J59 133 collapse of Derby's second ministry, and for what Bright J59 134 contemptuously described as "fancy franchises" - J59 135 electoral qualifications based upon educational criteria, and on J59 136 savings bank investments. However, the Bill also proposed the J59 137 abolition of the forty-shilling freehold county vote, transferring J59 138 existing voters to the boroughs in which their property was J59 139 situated, but disenfranchising those whose qualification lay in a J59 140 county constituency alone. The ministry's defeat came with the J59 141 division on Russell's amendment that it was "neither just J59 142 nor politic to interfere" in the forty-shilling freehold J59 143 franchise. The contemporary Tory response to the freehold land J59 144 movement suggests that Cobden's enthusiasm for it may not have been J59 145 as tactically inept as his recent biographers, and before them J59 146 Roebuck and Bright implied.

J59 147 None the less, it failed "to effect a breach in that J59 148 fortress of landlordism, the county representation", as the J59 149 1852 Reformers' Almanac had predicted. The political J59 150 appeal of the movement depended heavily on its being seen as a J59 151 cognate organization within a concerted campaign for reform. The J59 152 steady decline of 'Liberal' interests in the counties (from a peak J59 153 of 71 in 1835 to 36 in 1847) was hardly arrested by the movement at J59 154 the general election of 1852, which saw a total of only thirty J59 155 Radical Liberals returned for the English counties. Parliamentary J59 156 patrons reassessed their support in the light of both the election J59 157 and the difficulties experienced by the Parliamentary and Financial J59 158 Reform Association that year. Bright was the first to decline J59 159 re-election to the board of the National Freehold Land Society (by J59 160 1858 only Gilpin remained). A careful estimate of the allotments J59 161 created by freehold land societies, made by Thomas Beggs the J59 162 following year, put the total at 19,500 - an impressive figure but, J59 163 diluted over forty-four counties, unlikely to force a rout at the J59 164 polls. Seven years later, in a speech opposing Disraeli's Reform J59 165 Bill, Bright estimated the total votes created at 36,000. The only J59 166 specific victories ascribed to the movement by its supporters were J59 167 East Surrey (where the National Society consolidated Locke King's J59 168 thin majority), Middlesex ("though by the hair of his [the J59 169 Radical member's] head"), and North Warwickshire.

J59 170 If election results disappointed parliamentary Liberals, they J59 171 only reflected the experience of the Freehold Land Union and the J59 172 Freeholder in trying to marshal the constituent parts of the J59 173 movement. The Union had no coercive powers, and it was a frequent J59 174 complaint that societies' officials failed to take it seriously, or J59 175 provide statistical data to enable the movement's progress to be J59 176 monitored. Cassell, editor of the Freeholder, another casualty J59 177 of 1852, voiced similar complaints. Middle-class liberalism had J59 178 failed to hijack what was essentially a working-class movement. As J59 179 such its aims and aspirations differed form those of its patrons, J59 180 as the authors of Progress of the Working Classes J59 181 observed:

J59 182 Freehold Land Societies started with the intention of J59 183 manufacturing forty-shilling freeholders. ... That objective is now J59 184 quite a matter of secondary consideration. The shareholders may not J59 185 care less for the freehold, or the political power it confers; but J59 186 their leading desire is to add a house to the land, and thus secure J59 187 free homes.

J59 188 From the mid-1850s the direction of freehold land societies' J59 189 literature increasingly emphasized the building function. This was J59 190 not an abrogation of their pedigree, since mutual improvement had J59 191 always shared with radical politics the shaping of the movement. J59 192 Even the most commercial societies continued to advertise an J59 193 intention to help "prevent ... the evil operation of those J59 194 laws that fostered the possession of large estates, and gave to the J59 195 few an undue influence over the many". If this was bravado, J59 196 it was none the less bravado that seems to have sold shares. J59 197 Likewise, the Freehold Land Times, which commenced in J59 198 1854 and was, in contrast to the earlier Freeholder, a J59 199 strictly commercial venture, maintained a policy of editorial and J59 200 news coverage of a Liberal bent, in support for example of land J59 201 reform and free trade, and attacking the conduct of the Crimean J59 202 War.

J59 203 What militated most against the ambitious agenda parliamentary J59 204 Liberals tried to impose on the movement, was the range and J59 205 character of the societies it awkwardly embraced. Not only were J59 206 certain societies uninterested in Liberal parliamentary J59 207 aspirations; some were actively hostile. There was little in common J59 208 between the National and even the larger provincial societies, and J59 209 still less between the latter and the small groups, based on local J59 210 and workplace communities, which were the typical form the movement J59 211 took, for example, in Sheffield. These contrasts themselves J59 212 reflected the often sharply-differentiated experience of labour in J59 213 the major industrial cities. To a significant extent Chartism had J59 214 exposed such differences rather than reconciled them: it was hardly J59 215 likely that a modest venture like the Freehold Land Union would J59 216 succeed.

J59 217 J60 1 <#FLOB:J60\>Texts which challenge these assumptions commonly find J60 2 themselves in the no-person's land between writings for adults J60 3 (so-called) and writings for children (so-called).

J60 4 In peer-texts, the adult reader (real or otherwise) can adjust J60 5 to the degree of control which the author appears to be exercising. J60 6 As an adult reader, my selection of a text may be governed, in J60 7 part, by the amount of effort I wish to bring to it and by a J60 8 judgement of how much effort is warranted. With books 'for J60 9 children', or 'unskilled' readers, because of the status of the J60 10 audience, the author-reader (or narrator-narratee) relationship is J60 11 a more than usually unbalanced power relationship. The audience is J60 12 created by the writer much more directly than with a peer-text, in J60 13 the sense that the text does more than display its codes, grammar, J60 14 and contracts; it suggests what the reader must be or become to J60 15 optimize the reading of the text. Drawing on the power-codes of J60 16 adult-child, book-child, and written-oral relationships, it J60 17 prescribes what the reader must be, and indeed, because J60 18 there is both an authoritarian and an educational element involved, J60 19 what the reader can be. The exercise of such power is by no J60 20 means inevitable, although it is so characteristic as to define the J60 21 children's book for many readers. Very often there seems to be a J60 22 deliberate attempt to limit the child-reader's interaction with the J60 23 text. This may seem to be benevolent, if one believes that the J60 24 'open' text is fundamental to literary development or, as J60 25 Jacqueline Rose suggests merely a fact of life for the 'impossible' J60 26 category of children's fiction.

J60 27 What Texts Imply

J60 28 Criticism, especially of children's literature perhaps, is J60 29 controlled by perception of genre; is children's literature J60 30 identifiable by lexical items, grammatical structures, higher level J60 31 narrative units, or an overall tonal strategy? For example, what J60 32 gives away the 'implied audience' for this quotation?

J60 33 "He woke up with a jerk, shivering with cold. He began J60 34 to stretch his cramped legs but they hurt. Opening his eyes, he J60 35 looked around in the darkness. He knew immediately where he was. He J60 36 had been locked under the stairs. He peered through the crack at J60 37 the side of the small door. It was pitch black."

J60 38 It could be that the verb "woke up", rather J60 39 than 'woke' or 'was awake' and the economical syntax (and lack of J60 40 punctuation) of the second sentence are intended to link the J60 41 discourse to the mind of the character. But unfortunately, the J60 42 stylistic simplicity of the passage - that is, its lack of J60 43 deviation or variety - merely points up the logical and referential J60 44 anomalies. (How could he "peer through" a J60 45 "crack" (or is it really a 'gap'?) which he could not see J60 46 (as it was "pitch black")? Indeed, how could he J60 47 know that he was under the stairs if it was so dark; and if he knew J60 48 by some means other than sight, why are we not informed about it?) J60 49 The summarizing mode is so pervasive that it constantly shifts J60 50 towards implicit authorial control, which in turn becomes a marker J60 51 (or an assumed marker) of the genre of children's literature. And J60 52 this is quite apart from the grammatical features; five of the J60 53 seven sentences have the same structure (six if we discount the J60 54 clause "Opening his eyes"). Yet Michelle Magorian's J60 55 Goodnight Mr Tom not only won the British Library J60 56 Association's Carnegie Medal, but also (ironically enough), the J60 57 International Reading Association's Children's Book Award (1982). J60 58 Since this extract is characteristic of the novel, we may have here J60 59 some indication of the relative stress laid by judges upon content J60 60 and style.

J60 61 Magorian's text tells rather than shows, explicates rather than J60 62 demonstrates; and books which retain this dominating narrational J60 63 presence, the residual or 'transferred' storyteller, are a textual J60 64 echo of storying as an event which the storyteller essentially J60 65 controls. In general, it seems that this control is only J60 66 reluctantly relinquished (which may say something about the J60 67 adult-child relationship), which can scarcely be for so simple a J60 68 reason as that the reading audience cannot understand the text J60 69 without a built-in prompter. In fact, even skilled readers have J60 70 difficulty with the voice of the storyteller addressing the J60 71 audience 'directly' in 'two-dimensional' printed texts. As J60 72 the history of the early novel demonstrates, the act of storying J60 73 involves a narrative voice or stance, an implied narrator or author J60 74 or quasi-storyteller (or a device to replace it); and this produces J60 75 a grammatical and psychological situation of immense complexity.

J60 76 When, as in texts designed to be read to children (or, J60 77 indeed, any audience) there is a first-person marker, there can be J60 78 problems, as we have seen from The Tale of Tom Kitten. J60 79 One of the most complex instances is the opening of Milne's J60 80 Winnie-the-Pooh. The narrative begins with a direct J60 81 address to the implied reader, marked by second-person form: J60 82 "Anyhow, here he is ... ready to be introduced to J60 83 you." It then moves to a situation where the first-person J60 84 narrator describes how she or he tells a story to Christopher J60 85 Robin, who now becomes both a character and an addressee: J60 86 "You aimed very carefully at the balloon, and fired. 'Did I J60 87 miss?' you asked."

J60 88 The problems confronting a reader of that text and a listener J60 89 to that reader are formidable, not least because the reader implied J60 90 (and thus required) by the text is not the actual receiver. Hence J60 91 the linguistic needs are different. There is an entertaining J60 92 paradox here. The storyteller's summaries, intended to make things J60 93 easier for the listener, are quite likely when they appear in J60 94 a text being read silently to make things more difficult for the J60 95 reader. They have not sprung from a genuine need (on the part J60 96 of the reader), and as a result they require an artificial J60 97 convergence of text-codes and reader-codes, rather than, as in the J60 98 case of the 'given' text, allowing an exploration of codes which J60 99 may not cohere, and may not need to. (The J60 100 implications of this can be seen in Robert Leeson's account of the J60 101 history of children's literature, which emphasizes the interplay of J60 102 oral and written patterns in a socio-political context.)

J60 103 An example of both the summary and the quasi-storyteller's J60 104 voice can be seen in Ruth Park's novel Playing Beattie J60 105 Bow (which won the Australian Children's Book of the Year J60 106 Award in 1981):

J60 107 As she stood there, looking up at the askew, rusted J60 108 pulley, and the edge of the roof above it, a small patch of the sky J60 109 suddenly lost its stars.

J60 110 Someone was lying on the warehouse roof looking down at her.

J60 111 Chapter 7

J60 112 When Abigail realised that she was being spied upon J60 113 ...

J60 114 Here we have three renditions, or variations, of the same J60 115 essential semantic set, which progressively 'close' the text. J60 116 "A small patch of the sky suddenly lost its stars" J60 117 requires a considerable interpretive effort by the reader, and it J60 118 carries several possibilities. "Someone was lying on the J60 119 warehouse roof" restricts these possibilities. J60 120 "Looking down on her" and "realised that J60 121 she was being spied upon" similarly move from 'showing' to J60 122 'telling', from 'open' to 'closed'. Of course, it could be argued J60 123 that this progression reflects the deductions made by Abigail, so J60 124 that Park holds to the contract of narration through a single J60 125 consciousness. However, the progression from stylistic deviation J60 126 (the adverb in an adjectival position in "askew, rusted J60 127 pulley") to clich<*_>e-acute<*/> ("being spied J60 128 upon") re-assumes control. This is further corroborated by J60 129 the explanatory work of the first sentence in the new chapter, and J60 130 of course we need not assume that the presence of a chapter J60 131 division requires a break in the flow of reading.

J60 132 The Reader and Meaning

J60 133 Children are developing readers; their approach to life J60 134 and text stems from a different set of cultural standards from J60 135 those of adult readers, one that may be in opposition, or perhaps J60 136 based on orality. Hence they do 'possess' texts, in the sense that J60 137 their meanings are their own and private, even more than adults. J60 138 Adult readers know the rules of the game, even if they don't know J60 139 that they know; and their understanding, as we have seen, may rest J60 140 on belonging to 'interpretive communities' which not only know the J60 141 rules of the game but share their knowledge and attitudes. I would J60 142 like to lay bare some of these rules, and suggest that J60 143 child-readers cannot possibly have access to all of them. So, J60 144 regardless of what the text prompts, they are not necessarily in a J60 145 position to make use of it.

J60 146 But surely we can have some idea of what children J60 147 understand, otherwise the whole edifice of communication, J60 148 publishing, and language teaching for children comes tumbling down. J60 149 For example, what about comprehension texts, still so much alive in J60 150 public examinations?

J60 151 It seems obvious that all we are doing if we ask questions J60 152 about the 'content' or 'meaning' of a text is testing a child's J60 153 social competence (which is, perhaps, all we should do, or should J60 154 hope to do). That is, children who are successful in comprehension J60 155 tests demonstrate no more than that they can find the answer J60 156 implicit in the question. The 'real' meaning of the text to the J60 157 individual remains hidden; children (perhaps for ever afterwards) J60 158 develop the skill to say what they are supposed to say, and may J60 159 well assume that their private understandings are in some way J60 160 'wrong' - just as those who set the examination questions must J60 161 assume that their own reading of the text is in some way J60 162 'right'.

J60 163 In his excellent book Developing Response to Fiction, J60 164 Robert Protherough suggests that there is a spectrum between what J60 165 is 'objectively' correct - that is, something which all speakers of J60 166 the language will agree on as being 'there' in the text - and J60 167 things which are subjective and purely personal. His spectrum J60 168 (which could, I think, bear some modification) runs, in outline, J60 169 thus:

J60 170 1 Matters of fact

J60 171 2 Clear implications

J60 172 3 Manifest literary effects [e.g. symbols, motifs, shifts of J60 173 viewpoint]

J60 174 4 Shared associations

J60 175 5 Significance to the reader based on 'a particular stance' J60 176 [that is, a doctrine or ideology]

J60 177 6 Private associations

J60 178 Some of these - perhaps the first four - might seem to be the J60 179 common property of all readers. We read within a reading community, J60 180 and therefore can share meanings and understanding. But is this J60 181 really so?

J60 182 To look at it another way, are there degrees of understanding, J60 183 which, when we are writing or prescribing fiction, we will accept? J60 184 Is there another spectrum between total understanding of what the J60 185 writer intended and a free-form, totally personal reading, which, J60 186 say, takes After the First Death as a drawing-room comedy J60 187 or The Lion, The Witch, and the Wardrobe as a pagan text? J60 188 (The latter example is not as grotesque as it may appear. The book J60 189 has been banned in certain areas of the USA on just those grounds.) J60 190 Is there such a thing as 'total' comprehension? And are there J60 191 degrees of comprehension which we might accept as adequate, or J60 192 normal, or worth giving a good mark to?

J60 193 It is obvious that there are limits to the shared making of J60 194 meaning. What the author meant is, strictly speaking, unknowable, J60 195 even to the author. But we have to assume a certain congruence J60 196 between what you see and what I see and what a child-reader sees; J60 197 otherwise the whole business of making books (and, especially, J60 198 talking about them) becomes a nonsense. There must be a middle J60 199 ground of common-sense agreement about what meaning is.

J60 200 This may not seem to have got us very far; but it may at least J60 201 have made us cautious about assuming any similarity of J60 202 understanding between readers. What we need to do now is to J60 203 investigate the way in which texts work - what the shared rules are J60 204 - so that we can see where individual readers are likely to go J60 205 their own ways.

J60 206 Equally, the way texts are organized and our understanding of J60 207 that organization have a profound effect on the way we see the J60 208 world. As Roger Fowler puts it: "Linguistic codes do not J60 209 reflect reality neutrally; they interpret, organize, and classify J60 210 the subjects of discourse. They embody theories of how the world is J60 211 arranged: world-views or ideologies. For the individual, these J60 212 theories are useful and reassuring, making his relationship with J60 213 the world simple and manageable." J60 214 J61 1 <#FLOB:J61\>Chapter 8

J61 2 Your word is my command: the structures of language and power J61 3 in women's science fiction

J61 4 Lucie Armitt

J61 5 One of the frustrations felt by some readers and critics of J61 6 science fiction stems from the fact that, despite the potential for J61 7 innovation inherent in the apparently limitless narrative J61 8 possibilities of the genre, in actuality such innovation is rarely J61 9 fully realised within either content or form.

J61 10 My first serious encounter with science fiction was through an J61 11 interest in the political implications of language and its J61 12 structures for women. Language is of paramount importance with J61 13 regard to how we structure reality (providing a cognitive framework J61 14 for compartmentalising objects and sensations into linguistic units J61 15 of meaning). Indeed it has been argued that "reality J61 16 construction is probably to be regarded as the primary function of J61 17 human language", a claim which emphasises the need for J61 18 women to challenge the patriarchal bases of language if we are also J61 19 to challenge the patriarchal bases of society. However, trapped as J61 20 we are within a patriarchal linguistic and social framework, it is J61 21 very difficult for any writer to distance herself from that J61 22 framework and write through and about alternative structures whilst J61 23 still aiming to depict reality as it is lived and experienced. As J61 24 this essay sets out to demonstrate, it is not enough merely to J61 25 challenge surface manifestations (with revisions of words such as J61 26 'chairman', 'mastery', 'authoress' and so on, important though such J61 27 revisions are), but we must also analyse and subvert the deep J61 28 structural principles of language. Because of its ability to J61 29 provide the writer with this much-needed distancing from lived J61 30 reality, science fiction is an obvious choice for the writer intent J61 31 on such exploration. The two novels upon which this essay focuses J61 32 (Doris Lessing's The Marriages Between Zones Three, Four and J61 33 Five, and Suzette Elgin's Native Tongue) succeed in J61 34 illustrating two important issues: first, that the power structures J61 35 upon which societies depend are structured by and through the J61 36 structures of language; and, second, that by providing a fictional J61 37 context for what has otherwise been largely theoretical J61 38 abstraction, contemporary science fiction by women has made a by no J61 39 means insignificant contribution to the current theoretical debate J61 40 on women's relationship to language - and thus power.

J61 41 An essential starting point for any critique of language as a J61 42 power base of or for society is to demystify this ability J61 43 to shape and control our thoughts, perceptions and behaviour. In J61 44 The Marriages Between Zones Three, Four and Five Doris J61 45 Lessing begins with a vision of language as pre-eminently J61 46 deterministic of behaviour patterns in both the individual J61 47 characters and the societies depicted. To language is attributed an J61 48 autonomy outside of human communication and almost pre-existing of J61 49 any social referent whatsoever. She personifies language as a J61 50 pseudo-deity whose origins are unquestioned, and whose demands J61 51 remain unchallenged: "there was an Order, and it said, J61 52 simply, that she must go", (p. 13). Throughout the novel J61 53 the main protagonists' behaviour is directed and determined by this J61 54 "Order", a term for which we are given no further J61 55 definition, and which incorporates for the reader the dual J61 56 implications not only of the linguistic imperative but also of a J61 57 hierarchical power structure (both implications being significant J61 58 to its understanding). So the main characters of the novel are J61 59 impelled to act by unspecified drives and urges which have no J61 60 contextual reference and which are never fully explained, but which J61 61 exist and are articulated and understood solely at the level of J61 62 language. The narrative opens:

J61 63 Rumours are the begetters of gossip. Even more are they J61 64 the begetters of song. We, the Chroniclers and song-makers of our J61 65 Zone, aver that before the partners in this exemplary marriage were J61 66 awake to what the new directives meant for both of them, the songs J61 67 were with us, and were being amplified and developed from one end J61 68 of Zone Three to the other. (p. 11)

J61 69 It is perhaps fitting that rumour should be the catalyst which J61 70 initiates the narrative, it being the one form of utterance which J61 71 typically lacks a definite source or origin for its existence. In J61 72 addition, it is all the more fitting that the event concerned is a J61 73 marriage, because mainstream linguistic theory has frequently used J61 74 the social contract of the marriage vow as the primary illustration J61 75 of the 'illocutionary speech act', a term which is of value to this J61 76 argument because it hints at the power which even conventional J61 77 linguistic theory attributes to language as the initiator, as well J61 78 as the descriptor, of social codes. The term 'speech act' refers to J61 79 an utterance which not only articulates, but actually brings into J61 80 being a particular state of affairs. Thus, for example, in speaking J61 81 the words which form the marriage vows, the marriage itself is J61 82 brought into existence. It is created by and through J61 83 language. As is clear from the quotation cited above, once the J61 84 rumour is initiated in The Marriages, the marriage is J61 85 seen to exist, and neither Al-Ith nor Ben Ata has the power to J61 86 change events, which are from then on pre-ordained.

J61 87 I would not wish to suggest, however, that Lessing's vision of J61 88 language is that of an entirely self-referential or J61 89 self-perpetuating system (an approach which would toe the line of J61 90 traditional linguistic theory). Many of the theories put forward J61 91 both by psychoanalytic and later structuralist theoreticians could J61 92 be usefully applied to this text, particularly where they have J61 93 relevance to women's marginal position to the Symbolic Order or to J61 94 the dominant discourse. At one point in the narrative Lessing J61 95 refers to the orders by which Zone Four functions as J61 96 "ukases" (p. 251), a term deriving from Tsarist J61 97 Russia, and which has come to take on the meaning of 'any arbitrary J61 98 order.' The implications of this word are doubled by the pun J61 99 inherent in this definition which refers to the arbitrary nature of J61 100 the power base upon which Zone Four (and by extension the J61 101 patriarchal establishment) is based, and which makes fully apparent J61 102 (i) the intrinsic nature of the relationship between language and J61 103 the structures of power; and (ii) the fact that such structures are J61 104 not pre-ordained at all, but entirely arbitrary.

J61 105 H<*_>e-acute<*/>l<*_>e-grave<*/>ne Cixous distinguishes between J61 106 the Realm of the Gift and the Realm of the Proper J61 107 (Propre), a distinction worth applying here as Zone Three J61 108 could be considered representative of the former, and Zone Four the J61 109 latter. From a linguistic perspective this is manifested through J61 110 Zone Four's barked commands and imperatives (one might well say J61 111 that 'sentences are passed' on the ordinary people of the zone by J61 112 those positioned higher up the military hierarchy), while Zone J61 113 Three's communication works via finely tuned spatial techniques J61 114 which take us beyond the limits of word and sentence. The confusion J61 115 and dislocation felt by Jarnti and Ben Ata at encountering Zone J61 116 Three (either directly or via Al-Ith) is due to the fact that - J61 117 their commands being misplaced and indeed irrelevant in that region J61 118 - they lack any ability to shape or construct reality through J61 119 language. The linguistic units which frame the commands J61 120 simultaneously pre-package their perceptions of reality. Under the J61 121 rules of the Realm of the Proper, language J61 122 ap-propr-iates, leaving no room for the creation or J61 123 generation of the alternative visions associated with the Realm of J61 124 the Gift.

J61 125 Science fiction often draws very closely on narrative devices J61 126 and structures common to other forms of popular fiction, and in J61 127 some ways the narrative concerns of The Marriages works J61 128 through similar methods to the folktale or quest allegory. The J61 129 story of a queen of one land who marries the king of another in J61 130 order to restore harmony to her realm is a conventional one within J61 131 such literature, and in line with the allegorical form both Al-Ith J61 132 and Ben Ata are personifications of the abiding characteristics of J61 133 their own respective zones. In this context Waelti-Walters' J61 134 comments on the fairytale form are significant:

J61 135 Woman is gift and possession ... the touchstone of the J61 136 whole system of values men in our civilisation have built up J61 137 between themselves and which has taken symbolic form in J61 138 language.

J61 139 Initially, as implied above, Al-Ith embodies the Realm of the J61 140 Gift and Ben Ata the Realm of the Proper, the two forming an J61 141 irreconcilable dichotomy. Through their enforced marriage, however, J61 142 these contrasting features gradually become blurred, and Al-Ith in J61 143 particular forms the site of the conflagration of the two, rendered J61 144 visible through her pregnancy with Arusi. Thus she takes on board J61 145 this role of the touchstone of the value system, her victimisation J61 146 at the hands of a hostile system becoming more and more apparent as J61 147 the narrative progresses. Waelti-Walters' comment refers to the J61 148 linguistic implications of this (the use of the term 'symbolic' J61 149 here relating to the importance of language in the forging of J61 150 social codes, rather than the more specific psychoanalytic J61 151 definition of the term). The basic structuralist argument, deriving J61 152 from the work of Ferdinand de Saussure, argues that language J61 153 operates via a system of signs (the sign being the twofold J61 154 combination of the sound/word and the mental concept) which take on J61 155 meaning in relationship to each other and the system as a whole. J61 156 Later advocates of structuralism came to believe that this J61 157 linguistic model could usefully be applied to other social codes J61 158 and to the workings of literary narratives, one of the most J61 159 important of these being Propp's Morphology of the Folk J61 160 Tale. Lessing's narrative works quite closely along J61 161 folk-legend parallels, about which Propp notes: J61 162 "Functions constitute the basic elements of the tale, those J61 163 elements upon which the course of the narrative is built" - J61 164 the 'function' in this sense being the way in which any particular J61 165 action or event takes on significance (or indeed signification) J61 166 within the narrative as a whole. In the traditional folktale, J61 167 'functions' take on significance primarily as they relate to, or J61 168 derive from, the underlying moral code (which structuralist critics J61 169 would also interpret as a linguistic code), this being the J61 170 real (if frequently disguised) power behind its telling. As J61 171 various critics, including Waelti-Walters, have shown, this moral J61 172 code is not an apolitical guideline for children, but a product of J61 173 the prevailing ideology out of which the tale arises, and which J61 174 seeks primarily to protect the status quo. This is where J61 175 The Marriages departs from the afore-mentioned model. J61 176 Lessing not only foregrounds ideology within her narrative J61 177 framework, but demonstrates the significance of language to it. The J61 178 Chronicler is the persona whose main function is to stress the J61 179 folk-legend quality of the tale. He it is who interrupts the J61 180 narrative, just as the reader forgets that this is a story told by J61 181 a storyteller, to remind us that chroniclers in the two zones J61 182 frequently take one scene and depict it in very different ways, J61 183 turning it in the process into history or myth (fact or fiction) as J61 184 best suits their purposes. So Lessing's Chronicler reminds us that J61 185 language plays a central role in the construction of so-called J61 186 'truth' (to 'chronicle' an event being to present it as 'fact'); J61 187 thus truth is never absolute, but at all times remains a subjective J61 188 - and in this context an ideological - construct, something that J61 189 can work for good or ill. As the Chronicler states: J61 190 "Describing we become. We even - and I've seen it and have J61 191 shuddered - summon" (p. 243).

J61 192 But language is not just depicted as a key to power in this J61 193 novel, it also becomes one of the fundamental keys to knowledge. It J61 194 is through writing his version of the story of Al-Ith that the J61 195 Chronicler learns: "I write in these bald words the deepest J61 196 lessons of my life" (p. 242), lessons which Al-Ith will J61 197 come to share. Referring to the intrinsic relationship between J61 198 language and our perceptions of reality, Grace comments that, J61 199 through language, "we do not just see shifting patterns of J61 200 light, we see people and objects and actions". When Al-Ith J61 201 attempts to initiate herself into Zone Two without being granted J61 202 access to it by 'the Order', she stands just over the threshold of J61 203 the zone and considers:

J61 204 that all around her, above her, were people - no, J61 205 beings, were something, then, or somebody, invisible but there.... J61 206 Almost she could see them ... as if flames trembled into being.... J61 207 J62 1 <#FLOB:J62\>KNOW THYSELF vs COMMON KNOWLEDGE:

J62 2 BLEICH'S EPISTEMOLOGY SEEN THROUGH TWO SHORT STORIES BY J62 3 BALZAC

J62 4 Reader-response theorists, still haunted by the spectre of the J62 5 'affective fallacy', yet equally aware of the dangers of an J62 6 objectivist stance, are faced with a problem of authority: who or J62 7 what is the ultimate source of meaning? At one end of the spectrum, J62 8 authority may be invested in the actual author of the text, as in J62 9 the theory of E. D. Hirsch; at the other, meaning may be a function J62 10 of the individual reader's identity, an approach favoured by Norman J62 11 Holland. Other works, such as Wolfgang Iser's The Act of J62 12 Reading, or Stanley Fish's Affective Stylistics, J62 13 grant the written text a measure of authority by claiming that the J62 14 reader's response is guided and limited by specific objective J62 15 textual structures.

J62 16 Epistemologies may shift authority from the subjective to the J62 17 intersubjective: meaning is a function not of the individual reader J62 18 but of intersubjective communities. The later work of Stanley Fish, J62 19 introducing the notion of 'interpretive communities', falls into J62 20 this category. David Bleich's Subjective Criticism, J62 21 whilst appearing to favour the subjective end of the spectrum, also J62 22 has something of the 'community spirit'. As will emerge, the J62 23 relationship posited between individual subject and community is a J62 24 problematic one.

J62 25 Reading the theory 'through' Balzac's Le Colonel J62 26 Chabert and Adieu, I shall explore a number of problems J62 27 surrounding Bleich's epistemology, and consider some of the wider J62 28 implications arising from the concept of community.

J62 29 I. Bleich's Epistemology

J62 30 According to Bleich, paradigms or world-views are set up to J62 31 meet the epistemological needs of the present. The objective J62 32 paradigm which he challenges is no exception: "The notion J62 33 of objective truth has the epistemological status of God: it is an J62 34 invented frame of reference aimed at maintaining prevailing social J62 35 practices." This paradigm, claims Bleich, has outlived its J62 36 usefulness. Its epistemological presuppositions must be revised in J62 37 the light of work carried out by figures such as Einstein, J62 38 Heisenberg, and Gombrich. 'Facts' or 'knowledge' do not exist J62 39 independent of the investigator: they are a function of his motives J62 40 and the means of investigation. 'Facts' are not found, but made.

J62 41 Bleich's 'subjective paradigm' is intended to supersede the J62 42 objective approach. How does this new epistemology operate in the J62 43 field of reader-response? According to Bleich, though the literary J62 44 text has physical properties, its meaning depends entirely on a J62 45 process which he calls 'symbolization'. Symbolization, the reader's J62 46 initial response, can be defined as a peremptory perception and J62 47 evaluation of the text. It is an imaginative interiorization which J62 48 is differentiated from a purely sensorimotor response.

J62 49 Interpretation of the text is referred to as motivated J62 50 'resymbolization', where resymbolization is defined as the J62 51 mentation involved in a conscious response to the symbolization. It J62 52 is a reframing of the symbolization which occurs when the present J62 53 adaptive needs of the individual demand an act of explanation or J62 54 interpretation. The interpretation should not be considered in J62 55 terms of a true/false polarity, but rather in terms of validity or J62 56 conviction. It is not concerned with a recovery of 'fact' or J62 57 'truth': its success depends upon its meeting of present demands or J62 58 adaptive needs: "The logic of interpretation is that its J62 59 resymbolising activity is motivated and organised by the conscious J62 60 desires created by disharmonious feelings or self-images; the goal J62 61 of these desires is increasing the individual's sense of J62 62 psychological and social adaptability" (Subjective J62 63 Criticism, p. 83).

J62 64 Central to Bleich's epistemology is the individual's capacity J62 65 for reflective thought. Bleich refers to a 'subjective dialectic', J62 66 the individual's ability (for example, when hearing a story) to ask J62 67 himself questions and receive answers. This activity is a sign that J62 68 symbolization (imaginative response), and not just a sensorimotor J62 69 activity (registering the visual stimuli of print), has taken J62 70 place. In the context of the reading act, Bleich introduces a J62 71 valuable hermeneutic tool: the 'response statement'. This is a J62 72 recording by the reader of his subjective responses as he reads. J62 73 The 'response statement' is used alongside the subject's 'meaning J62 74 statement' or interpretation of a text, with a view to bringing to J62 75 reflective attention the subjective stratum of response, and J62 76 thereby acknowledging its influence on the apparently objective J62 77 'meaning statement'. The reading subject is, we might say, J62 78 answerable to and for himself; he is responsible for his reading. J62 79 The important thing is to know that you know.

J62 80 Knowledge, however, does not stop with the individual subject J62 81 and his reflective powers: it must be 'negotiated' in a community: J62 82 "To know anything at all is to have assigned a part of J62 83 one's self to a group of others who claim to know the same thing J62 84 .... The degree to which knowledge is not part of a community is J62 85 the degree to which it is not knowledge at all" J62 86 (Subjective Criticism, p. 296). The community is also J62 87 deemed to have reflective powers, to be conscious of, and able to J62 88 articulate its own motivation for knowledge-seeking. It, too, is J62 89 answerable to and for itself; it, too, is responsible. Bleich J62 90 considers this new epistemology to be ethically superior to its J62 91 objective counterpart: operating according to the rules of the J62 92 latter, the subjective stratum of any knowledge-seeking act goes J62 93 unnoted and probably unnoticed. Interpretation operating under the J62 94 objective paradigm has no authority: authority requires a J62 95 hermeneutic moment which reveals the interests or motivations of J62 96 the seeker of knowledge or explanation.

J62 97 II. Balzac and Bleich

J62 98 Both Le Colonel Chabert and Adieu represent J62 99 characters prompted by their present adaptive needs to engage in J62 100 knowledge-seeking acts, and to form some kind of intersubjective or J62 101 community relationship. Both texts, to a certain extent, play out J62 102 or dramatize some of Bleich's ideas. Tompkins has noted that J62 103 Bleich's privileging of self-knowledge, manifested in the J62 104 'subjective dialectic' and 'response statement', marks an J62 105 inconsistency in his epistemology. I shall show that Le J62 106 Colonel Chabert dramatizes the very epistemological issues at J62 107 stake. Both short stories, in fact, not only 'stage' some of the J62 108 theoretical concepts; they also serve to challenge the theory and J62 109 to reveal some of the implications which arise from the J62 110 relationship between individual subject and intersubjective J62 111 community.

J62 112 Le Colonel Chabert tells the tale of a Napoleonic J62 113 colonel injured and left for dead after the battle of Eylau. J62 114 Suffering from a severe head-wound, the colonel digs his way out of J62 115 a mass grave, recovers his health and memory, and returns to what J62 116 is, by then, Restoration Paris. There he enlists the help of the J62 117 solicitor Derville, in an attempt to regain public recognition. J62 118 After a crucial meeting with Chabert, his wife, since remarried, J62 119 realizes that his physical appearance has so altered as to render J62 120 him unrecognizable to the public. After a series of manoeuvrings, J62 121 she successfully thwarts Chabert's attempts, and the colonel fades J62 122 into obscurity.

J62 123 Le Colonel Chabert prompts several questions about J62 124 the nature of self-knowledge. For example, can the individual's J62 125 grasp of self-identity be challenged by an external authority? J62 126 Early in the short story, one of Derville's clerks vouches for the J62 127 authority of the subject: "Je l'appelle pour lui J62 128 demander s'il est colonel ou portier, il doit savoir, J62 129 lui" (p. 31, my italics). As far as Chabert J62 130 is concerned, his self-knowledge, his knowledge of who he is, is, J62 131 quite literally, self-evident. There can be no question of J62 132 intersubjective negotiation:

J62 133 -Il faudrait peut-<*_>e-circ<*/>tre J62 134 transiger, dit l'avou<*_>e-acute<*/>.

J62 135 -Transiger, r<*_>e-acute<*/>p<*_>e-acute<*/>ta le colonel J62 136 Chabert. Suis-je mort ou suis-je vivant? (p. J62 137 58)

J62 138 To this extent, Bleich's privileging of knowledge of self seems J62 139 to find literary support. However, Chabert's self-knowledge is J62 140 challenged and indeed denied by the 'comtesse', a J62 141 representative of the community which is Restoration society. As J62 142 Chabert discovers, even self-knowledge is negotiable: J62 143 "tout se plaide" (p. 72). J62 144 Derville, using terms which would not be out of place in Bleich's J62 145 own work, makes the point that self-knowledge becomes knowledge J62 146 only if and when negotiated in a motivated community: J62 147 "Vous <*_>e-circ<*/>tes le comte Chabert, je le J62 148 veux bien, mais il s'agit de le prouver juridiciairement, J62 149 <*_>a-grave<*/> des gens qui vont avoir J62 150 int<*_>e-acute<*/>r<*_>e-circ<*/>t <*_>a-grave<*/> nier votre J62 151 existence" (p. 70, my italics). This raises a J62 152 number of points which are inadequately covered by Bleich. Exactly J62 153 how does his epistemology reconcile self and community? On the one J62 154 hand, it seems impossible to deny the privileged nature of J62 155 self-knowledge, specifically, of the reflective ego: to do so would J62 156 be to deny the basic ontological structure of the self. The J62 157 'comtesse' sees only a part of the problem when she tells J62 158 Chabert that "renoncer <*_>a-grave<*/> J62 159 vous-m<*_>e-circ<*/>me" would mean J62 160 "commettre un mensonge <*_>a-grave<*/> toute J62 161 heure du jour" (p. 107). To deny the reflective J62 162 ego is, ultimately, to deny life itself. Chabert himself touches on J62 163 the problem when he expresses a desire for the ontologically J62 164 impossible: "Je voudrais n'<*_>e-circ<*/>tre pas J62 165 moi." In the closing pages of the text, his J62 166 attempted self-renunciation hints at the fact that to abolish the J62 167 reflective ego is to cease to be a human subject: J62 168 "Pas Chabert! pas Chabert! ... Je ne suis plus un J62 169 homme, je suis le num<*_>e-acute<*/>ro 164, septi<*_>e-grave<*/>me J62 170 salle" (p. 121).

J62 171 If self-knowledge is an ontological necessity, how can it be J62 172 integrated into an epistemology such as Bleich's, which insists on J62 173 the intersubjective negotiation of all knowledge? Self-knowledge is J62 174 by definition solipsistic. If "tout se J62 175 plaide", how can proof be offered to a third J62 176 party? How can Chabert's reflective awareness of himself become J62 177 material which is accessible to intersubjective negotiation?

J62 178 This problematic relationship between transcendent subject and J62 179 community is briefly raised by Tompkins, who compares Bleich's J62 180 concept of community to that of Stanley Fish:

J62 181 Both statements suggest that individual or subjective J62 182 knowledge exists prior to the formulation of publicly shared J62 183 assumptions and that membership in an interpretive community is a J62 184 conscious act entered into freely by each individual .... This J62 185 description [of the interpretive community] stands Fish's concept J62 186 on its head. Instead of the individual's being constituted by the J62 187 assumptions of the group, the group is formed by individuals who J62 188 then negotiate its assumptions into existence. (p. 1073)

J62 189 The conflict appears to be irreconcilable. One can adopt Fish's J62 190 stance, whereby the subject is constituted by group assumptions and J62 191 correlatively loses the freedom of the self-reflective, J62 192 transcendent self. In this case we can no longer talk of authority J62 193 or responsibility: the subject is not answerable to or for himself. J62 194 Alternatively, one can seek to preserve, as Bleich does, the J62 195 freedom of the transcendent self, in which case we cannot accept J62 196 that all knowledge is negotiated in a community.

J62 197 Le Colonel Chabert plays out this epistemological J62 198 conflict to perfection. In Chabert's case, joining the community J62 199 represented by the 'comtesse' would mean denying his own J62 200 existence; it results in an ontological impasse. Extending this, J62 201 and taking Chabert's case as exemplary, we might ask how can J62 202 individuals be convinced to join a particular community? Thomas J62 203 Kuhn, referring to paradigm switches and communities of scientists, J62 204 suggests that persuasion can go only so far: "The transfer J62 205 of allegiance from paradigm to paradigm is a conversion experience J62 206 that cannot be forced." Richard Norman, considering a J62 207 similar problem in terms of seeing 'gestalt' figures (duck/rabbit; J62 208 old/young woman), states: "I cannot 'choose' or 'decide' to J62 209 see the figure in a certain way. If I am able to see it as a J62 210 picture of an old woman, this is because that way of seeing forces J62 211 itself upon me. We can speak of the dawning of an aspect." J62 212 Proffered 'proof' about the 'gestalt' figure ('here is the nose, J62 213 here the chin', etc.) is proof only to one who is already J62 214 in-context: "The 'evidence' each gives will be understood J62 215 in the relevant way only by one who is already convinced" J62 216 (p. 335). Neither Fish, who takes the same line (the subject is J62 217 always-already<&|>sic! in context), nor Bleich considers in J62 218 sufficient detail the processes of joining and leaving J62 219 communities.

J62 220 The conflict between Chabert and the 'comtesse', a J62 221 spokeswoman for the Restoration community, extends the issue into J62 222 the domain of ethics: what happens when there is a conflict of J62 223 interests between communities? How are differences settled? How, J62 224 exactly, does communication and negotiation take place? In the case J62 225 of Chabert, the innocent victim seems to fall prey to an anonymous J62 226 corporate body.

J62 227 Bleich notes: "If there is no external standard, J62 228 collective interests are the highest authority, and knowledge J62 229 depends ultimately on how individuals form groups and circumscribe J62 230 the existence of other groups" (Subjective J62 231 Criticism, p. 264). J62 232 J63 1 <#FLOB:J63\>GENDER AND NARRATIVE IN THE FICTION OF APHRA J63 2 BEHN

J63 3 By Jacqueline Pearson

J63 4 Aphra Behn is still a neglected writer, and even among those J63 5 critics who have given her work more than a passing acknowledgement J63 6 there is a striking lack of consensus about the nature and extent J63 7 of her achievement, especially in her fiction. She has been praised J63 8 as innovative and original: Robert Adams Day thus finds J63 9 Oroonoko "entirely original" in its narrative J63 10 methods and praises its "astonishing innovations". J63 11 On the other hand, many accounts of the rise of the novel ignore or J63 12 understate her contribution: Ian Watt's classic study, The J63 13 Rise of the Novel, makes only two brief references, and even J63 14 in a full-length study of Behn's work, F. M. Link finds her fiction J63 15 unoriginal and concludes that she "made no significant J63 16 contribution to the development of the [novel] form". And J63 17 if there is little agreement on questions of originality and J63 18 influence, there is no more on the themes of the novels. J63 19 Oroonoko, for instance, has been seen as expressing J63 20 "republican prejudices", or as demonstrating a J63 21 strongly royalist viewpoint, or both. There is especially a lack of J63 22 consensus on Behn's treatment of gender: some critics find her a J63 23 vigorous feminist, making 'suffragette' claims for women, while J63 24 others argue that she compromised with a male-dominated literary J63 25 establishment and that her work consequently displays a J63 26 "masculine set of values".

J63 27 A high proportion of recent criticism of Behn's fiction adopts J63 28 a na<*_>i-trema<*/>vely literalist reading, taking Oroonoko J63 29 and even The Fair Jilt and other tales quite simply as J63 30 self-revelation, as direct autobiography. As a result of, and a J63 31 reaction against, this kind of reading, the most promising recent J63 32 reassessments have focused on the role of the narrators. The J63 33 narrator has been seen to provide circumstantial detail, local J63 34 colour, a vivid immediacy, and a "breezy colloquial J63 35 quality", to offer "a viable standard of judgement J63 36 for the readers", to unify the novel and involve the reader J63 37 emotionally in the narrative, and also "to attest the truth J63 38 of the whole story". I shall argue, however, that the J63 39 situation is still more complex, and that the narrator, who is not J63 40 coterminous with 'Aphra Behn', is a complex and subtle part of J63 41 Behn's treatment, both open and implied, of issues of gender and J63 42 power. In order to do this, I shall briefly examine the strategies J63 43 with the narrator in a range of Behn's fiction, before going on to J63 44 a fuller analysis of the role of the narrator in Oroonoko.

J63 45 I J63 46 In Behn's fourteen fictions, the narrator is never definitely J63 47 male: six give no clue to gender, though she sometimes seems to be J63 48 female by implication, and in eight, 'The Unfortunate Happy Lady', J63 49 Oroonoko, The History of the Nun, 'The Nun', 'The J63 50 Lucky Mistake', 'The Unfortunate Bride', 'The Wandering Beauty', J63 51 and 'The Unhappy Mistake', she is definitely female. In the J63 52 simplest cases the female sex of the narrator lends an authority to J63 53 her accounts of women's lives and natures, and reflects the J63 54 empowering of women, or the mockery of men, within the narratives. J63 55 In more complex tales, the female narrator is depicted, like the J63 56 female characters, as embedded within patriarchy and limited by it. J63 57 These women are torn by contradictions, powerful and governing J63 58 within their fictions, powerless outside them, and their narratives J63 59 are deeply coloured, even undermined, by these contradictions. What J63 60 they present as simple narratives, entertaining stories or moral J63 61 tales, turn out to encode quite different meanings, more sinister, J63 62 revealing, and subversive, over which the narrators have less J63 63 perfect control. Narrators are given to Freudian slips, unnoticed J63 64 and unacknowledged self-contradiction, uncomfortable ambivalences, J63 65 not fully articulated, about the tales they tell. It is these J63 66 complex, uncomfortable, flawed, or even duplicitous narrators who J63 67 are Behn's most effective tool in her analysis of patriarchy.

J63 68 The most obvious of the contradictions imported by the J63 69 narrators is their misogyny, paradoxically most apparent in the J63 70 case of specifically female narrators. Tales may condemn female J63 71 weakness, like 'The Nun', or celebrate female strength, like 'The J63 72 Unfortunate Happy Lady', yet in both the female narrator displays J63 73 an oddly masculine misogyny: "our Sex seldom wants matter J63 74 of Tattle", "how wretched are our Sex, in being the J63 75 unhappy Occasion of so many fatal Mischiefs", "'tis J63 76 the humour of our Sex, to deny most eagerly those Grants to Lovers, J63 77 for which most tenderly we sigh, so contradictory are we to our J63 78 selves". The authority of the female narrator is thus used, J63 79 paradoxically, to give an authoritative insider's view of female J63 80 weakness, as the female narrator seeks male approval by attacking J63 81 her own sex or by modest self-deprecation. Such ambiguities are J63 82 perhaps best understood as the result of the self-divisions - J63 83 "so contradictory are we to our selves" - J63 84 experienced by the female narrator, anxious to succeed within a J63 85 male-dominated literary establishment and consequently obliged to J63 86 accept its standards, and yet also, sometimes admittedly, sometimes J63 87 not, highly critical of that male world and its male J63 88 inhabitants.

J63 89 The language of the narrators is important in establishing J63 90 these contradictions. It is typically marked by apparent humility J63 91 or self-deprecation, though this is often actually "a means J63 92 of self-assertion and a means of commenting upon the limited roles J63 93 that women are expected to play". This can be seen J63 94 especially in the use of apparently or mockingly humble adverbs, J63 95 like 'perhaps' or 'possibly'. These adverbs appear to create an J63 96 allegedly female tentativeness, though one very often suspects J63 97 irony in this apparent self-deprecation, since it actually creates J63 98 a playful knowingness, often exploited for erotic effect. In 'The J63 99 Unfortunate Happy Lady' the assumed na<*_>i-trema<*/>vety of the J63 100 female narrator's voice touchingly and humorously dramatizes the J63 101 real innocence of her heroine Philadelphia, and we hear J63 102 simultaneously the two polarized female voices of the tale, the J63 103 innocent and the experienced: "She apprehended, that J63 104 (possibly) her Brother had a Mistress ..." (p. 45). A J63 105 similar device can be seen in 'The Court of the King of Bantam', J63 106 where the narrator assumes an apparent coy and tentative tone which J63 107 is actually playfully revealing and which asserts her authority J63 108 over even the "unfeminine" explicitness of the tale: J63 109 "... her Bed. Where I think fit to leave 'em for the J63 110 present; for (perhaps) they had some private Business" (pp. J63 111 31-2).

J63 112 Similarly, the narrator's claims to ignorance demonstrate not J63 113 so much narrative failure as the cool and perfect control of the J63 114 gentlewoman-amateur, an exasperating mockery of the reader for our J63 115 need for detail, verisimilitude, and coherent narrative structure: J63 116 "The rest I have forgot", "I had forgot to J63 117 tell you ...", "I had forgot to tell my Reader J63 118 ...". In light tales like 'The Court of the King of Bantam' J63 119 the narrator's apparent lack of authority and of certain knowledge J63 120 on some details serves to comic effect. Here the narrator is not J63 121 specifically female, and may possibly be, by implication, male J63 122 ("my Friend the Count", p. 27; "In less J63 123 Time than I could have drank a Bottle to my Share", p. 30). J63 124 If this is so, the apparent imperfection of his narrative authority J63 125 would serve to parallel the tale's mockery of Sir Would-be King and J63 126 the travesty of patriarchal authority that he represents. In more J63 127 serious tales, and where the narrator is more explicitly female, J63 128 the effect can, as we shall see, be different. The narrator may, J63 129 for instance, deliberately limit her field of expertise. Thus in J63 130 The History of the Nun, the female narrator concentrates J63 131 on her story of passion and moral paradox, for "it is not J63 132 my business to relate the History of the War" (p.304) which J63 133 provides a background; or a possibly female narrator will point out J63 134 the imperfections and inconsistencies in presumably male J63 135 authorities - "Some authors, in the relation of this J63 136 battle, affirm, that Philander quitted his post as soon as the J63 137 charge was given ...". Writing in a world where female J63 138 authorship was the subject of a vigorous and largely hostile J63 139 scrutiny by the representatives of the dominant culture, Behn has J63 140 her female narrators humbly accede to the view that female J63 141 creativity should be confined to certain fields, but this J63 142 transparently ironic humility does not so much accept the J63 143 conventional limitations as draw mocking attention to them. J63 144 "History" may be the locus of a specifically male J63 145 authority, but male "authors" are mocked by implication for J63 146 the imperfectness of their authority.

J63 147 Behn's narrators also offer a critique of the inequalities J63 148 encoded in the gendered language of society. Some words are J63 149 revealed to have different meanings depending on whether applied to J63 150 male or female subjects. Thus Sylvia would be "undone" by J63 151 losing her virginity, while Philander is "undone" by J63 152 failing to have sex and proving impotent at his first encounter J63 153 with her. Love Letters, and many other Behn tales and J63 154 plays, criticize society's language of gender not only by explicit J63 155 statements of the equality of men and women - they respond to J63 156 sexual passion with "equal fire, with equal languishment", J63 157 with "equal ravishment" (LL, pp. 53, 243) - but also J63 158 by allowing women to appropriate for their own uses a sexual J63 159 vocabulary in which they have previously been the objects of male J63 160 language. Thus men can be 'beautiful' and 'lovely' in the eyes of J63 161 women as much as women can in the eyes of men, women are J63 162 conventionally addressed by men as "charmer", but the word J63 163 can also, more unconventionally, be used by women of men, and while J63 164 Behn does not explicitly reject the conventional belief that women J63 165 are "the soft ... Sex", she also allows sympathetic J63 166 males to display "softness". Love Letters and J63 167 other tales thus imply a biological equality between the sexes, but J63 168 also allow their narrators to explore the socially constructed J63 169 inequalities.

J63 170 II J63 171 Behn's narrators, then, are often specifically female. The J63 172 implied reader may also be female, as she is invited to enter J63 173 closed female worlds of nunnery or brothel. A female reader is J63 174 constructed within the text, by, for instance, Behn's use of the J63 175 first person plural, which is not the authorial 'we' and contrasts J63 176 sharply with the narrator's jauntily individualistic 'I', but which J63 177 implies a female reader and a sympathetic complicity between her J63 178 and the author - "how wretched are our sex ..." (A J63 179 similar device can be found in Behn's plays: in the prologue to her J63 180 first performed play, The Forc'd Marriage (1670), women J63 181 in the audience are identified as "Spies" for the J63 182 "Poetess".) Behn's dedication of work to individual women J63 183 also suggests that she aimed for a female readership: the printed J63 184 text of The Feign'd Curtezans (1679) is dedicated, for J63 185 instance, to Nell Gwyn, and The History of the Nun to J63 186 Hortense Mancini, Duchess of Mazarine. The historical evidence, J63 187 fragmentary though it is, also suggests that many women read Behn's J63 188 work. Almost all women writers between the 1670s and the middle of J63 189 the eighteenth century are aware of Behn's example and had probably J63 190 read some of her works: Anne Finch knew Behn's work as a poet, J63 191 Catherine Trotter had read Agnes de Castro, which she J63 192 adapted for the stage in 1695, Jane Barker had by 1726 read at J63 193 least The History of the Nun and 'The Wandering Beauty', J63 194 which she retells in The Lining of the Patchwork Screen, J63 195 and many other women writers refer to Behn and her work. Mary J63 196 Wortley Montagu, whose reading is better documented than that of J63 197 most women, was acquainted with Behn's The Emperor of the J63 198 Moon, knew her poems well enough to quote two of them, fifteen J63 199 years apart, and had probably also read Oroonoko. In the J63 200 mid-eighteenth century Oroonoko was adapted and included in J63 201 one of the first women's magazines, thus reaching a still wider J63 202 middle-class audience. And, to take a final, celebrated example, J63 203 Mrs Keith of Ravelstone, as a young girl in London in about the J63 204 1760s, heard Behn's novels "read aloud for the amusement of J63 205 large circles of the first and most creditable society". J63 206 The female reader is both constructed within the text and a J63 207 historical reality outside it.

J63 208 However, while Behn expected and encouraged women to read her J63 209 tales, her female narrators sometimes imply, image, or address J63 210 specifically male readers, often for critical or ironic purposes. J63 211 The tales are full of female authors, writers, narrators, actors, J63 212 orators, and painters, producing works which are shown consumed by J63 213 a specifically male public. J63 214 J64 1 <#FLOB:J64\>The architecture and art of medieval Greece

J64 2 Byzantine church building

J64 3 Leaving aside the sparse but picturesque remains of strictly J64 4 Western church building in medieval Greece, as at Isova or J64 5 Zaraka (names in italics refer to sites discussed in the J64 6 Entries), we are left with two principal building forms of the J64 7 Byzantine world. The earliest and the simplest is the type known as J64 8 a basilica, where a rectangular central hall or nave, usually J64 9 separated from side aisles by rows of columns, and with clerestory J64 10 windows and a wooden roof, has a projecting apse at the opposite J64 11 end from that of the entrance; it was normally approached through a J64 12 colonnaded atrium, or open courtyard. Originating in pagan J64 13 antiquity as an assembly hall, this is the basic form of early J64 14 Christian basilica, and from its adoption for Christian worship in J64 15 the fourth century it became standard practice to give it an J64 16 east-west orientation.

J64 17 figure&caption

J64 18 In size it could vary greatly, and the largest, a few of which J64 19 still survive (e.g. the church of the Akheiropoietos, J64 20 Thessaloniki), were of great grandeur, many of them J64 21 embellished with spectacular mosaics. By the later seventh century J64 22 there were scores of buildings of this type on the Greek mainland, J64 23 although only a few are still visible, and even fewer standing. In J64 24 succeeding centuries the basilica design, when used for larger J64 25 churches, seems to have been reserved mainly for those which were J64 26 the seat of a bishop, and it was associated with this use (as in J64 27 the original building of the Mitropolis, Mystra) into the J64 28 later Byzantine period. It has been suggested that the conversion J64 29 of the Parthenon in Athens to a church provided a model J64 30 for this association.

J64 31 Early variations on the basilica form, besides the addition of J64 32 further side aisles, could involve the insertion of a transverse J64 33 aisle, or transept, close to the east end, and a porch, or narthex, J64 34 across the west end. During the sixth and seventh centuries the J64 35 grandest basilicas began to have a central cupola, as in the J64 36 Basilica B, Philippi. In the middle and later Byzantine J64 37 periods it was not common to build large new basilicas - those at J64 38 Serres, Veroia and of the Koimisis, Kalabaka J64 39 (which in any case probably replaced an early Christian building of J64 40 the same design), are unusual in their size. The great majority of J64 41 basilicas built after the ninth century tend to be quite modest in J64 42 size, and the absence of a cupola may be due for the most part to a J64 43 simple shortage of funds; the large number to be found at J64 44 Kastoria or Veroia are mostly small or very small and J64 45 must constitute a local exception, while those of Ag J64 46 Khristoforos, Mystra and Ag Ioannis Khrysostomos, J64 47 Geraki, are more typical.

J64 48 While the basilica form had links with antiquity, the second of J64 49 the two Byzantine building forms was developed exclusively in the J64 50 service of Christian architecture. Byzantine church building after J64 51 the seventh century becomes overwhelmingly an architecture that J64 52 exploits the dome. The basilica form was progressively superseded J64 53 from the sixth century by a variety of designs which have in common J64 54 a structure that is sufficiently strong to support a drum and J64 55 cupola over an open, square space. The main centres for J64 56 experimentation in these new designs would most probably have been J64 57 in Asia Minor and in Constantinople itself, and there is still J64 58 discussion as to where certain features originated. (The supreme J64 59 dome of the entire Byzantine period was always that of Ag Sofia, J64 60 Constantinople, which was built between 532 and 537, but the J64 61 architects of this building are known to have come from Asia J64 62 Minor.) The drum could be supported either on the four wall J64 63 sections abutting on to the crossing, on free-standing masonry J64 64 piers, on columns, or (very commonly) on a combination of one or J64 65 other system. In all cases the weight of the octagonal or circular J64 66 drum, with the cupola above it, had to be transferred down to the J64 67 four supporting members, and this was actually achieved by means of J64 68 pendentives; in their finest form these are a sophisticated piece J64 69 of engineering in which a concave section of a sphere receives the J64 70 weight of a quarter of the drum on its upper section and passes it J64 71 down to the single member (pier, column or wall section) out of J64 72 which it is built.

J64 73 The most common solution to this problem throughout all of J64 74 Greece became the so-called 'inscribed cross' or 'cross-in-square' J64 75 plan, and this was in common use by the eleventh century. It J64 76 involved the dome over the central square space of the naos being J64 77 supported either on four columns or, very commonly, on two columns J64 78 and the two wall sections either side of the main apse; to north, J64 79 south and west there were usually barrel vaults, with often a J64 80 fourth one to the east adjoining the apse. On the plan the dome and J64 81 its adjacent vaults thus appear as 'inscribed' within an J64 82 approximately square outer space. There are many possible J64 83 variations on this; these include greater or smaller depth of J64 84 lateral barrel vaults, the insertion of buttressing J64 85 semi-domes on eastern and western sides of the drum, J64 86 extending the western bay to enlarge the space of the naos, J64 87 inserting galleries in the crossing and thus adding a second J64 88 storey, and so on, but the one constant factor remained the J64 89 emphasis on the central dome.

J64 90 figure&caption

J64 91 At an early stage further spaces were added, such as the two J64 92 subsidiary apsed chambers built either side of the main apse; these J64 93 became standard in most churches, and are called the prothesis and J64 94 diaconicon; only the smallest churches might not have them. Almost J64 95 invariably the inscribed cross design incorporated a narthex to the J64 96 west, with sometimes a second, or exonarthex, as well, and larger J64 97 churches often had a gallery above this. Further domes could also J64 98 be added - either one over the bema, or over narthex and gallery, J64 99 or four more over each of the four corner bays. Indeed, either the J64 100 multiplication of domes, or the expansion of the central dome to J64 101 exceptional diameter or height, became the chief ways in which J64 102 builders or patrons could express the importance of a new church; J64 103 such expression was otherwise left to lavishness of patterns in J64 104 exterior brickwork, richness of interior decoration in the carving J64 105 on capitals and templa, and to colourful frescos or mosaics on the J64 106 upper surfaces of the building.

J64 107 In some, less usual, cases there was experimentation with the J64 108 use of further apsidal forms (or absidioles) to north and south, J64 109 and even to the west as well; examples of this are the churches of J64 110 Ag Apostoloi, Athens (Agora), the original form of J64 111 Ag Andreas, Peristera and the katholikon of the monastery J64 112 at Molyvdoskepastos. Such apsidal forms provided a more varied J64 113 exterior profile, with inside a more fluid and varied spatial J64 114 quality. Far more common, as a way of imparting individuality to J64 115 the exterior of churches, was the use of brick ornament; this could J64 116 be quite restrained, as in the ubiquitous form of string courses of J64 117 bricks set with projecting corners to produce the 'dog-tooth' J64 118 effect, or more lavish and varied as at Elassona or J64 119 Gastouni, or they could be set in blind arcades or niches. J64 120 This ornament is never found inside, as interiors were completely J64 121 covered by carefully developed didactic programmes of decoration in J64 122 fresco or mosaic, as we shall now see.

J64 123 Byzantine artistic practice

J64 124 Over the centuries the Byzantines developed an approach to J64 125 imagery markedly different from that prevailing in western Europe. J64 126 Mainly codified before and during the period of Iconoclasm J64 127 (726-843), it involved the concept of an image, when created in the J64 128 correct way, being a receptacle for the spirit of the person or J64 129 subject it portrayed. For the Byzantine beholder, an image, or J64 130 icon, was not just a likeness of a saint, but contained J64 131 the essence of the prototype, and so actually was the subject J64 132 portrayed. Other factors followed from this. If by reason of the J64 133 life he or she had led while in the world, a saint was worthy of J64 134 veneration, so the image of the saint was equally worthy; one of J64 135 the church fathers had already written in the fourth century: J64 136 "Honour paid to the icon passes on to the J64 137 prototype". To pay honour and respect to a person must J64 138 involve some form of communication, and so the iconic image is J64 139 invariably shown in a frontal pose (or nearly so) and with the eyes J64 140 open. To portray a figure in a way in which the beholder could not J64 141 communicate with it would have been pointless as it could not have J64 142 been 'understood'.

J64 143 The Byzantines had a strongly hieratic view of the universe, J64 144 and this too was clearly expressed in their art. Just as the most J64 145 sacred subjects, such as the feasts of the Dodecaorton, are of J64 146 greater sanctity than lesser episodes of Christ's life, so images J64 147 of these feasts are also more venerable. A hierarchy of sanctity J64 148 grew up as a product of this outlook, and the most complete J64 149 expression of the Byzantine view of the Christian universe became J64 150 the extended sequence of images that can be found in a Byzantine J64 151 church. The image of the most venerable of all subjects, Christ J64 152 Pantocrator, can usually be found in the highest part of the church J64 153 building - the summit of the cupola; below this are often found J64 154 adoring archangels. The dome of the church becomes identified in J64 155 this way with the dome of heaven. The second most sacred part of J64 156 the church, the main apse, houses the next most holy of images - J64 157 that of the Mother of God. Either some of the feasts of the J64 158 Dodecaorton, or else the four evangelists, can then be found in the J64 159 four pendentives supporting the cupola; below these, in the J64 160 soffits of the arches and in wall niches and subsidiary vaults, are J64 161 usually images of single saints from the service books of the J64 162 orthodox church. The image which displays the church's dedication J64 163 can often be found on the north wall adjacent to the templon. On J64 164 the lowest wall areas there was either a cladding of marble sheets, J64 165 often decoratively cut to form symmetrical patterns (as at J64 166 Osios Loukas), or more commonly, plaster painted to J64 167 imitate marble; this was ultimately a continuation of an antique J64 168 Roman practice.

J64 169 It should be emphasized that complete uniformity can never have J64 170 been intended; no two church interiors are ever identical, but J64 171 there are enough features in common to demonstrate the existence of J64 172 a general approach. It has even been suggested that the marriage of J64 173 the architecture with the imagery that it housed became so intimate J64 174 that it is difficult to visualize one without the other; certainly J64 175 some form of close collaboration between builders and painters or J64 176 mosaicists must have existed for such a perfectly integrated J64 177 concept to have become established. It was also within this context J64 178 that the use of portable icons was established from earliest times. J64 179 Today icons will be found in large numbers on later decorative J64 180 wooden iconostases, but the custom of displaying in some special J64 181 way icons of particular significance, some perhaps relating to the J64 182 dedication of the church, is probably quite an early one. The J64 183 practice of kissing icons, which is still universally observed, J64 184 also certainly has early origins.

J64 185 The modern visitor to a Byzantine church should be aware that J64 186 the forms of the building in which he finds himself were developed J64 187 in partnership not just with the imagery that they contain; they J64 188 would also have grown up in answer to liturgical needs, and can to J64 189 this extent be assessed on purely functional grounds. The orthodox J64 190 liturgy, although subject to modification and adaptation, from J64 191 early days took on certain formal needs which remained fairly J64 192 constant. In later centuries new demands also arose which needed to J64 193 be accommodated; the rite of baptism for example, had been very J64 194 important in the early church, thus necessitating a baptistry J64 195 building (as at Amfissa). This sacrament was given less J64 196 emphasis as the centuries passed, and separate baptistries were not J64 197 built, but it was revived towards the end of the Byzantine period, J64 198 and started again to have its own accommodation, usually in part of J64 199 the narthex. J64 200 J65 1 <#FLOB:J65\>Having produced his famous Tower monument in praise of J65 2 the Communist International (an uncompromisingly modern and J65 3 technologically daring project), Tatlin concerned himself with the J65 4 invention of an 'air bicycle', or glider, the Letatlin. It was J65 5 intended to be based on bird flight. It was ridiculed as primitive J65 6 by some of the technicist Soviet critics of the time. Tatlin J65 7 replied, in terms similar to those of Malevich's memoirs:

J65 8 I want to give back to people the feeling of flight. J65 9 This we have been robbed of by the mechanical flight of the J65 10 aeroplane. ... We can no longer feel the movement of our body in J65 11 the air.

J65 12 As he grew older and more isolated, Tatlin was referred to by J65 13 some in vaguely pitying terms as a clown, or Big Fool. In fact his J65 14 life exemplified the real dilemmas of a struggle with the dichotomy J65 15 between art and technology. If in his Tower he used his J65 16 revolutionary outlook and artistic imagination and daring to reveal J65 17 the possibilities of technological innovation, which would J65 18 otherwise be assimilated to archaic, traditional patterns of J65 19 thinking and daily life, with his glider he seemed to resist the J65 20 inevitable drift and rationale of mechanical, industrial culture. J65 21 He insisted that he conceived the glider as an artist, that it J65 22 should be aesthetically perfect. Against the 'iron laws of J65 23 technology' it was considered a failure. Yet although his proposal J65 24 was taken for what it partly was - a critique of social life and J65 25 values - in some ways Tatlin was still bound by technological J65 26 priorities. He still felt, like Leonardo before him, that the J65 27 'feeling of flight', which is really a psychological desire and J65 28 dream of freedom, could be satisfied by a machine. The drama of his J65 29 whole position was to be caught between the real and the imaginary, J65 30 between art as a representation of life, and life itself.

J65 31 MEXICANIDAD J65 32 Quite different issues are raised by the itineraries of the J65 33 Mexican painter Diego Rivera. He made the aspiring artist's journey J65 34 from periphery to centre: he travelled to Italy, and to Paris where J65 35 he painted as one of the cubists. He also absorbed at first hand J65 36 the energy of the two emergent global powers, the USA and the USSR, J65 37 and he joined those other Mexican intellectuals whom the J65 38 revolutionary upheaval of 1910-20 spurred to travel widely over J65 39 their own country for the first time. But the key move of Rivera's J65 40 was from the metropolitan art centres back to his own country at a J65 41 time when Mexico was reasserting itself against the effects of J65 42 centuries of invasion and brutalization by outsiders. He radically J65 43 changed his style and began to produce the monumental, didactic J65 44 murals in which he set out to reinvent the history of Mexico (and J65 45 its modern development) in visual images for a still-overwhelmingly J65 46 non-literate population.

J65 47 In Paris Rivera may well have absorbed the avant-garde J65 48 discovery of 'primitive' and non-European art: some of the J65 49 evocations of Mexico in his early murals appear to owe something to J65 50 the paradisal tropics of Gauguin and Rousseau. But in Mexico such J65 51 'primitivism' became absorbed in the powerful and complex movement J65 52 of indigenism or nativism which was a driving force of the J65 53 revolution in the area of culture and which spread across all J65 54 disciplines (art, architecture, archaeology, folklore, music, J65 55 poetry, education, literature, medicine, and so on). J65 56 Anthropologists define nativism as an organized and conscious J65 57 effort on the part of members of a society to revive parts of its J65 58 own culture - in Mexico this implied not only national resurgence J65 59 but also the rediscovery of indigenous cultural values despised by J65 60 most of the ruling class within Mexico itself.

J65 61 Characteristically, Rivera's style evolved from primitivist J65 62 projections to an extraordinary effort of factual research into J65 63 ancient and folkloric Mexico, which he converted into painted J65 64 images. In a late mural like the one in the Hospital de la Raza in J65 65 Mexico City, Rivera gave a vast and detailed portrayal of Aztec J65 66 medical practices equal space with a portrayal of modern medicine. J65 67 This in itself was a bold assertion, but when one looks longer at J65 68 the layout of this painting one sees complex tensions of fact, J65 69 fantasy, and desire. The Aztec scenes are painful but open and J65 70 communal; the modern ones painless but torn by class strife and J65 71 segmented into alienating boxes. The very colour and conviviality J65 72 of the Aztec scenes show the way in which Rivera has marshalled and J65 73 constructed an ancient world to point beyond the present to a J65 74 Utopian communist future.

J65 75 Rivera's gargantuan project and his position in J65 76 twentieth-century art raise many questions. Interestingly, some of J65 77 the contradictions of his art become clearest in the comparison J65 78 between his own painting and that of his wife Frida Kahlo, J65 79 certainly his equal as an artist. In keeping with his J65 80 socio-political role as educator and illustrator, his relationship J65 81 to the masses always seems to be one of depicting and organizing J65 82 from the outside. He stylized both ancient and modern J65 83 Mexicans as a people, as a historical force. Rivera's contemporary, J65 84 the Peruvian critic Jos<*_>e-acute<*/> Carlos J65 85 Mari<*_>a-acute<*/>tegui, had said that the "idealization J65 86 and stylization of the Indian" was an inevitable feature of J65 87 'indigenist' literature in Latin America. An 'indigenous' J65 88 literature would appear only "when the Indians themselves J65 89 are able to produce it".

J65 90 It is revealing too that women could enter Rivera's murals only J65 91 allegorically. Even though very often they were his friends or J65 92 lovers, in the mural they took on a generalized, abstracted J65 93 persona, an actor either in historical events or in the world of J65 94 concepts: 'fertility', 'America', 'agitation', etc. Frida Kahlo's J65 95 method - although she lived with Rivera for so many years and was J65 96 as militantly communist as he - was almost opposite. She had no J65 97 links with the tradition of 'high art' and architecture. She was J65 98 self-taught and because of ill-health was almost forced to work on J65 99 a domestic scale, and even from her bed. Where Rivera generalized, J65 100 Kahlo particularized, thinking on the level of the individual and J65 101 lived experience. When Kahlo came to represent the Indian, it was J65 102 the Indian in herself. My Nurse and I (1937) conveys the J65 103 idea of being nurtured by the indigenous culture of Mexico in as J65 104 intimately close, as physical, a metaphor as one can imagine (Kahlo J65 105 herself was of mixed ancestry: German on her father's side, Spanish J65 106 and Indian on her mother's). By showing herself as an adult she J65 107 implies the continuation of this nourishment throughout her life. J65 108 She suggests the 'continuity of consciousness', claimed by J65 109 revolutionary archaeologists between the ancient and contemporary J65 110 Indians of Mexico, by giving her nurse an archaic mask; this in J65 111 turn modifies a comforting image into one of power.

J65 112 While Kahlo and Rivera shared the same enthusiasm for Mexican J65 113 popular culture, Kahlo's relationship to these sources was less J65 114 schematic, more organic than Rivera's. Her own work was strongly J65 115 influenced by the style of popular retablo, or J65 116 ex-voto paintings, where the traumas of individual J65 117 lives are rendered with cruel and often bloody directness. In one J65 118 sense these vernacular images represent the breaking-out of a J65 119 suppressed voice; the same can be said of Kahlo's work as a woman J65 120 artist. Through the individual she arrived at the general, the J65 121 collective.

J65 122 Nativism seems to have been an extremely complex phenomenon in J65 123 Mexico. A more subtle analysis would have to take account of class J65 124 (in the dichotomy between Rivera's 'personality' portraits of the J65 125 rich and famous and his 'humble' portraits of the poor, it is hard J65 126 not to see a form of primitivism which has been avidly consumed as J65 127 such by Mexico's upper classes). And this is further complicated by J65 128 the tendency of Europeans to see Mexican nativism in primitivist J65 129 terms. In fact there is a sense in which the Indian in Mexico J65 130 became part of Europe's argument with itself. The elements in J65 131 Mexican culture formerly thought inferior to European elements were J65 132 made visible and celebrated in Mexicanidad, and gradually J65 133 became fashionable, rather in the way Africanity did in Paris in J65 134 the 1920s. Rivera and the other artists had done much to define J65 135 this Mexicanity. And although, unlike Africanity, Mexicanity was J65 136 constructed 'at home', it could be said to have created an J65 137 imaginary Mexico which would later have to be 'deconstructed' in J65 138 the name of the real.

J65 139 ANOTHER AVANT-GARDE

J65 140 There is a significant sequel to this story of the Mexican J65 141 experience: its almost complete suppression in European and North J65 142 American histories of twentieth-century art soon after the Second J65 143 World War. In his Concise History of Modern Painting, J65 144 which has become a standard textbook since its publication in 1959, J65 145 Herbert Read actually declared that he was deliberately leaving out J65 146 the Mexican artists. He left out much else besides, in fact the J65 147 whole relationship of art to social change. But this was not J65 148 unusual. With the growth of the art market, the philosophy of the J65 149 Cold War, the aggressive exportation of western culture and J65 150 life-style to the rest of the world, especially by the USA (an J65 151 exportation, overwhelmingly, of objects), modern art was J65 152 constructed almost as a western capitalist monopoly enterprise. J65 153 This has made the cultural expression of the power relations more J65 154 complex. We see, for example, the projecting on an international J65 155 scale (with the help of the art market, of corporate J65 156 self-interest, of museums and publicity), as in some way J65 157 universally 'modern', of art which actually embodies local or J65 158 national myths ("Pop Art is American ethnic art" - J65 159 Rasheed Araeen). Many major 'Third World' cities today (Rio de J65 160 Janeiro, Caracas, Buenos Aires, Manila, Delhi, etc.) are as J65 161 instantly well informed of developments in Paris or New York as J65 162 those cities themselves, sometimes more so. But the traffic is J65 163 always in one direction, confirming their satellitization in the J65 164 world of multi-national capitals. Their reception, their J65 165 critical transformation of new ideas, let alone their own J65 166 discoveries, remain unknown and forcibly localized. New dialectics J65 167 between local and global, between art centres and the rest, are J65 168 coming into existence. On the one hand, artists of Third World J65 169 origin find themselves marginalized within western capitals or J65 170 separately categorized according to an exclusive mainstream. On the J65 171 other, there are numbers of artists (film-makers, theatre groups, J65 172 etc.) working 'locally' in the Third World but with modern J65 173 techniques and world connections.

J65 174 Any attempt to describe the relations of these artists to the J65 175 dominant centres of power in the post-war world would produce the J65 176 most complex map: of emigrations, of returns, of exile, of staying J65 177 home. Such conditions are not generally known, let alone the J65 178 significance of an individual's movement and practice between J65 179 different social spaces and cultural contexts. But a particular J65 180 perspective is shared by all those who take the question radically J65 181 'from the other side'. Their consciousness of the daily J65 182 socio-economic realities of 'underdevelopment' - of which a J65 183 colonized culture is part - becomes inseparable from their critique J65 184 of art, of its commercialization and neutralization in western J65 185 cultural institutions. This perspective is expressed too, I J65 186 believe, in radically different meanings and uses of J65 187 'primitivism'.

J65 188 Consider, for example, the work of two outstanding Brazilian J65 189 artists, Lygia Clark (1920-88) and Helio Oiticica (1937-80), of J65 190 which the 'international museum circuit' remains ignorant. In the J65 191 1950s a 'universal' modernist movement was tested out in Brazil: J65 192 constructivism. It flourished as part of the post-war economic and J65 193 construction boom, answering the desire of dynamic sections of the J65 194 Brazilian society to create an 'absolutely modern' environment. J65 195 Both Lygia Clark's and Helio Oiticica's early work was done in the J65 196 constructivist vein (and has always since retained some traces of J65 197 its geometric ordering). But almost at once they began to question J65 198 its application as a ready-made model to Brazilian conditions, and J65 199 to explode the limitations of its rationale.

J65 200 According to a technicist notion of constructivism, Lygia J65 201 Clark's and Helio Oiticica's work apparently 'went back' to J65 202 'primitive' materials, to the body, to 'primordial' sensations, J65 203 relationships. In fact its radicalism was of another kind. An acute J65 204 sensitivity to traditions and tensions in their own environment is J65 205 combined with a searching questioning of the artist's production in J65 206 a corporate, consumer society.

J65 207 By the early 1960s, Lygia Clark had broken with the traditional J65 208 idea of sculpture as a detached object, in which the body's energy J65 209 and the artist's expressive power are somehow captured, frozen. J65 210 J66 1 <#FLOB:J66\>PORTRAITURE AS HISTORY AND THEATRE

J66 2 The insistence that portraiture could serve the same edifying J66 3 function as history painting was to some extent opportunistic. J66 4 Theorists such as Richardson and Reynolds had their own interests J66 5 as practising artists to serve, and the promotion of portraiture as J66 6 a serious branch of art lent respectability to their efforts. J66 7 However, throughout the eighteenth century, artists, critics and J66 8 patrons acknowledged that portraiture could and did have a function J66 9 beyond the obvious one of 'face painting'. Its perception as J66 10 history accompanied the growth of a genre of theatrical J66 11 portraiture, which, in its initial stages, was primarily J66 12 commercial. By the end of the century, changes in the theatre J66 13 itself and in the perception of actors gave the stage added status J66 14 as a subject and concern for painters. It was understandable J66 15 therefore that artists should see in the theatre a means of fusing J66 16 portraiture and history, and Lawrence's portraits represent the J66 17 climax of this tendency.

J66 18 From the seventeenth century, the equation of portrait J66 19 patronage with personal vanity was denied by those who saw a more J66 20 public role for the genre. Many country houses had a 'gallery of J66 21 worthies' containing either painted or sculpted depictions of J66 22 famous men and, less frequently, women, from the past and present. J66 23 Such galleries were intended to suggest the learning and judgment J66 24 of the house's owner, but they were also interpreted as an J66 25 incitement to noble behaviour. The portrait's power was seen to lie J66 26 more in the choice of the person represented than in the quality of J66 27 the painting itself, so that the Brown Gallery at Knole contained J66 28 portraits of worthies to which the names were attached arbitrarily. J66 29 To counterbalance this tendency to ignore the portrait's aesthetic J66 30 value, Jonathan Richardson insisted that its true power lay in the J66 31 skill of the artist to reveal the nobility of the sitter's J66 32 character. Richardson, who equated the genre's status with that of J66 33 history painting, defined a portrait as "a sort of General J66 34 History of the Life of the Person it represents, not only to Him J66 35 who is acquainted with it, but to Many Others". By J66 36 successfully depicting a 'great man', a portrait painter could J66 37 therefore convey the idea of greatness to an observer, whether or J66 38 not that observer had any foreknowledge of the sitter. The portrait J66 39 thus transcended the specificity of its subject by focusing on the J66 40 salient virtues of the sitter's character. With such a theoretical J66 41 basis, portraits could be seen as having a public role, and one J66 42 critic even went so far as to suggest that portrait painting should J66 43 become for England what history painting was for Italy.

J66 44 Later in the century, this conception of portraiture permeated J66 45 Royal Academy lectures and discourses. Reynolds referred to the J66 46 genre infrequently and cautiously, but in his Fourth Discourse he J66 47 implicitly acknowledged the value of portraiture that was J66 48 "general" rather than "specific": "but it J66 49 happens in a few instances, that the lower may be improved by J66 50 borrowing from the grand. Thus if a portrait-painter is desirous to J66 51 raise and improve his subject, he has no other means than by J66 52 approaching it to a general idea". In this same Discourse, J66 53 Reynolds deplored the specificity that resulted when a portrait J66 54 painter attempted history, as "An History-painter paints J66 55 man in general; a Portrait-painter, a particular man, and J66 56 consequently a defective model." Here he identified the J66 57 essential problem of specificity, which not only hampered the J66 58 portrait painter searching for a means of universalizing his works, J66 59 but potentially threatened the universality of history painting J66 60 itself. Reynolds's warning was undermined by the wild success of J66 61 West's and Copley's modern history paintings, which drew their J66 62 popularity in part from the number of identifiable portraits J66 63 contained within them. This dissolving of the boundaries between J66 64 portraiture and history led Northcote to contend "Portrait J66 65 often runs into history and history into portrait, without our J66 66 knowing it", and by 1809, one critic could argue the J66 67 benefits of including portraits in history painting, and look J66 68 forward to the time when West's example "will ultimately J66 69 banish all imaginary monsters from the historical canvass J66 70 [sic]". Writing in the Anti-Jacobin Review, this J66 71 critic was responding to Opie's Royal Academy lectures, and J66 72 challenging that artist's scorn for those who were interested only J66 73 in "likenesses". Significantly, this partisan reviewer saw J66 74 portraits of 'great men' within a war-weary country as serving an J66 75 inspirational and educative function.

J66 76 The nature of the relationship between portraiture and history J66 77 painting was explored most subtly by Fuseli, who interpreted the J66 78 former as being concerned, broadly, with the depiction of J66 79 character: "Portrait is to historic painting in art, what J66 80 physiognomy is to pathognomy in science. That shows the J66 81 character and powers of the being which it delineates, in its J66 82 formation, and at rest; this shows it in exertion." J66 83 Fuseli's sensitivity to the role of portraiture within history J66 84 painting is revealed through his invention of a category of J66 85 "dramatic" painting, which, together with the J66 86 "historical" and the "epic", formed his three J66 87 branches of history in painting. The introduction of this dramatic J66 88 argument into a predominantly historical agenda is a significant J66 89 change of emphasis, and one which reflects the growing importance J66 90 of a theatrical model to the historical painters of the Royal J66 91 Academy.

J66 92 Theatrical portraiture emerged as a sub-genre in the 1740s and J66 93 50s when David Garrick became aware of the valuable publicity to be J66 94 gained from depictions of himself in character. Topical portraits J66 95 of this kind proved successful commercial speculations for both J66 96 actor and artist, particularly when shown at a public exhibition. J66 97 Because they were primarily commercial in their content and J66 98 intention, theatrical portraits were at first held in low regard. J66 99 Actors and actresses, struggling to increase their status in a J66 100 mobile society, would often take offence when portraits in street J66 101 dress were exhibited with allusions to their profession, perceiving J66 102 an important difference between a sensational image of public J66 103 performance and a private declaration of status. Lawrence himself J66 104 suffered the outrage of Elizabeth Farren, when her portrait was J66 105 accidently<&|>sic! exhibited under the title 'An Actress', rather J66 106 than 'A Lady'. Equally, critics of the Shakespeare Gallery warned J66 107 artists not to look to the stage for their inspiration, fearing J66 108 (not without cause) that the exhibition could turn into a mere J66 109 collection of costume portraits. However, the contempt with which J66 110 theatrical portraits seemed to be held was offset by more positive J66 111 responses. 'Anthony Pasquin's' often brutal criticisms of Royal J66 112 Academy exhibitions included praise of Samuel de Wilde's theatrical J66 113 portraits, and Reynolds, who professed to abhor "the J66 114 theatrical pomp and parade of dress and attitude" was the J66 115 eager purchaser of Zoffany's portrait of Garrick as Abel Drugger. J66 116 Reynolds also painted several theatrical portraits, including J66 117 Mrs. Abington as Miss Prue (1771; New Haven, Yale J66 118 Centre), which offended against his own rules of specificity.

J66 119 It was Reynolds who first used theatrical subjects as a means J66 120 of realizing his ambition to elevate the genre of portraiture. J66 121 Several of his theatrical portraits show his attempt to cast this J66 122 genre into a pseudo-historical mould. Garrick Between Tragedy J66 123 and Comedy (1760-62; private collection) plays learnedly on J66 124 the theme of the Judgment of Hercules and alludes to Shaftesbury's J66 125 Characteristicks. Such works came to be read as transcending J66 126 the narrow bounds of both face painting and theatrical portraiture. J66 127 Thus Mrs. Abington as the Comic Muse (c. 1769; J66 128 Waddesdon) was the subject of a commemorative poem by Edward J66 129 Jerningham, who declared that the portrait, then hanging at Knole, J66 130 acted as a sort of "talisman" against sadness, while J66 131 Reynolds's depiction of the actress Fanny Kemble in street dress J66 132 was referred to as an "instructive canvas" exciting J66 133 "moral worth". Here Richardson's idea that a J66 134 portrait could instruct by merely exhibiting the virtue of its J66 135 sitter makes its way into ephemeral criticism. But it was J66 136 Reynolds's portrait of Mrs. Siddons as the Tragic Muse J66 137 (plate 41) that went the furthest towards exemplifying his own J66 138 ideals, and which had the greatest impact on Lawrence.

J66 139 In his otherwise sparse and somewhat lame academy discourses J66 140 Lawrence's observations on Reynolds's portrait of Siddons emerge as J66 141 a strong testimony to his interest in this subject. Calling the J66 142 painting "a work of the highest epic character, and J66 143 indisputably the finest female portrait in the world", J66 144 Lawrence goes on to praise Reynolds and Siddons for the cooperation J66 145 which led to the painting's creation. By using the term 'epic', J66 146 Lawrence implicitly links the portrait with Fuseli's highest J66 147 category of history painting. Lawrence's own evaluation of J66 148 theatrical portrait to epic history painting was made possible by J66 149 the public perception of Siddons, and his choice of Siddons's J66 150 brother, John Philip Kemble, as a theatrical/historical subject J66 151 equally owed much to the public perception of that actor.

J66 152 The status of the actor in the eighteenth century was enhanced J66 153 when regulation of the theatres led to a monopoly, allowing J66 154 Garrick, as manager of Drury Lane, to gain an unprecedented J66 155 fortune. Garrick used his wealth and fame as a passport to genteel J66 156 society, and Kemble followed his lead when he assumed the J66 157 management of Drury Lane under Richard Brinsley Sheridan's J66 158 jurisdiction. Garrick's ascent had aroused as much suspicion as J66 159 admiration, with observers such as Horace Walpole rejecting his J66 160 claims to gentility as mere emulation. Kemble however benefited J66 161 from Garrick's example and inspired greater regard for his personal J66 162 qualities. Just as his private character was declared to be above J66 163 reproach, so his public persona was represented as exhibiting the J66 164 values of nobility, dignity and restraint. Haydon reported that J66 165 "Kemble as an Actor & West as a Painter have more claims on J66 166 society from the honour of their private characters, than the J66 167 greatness of their genius", but the Gentleman's J66 168 Magazine extended his claim without qualification, referring J66 169 to Kemble as a "man in whom private worth unites itself to J66 170 public activities". His image was enhanced by an J66 171 association with primarily tragic roles: unlike Garrick, who J66 172 "would play Punch rather than not be acting", J66 173 Kemble "was impressed with a higher sense of the J66 174 dignitiy and utility of his art", as the J66 175 portrayal of comic characters was felt to be "below the J66 176 attention of such a performer". That such qualities were J66 177 the result of a critical construction can be proven by comparing J66 178 them to adverse criticisms which viewed his acting as sloppy, J66 179 ranting and overtly emotional. Apologists explained his 'classical' J66 180 style in terms that were appropriated from Reynolds's J66 181 Discourses:

J66 182 The truth is, that the advocates for natural representation, J66 183 forget what Mr. Kemble always bore in mind, that it is the best J66 184 part of Nature only, which should be faithfully given; that J66 185 stooping to represent the common effects of common life, either in J66 186 person or action, degrades the character and the art.

J66 187 Kemble's Coriolanus was extolled as "the beau J66 188 id<*_>e-acute<*/>al of Rome"; his stage presence was J66 189 compared to the effect of the Apollo Belvedere and the Laocoon; and J66 190 his later biographer, the dramatist James Boaden, quoted Reynolds's J66 191 Seventh and Thirteenth Discourses in an attempt to classify the J66 192 special nature of his acting. Boaden referred to Kemble's method as J66 193 "academic or critical", and distinguished between J66 194 the "academic" and "vulgar" styles of performance, J66 195 just as Reynolds had polarized the "academic" and J66 196 "vulgar" styles of painting in his Discourses.

J66 197 The appropriation of Reynolds's terminology was more than J66 198 simple plagiarism. Establishing Kemble's credentials, even in J66 199 retrospect, was crucial, as from the end of the eighteenth century J66 200 the stage was seen as having a political as well as moral purpose, J66 201 "animating by example the love of a country, public J66 202 courage, and all those virtues that must grace the English J66 203 character, keep them free from the yoke of foreign tyranny and that J66 204 unparalleled catalogue of vices existing in a neighbouring J66 205 nation". Drury Lane was under the management of the Whig J66 206 and sometimes radical Sheridan, and when Kemble purchased part of J66 207 the rival theatre Covent Garden in 1803, popular polemic saw him J66 208 either as representative of ancient virtues, order and hierarchy, J66 209 or as tyrannical and autocratic - both contrasting with Sheridan's J66 210 policies. As a manager Kemble was attacked for his "want of J66 211 respect for the public" and "aristocratic J66 212 impulse" or praised for his disdain for popular opinion. J66 213 Both detractors and defenders contrasted Kemble's aloofness with J66 214 the populism of Sheridan, who put the desires of the majority of J66 215 his audience for proper seating and accessible entertainment before J66 216 those of the privileged few. J66 217 J67 1 <#FLOB:J67\>The problem is compounded by the fact that uncertainty J67 2 concerning the undoubtedly different appearance of the models that J67 3 lie behind the two books makes it highly problematic to assess the J67 4 nuances of their stylistic rapport. However, having said this, it J67 5 will be observed that although they share the same attitudes to J67 6 composition and to the rendering of bricks and drapery folds, the J67 7 miniatures of the two books are clearly distinguished from each J67 8 other in terms of facial types and the use of colour. The tones J67 9 favoured in the Lectionary, whilst bright, are nevertheless more J67 10 harmonious and less abrasive than those which were generally J67 11 preferred in the Troper. The lectionary is closer to the mainstream J67 12 of Anglo-Saxon art than is the Troper. As no more than one artist J67 13 seems to have worked in either manuscript, at the very least we can J67 14 safely conclude that the two books are not contemporary works of J67 15 the same artist. The main texts of the two volumes were written by J67 16 different hands of quite different aspects (neither of which have J67 17 been recognized elsewhere). Nor are the monumental display scripts J67 18 of the two books particularly close. On the other hand, a modest J67 19 scribal relationship between them is provided by their J67 20 inscriptions. The small inscription b.ciliccus J67 21 (basiliscus) written up the side of the page, beside the Q to J67 22 Luke's gospel in the Lectionary, is comparable in arrangement and J67 23 script (rustic capitals) to the inscriptions written around all J67 24 eleven surviving pictures in the Troper. Whether these similarities J67 25 amount to sufficient evidence for assigning both books to the same J67 26 centre of production is open to doubt: even if they be regarded as J67 27 earlier and later works of the same artist this need not J67 28 necessarily imply that they were executed at the same centre. The J67 29 safest and most reasonable course for the present is surely to J67 30 group them more loosely, as the products of related artistic J67 31 milieux. Now, far from being an anticlimax, the suggestion of a J67 32 looser association between them is itself of some import, implying J67 33 as it does that this divergent, progressive style - practised J67 34 either by two artists or by one developing his idiom over a number J67 35 of years, in either case possibly at different centres - was J67 36 potentially a more widespread phenomenon.

J67 37 Returning to the issue that is of primary concern here, namely J67 38 that of clarifying and defining more precisely the nature of the J67 39 innovative qualities of these two decorated books, there are three J67 40 basic points which should be made.

J67 41 In the first place, since art historical discussion has tended J67 42 to focus on the miniatures alone, emphasizing their atypical J67 43 qualities, it is worth stressing that the organization, J67 44 mise-en-page and writing of the two manuscripts as a J67 45 whole are not comparably unusual. Notwithstanding certain J67 46 constructional anomalies (particularly in the Troper), both books J67 47 are generally well produced, written in well sustained, fairly J67 48 elegant late English Caroline minuscule, with capitals in gold, J67 49 red, blue and lilac. The text hand of Caligula, notable for the J67 50 flowing, curling tail of g and for the use of two forms of q - a J67 51 regular minuscule, and a diminutive majuscule - is difficult to J67 52 parallel closely, but the general aspect of the neat, rectilinear, J67 53 mainly upright script with triangular serifs on the ascenders finds J67 54 analogies in several manuscripts of mid- to late eleventh-century J67 55 date, including volumes from Exeter and Bury St Edmunds. The script J67 56 of the Lectionary is strongly reminiscent of Worcester bookhands of J67 57 the third quarter of the eleventh century. In size and proportions J67 58 the Lectionary compares closely with the Gospel Lectionary of St J67 59 Margaret. The Troper is a little smaller than the Troper in the J67 60 Bodleian and somewhat larger than that in Corpus Christi College, J67 61 Cambridge. Like the Cambridge Troper (and other music manuscripts), J67 62 the minuscule text in Caligula is very small in relation to the J67 63 space between the ruled lines, allowing plenty of room for the J67 64 neums above; and, as is the case in both unillustrated Tropers, J67 65 coloured majuscules and rustic capitals play a crucial role in J67 66 subdividing and organizing the text. The use of projecting J67 67 ascenders and descenders to characterize and accentuate display J67 68 script is another important feature in the design of the Troper J67 69 which is paralleled in numerous other contemporary manuscripts.

J67 70 Secondly, consideration of such analogies as may be found for J67 71 the distinctive style of the artwork in these two books reveals it J67 72 as a native tradition reformulated apparently through familiarity J67 73 with certain continental works. Although rooted in the vocabulary J67 74 of Anglo-Saxon manuscript decoration of the 1020c, the style J67 75 attests to a new relationship with contemporary illumination in J67 76 Flanders and Lotharingia. Modest reflections of the same manner are J67 77 to be seen in several other English books dating from the last J67 78 three decades of the eleventh century.

J67 79 The ornamental initial in the Troper and many of the decorative J67 80 features of the Lectionary, particularly in its canon tables, are J67 81 appreciable as a development of the art of the Eadui Gospels and J67 82 the Arundel 155 Psalter, Christ Church manuscripts of the previous J67 83 generation. The treatment of the figures is reminiscent of that in J67 84 the Old English Hexateuch. The choice and use of colours - heavy J67 85 tones scumbled with white - invites comparison with the Hexateuch, J67 86 the mid-eleventh-century Canterbury Rule of St Benedict, J67 87 the Illustrated Miscellany and the later Exeter Gospels. On the J67 88 other hand, the best analogies for the general style of most of the J67 89 figural compositions are undoubtedly found in eleventh- and J67 90 twelfth-century manuscripts from Flanders and Lotharingia. The J67 91 influence of styles or indeed the participation of artists from J67 92 these areas is quite possible in view of England's continuing J67 93 contacts with Flanders and Lotharingia in the mid-eleventh century. J67 94 There were, it will be recalled, four Lotharingian bishops in J67 95 England during the reign of Edward the Confessor. Given that the J67 96 Gospel Lectionary is known to have been at Hereford in the late J67 97 eleventh century, while it has been suggested that the Troper may J67 98 have been made for Edward the Confessor, it is interesting to J67 99 recall that the bishop of Hereford during the period 1061-79 was J67 100 Walter of Lotharingia, who had come to England with Edward and had J67 101 formerly been a royal chaplain.

J67 102 Echoes of the style of the two books are found in two English J67 103 Romanesque manuscripts: the Shaftesbury Psalter and a closely J67 104 related copy of Boethius, De consolatione philosophiae, J67 105 which is now in the Bodleian Library; both are dated to the second J67 106 quarter of the twelfth century and are associated with the West J67 107 Country. However it is open to question whether these formal J67 108 similarities represent an actual connection between the two J67 109 Anglo-Saxon and the two Romanesque books, as opposed to being the J67 110 coincidental result of independent influence from related J67 111 Continental sources. Of more immediate relevance are certain J67 112 parallels to be seen in the decoration of a few English manuscripts J67 113 of late eleventh-century to early twelfth-century date. For J67 114 instance, to begin with the most limited case, the distinctive J67 115 manner of rendering bricks, but not any other feature, is matched J67 116 in one of the drawings in the Cotton Caligula A.XV computistica, a J67 117 Christ Church book dating from c. 1073. More impressive are J67 118 the analogies provided by Durham Cathedral Library, MS B. II. 16, a J67 119 copy of Augustine, In evangelium Iohannis, made at St J67 120 Augustine's, Canterbury, towards the end of the eleventh century J67 121 and sent to Durham shortly afterwards. As the decoration of the J67 122 book consists purely of initials, not miniatures, this naturally J67 123 limits the extent of the parallels; nevertheless the same foliate J67 124 forms reappear and a number of the figural initials are painted in J67 125 a manner akin to that used in the Troper - rich colours, thickly J67 126 applied, their tones modified by the application of white (fig. J67 127 25). Another book, Cambridge University Library, MS Ii. 3. 33, a J67 128 copy of Gregory's Registrum epistolarum ascribed to J67 129 Christ Church and dating from between 1079 and 1101, has three J67 130 decorated initials painted in a related style (all on fol. J67 131 5r), although with less use of white. The fact that these J67 132 analogies as indeed the earlier eleventh-century English ones occur J67 133 in manuscripts of Canterbury, particularly Christ Church, origin is J67 134 worth noting; on the other hand this in itself is insufficient J67 135 grounds for ascribing the Troper and the Lectionary to that centre. J67 136 The principal importance of the three later Canterbury books is J67 137 that they reveal that the stylistic innovations of the Troper and J67 138 particularly the Lectionary were not wholly isolated in the context J67 139 of English art in the second half of the eleventh century: J67 140 affiliated modes of decoration were evidently being explored, J67 141 albeit in a discreet way, at Canterbury in the decades after the J67 142 Norman Conquest.

J67 143 figure&caption

J67 144 The third point, which relates specifically to the Troper, is J67 145 the essential originality of its decorative programme. Fine J67 146 illustrated early medieval tropers are very rare, and the sum total J67 147 of decoration in the other two surviving Anglo-Saxon tropers is a J67 148 single ornamental initial. Furthermore, irrespective of the J67 149 irresolvable issue of whether or not there may once have existed J67 150 other Anglo-Saxon tropers with more extensive decoration, the J67 151 idiosyncratic arrangement of the miniatures within the surviving J67 152 portion of the Caligula Troper, the discrepancies in their size and J67 153 finish, their unusual subject matter and their divergent J67 154 iconographies and decorative features argue strongly in favour of J67 155 this being the first manuscript in which these pictures were added J67 156 to this text. Moreover, the way in which they were actually applied J67 157 to the book, far from reflecting an inherited layout, suggests J67 158 rather that their inclusion represented an afterthought or a change J67 159 in plan. Some of the pictures, only four of which occupy a full J67 160 page (the remainder vary in size), were clearly squeezed into J67 161 spaces that were hardly sufficient to receive them; and on one J67 162 occasion the addition of a miniature obliterated at least one line J67 163 of text and probably more.

J67 164 A thorough review of the iconography of the miniatures and J67 165 their various analogues remains a desideratum; nevertheless, the J67 166 basic point that there is a range of unusual subjects and an J67 167 eclecticism of iconography and design here for which a single model J67 168 could not account is not in doubt. Five of the miniatures - the J67 169 Naming of John, the Release of St Peter, the Temptation of St J67 170 Martin, the Annunciation to Joachim, and Joachima and Anna with J67 171 Mary - are the only known representations of the subject in J67 172 question in an Anglo-Saxon manuscript, while the treatment of the J67 173 subjects that do occur elsewhere remains distinctive, even when a J67 174 familiar iconographic type provides the basis of the work. Thus in J67 175 the case of the two-part miniature of the Passion of St Lawrence J67 176 (fig. 24), although the illustration to Psalm XXXIII in the Utrecht J67 177 Psalter and its early eleventh-century Anglo-Saxon copy, Harley J67 178 603, provide precedents for the iconography of the lower scene (his J67 179 martyrdom itself), the upper scene is unique in the insular milieu. J67 180 The inclusion of the diminutive, green, bestial demon at Decius' J67 181 shoulder, inciting him to evil, is particularly to be noted since J67 182 this seems to have been a favourite motif of the artist and its J67 183 'brother' appears in his representation of the Temptation of St J67 184 Martin. In tracing the pictorial ancestry of these demons to the J67 185 illustration to Psalm XXV in the Utrecht Psalter, and the figural J67 186 initial to Pusillus eram in the Corbie Psalter, we should J67 187 not overlook the possible influence of dramatic literary portrayals J67 188 of devils as very physical agents provocateurs such as J67 189 are found in the Old English poems Juliana and Andreas. J67 190 Similarly, in the case of the image of Stephen (fig. 23), whereas J67 191 in the contemporary Bury Psalter and the earlier Benedictional of J67 192 St <*_>AE-ligature<*/>thelwold we see the saint depicted at the J67 193 moment of his martyrdom according to a well established J67 194 iconographic tradition, here we are presented instead with an J67 195 iconic, frontal figure, holding palm, stole, open book and censer, J67 196 who fills his frame, and who invites comparison with other J67 197 monumental hieratic figures, ranging from the evangelists in J67 198 certain Insular gospel books to those in a late eleventh-century J67 199 copy of the New Testament from the area of Agen-Moissac. The fact J67 200 that this image and that of St Andrew are iconic, while the J67 201 miniatures for St Martin, St Lawrence and St Peter are narrative J67 202 (consisting of one, two and three scenes respectively) is J67 203 symptomatic of the eclectic qualities of this remarkable decorative J67 204 programme as a whole. J67 205 J68 1 <#FLOB:J68\>Humphrey Searle

J68 2 BY

J68 3 DAVID C. F. WRIGHT

J68 4 Many composers suffer neglect and some are, or become, J68 5 completely forgotten. Few composers, however, have been as badly J68 6 treated as Humphrey Searle.

J68 7 He was born in Oxford on 26th August, 1915, one of three sons J68 8 born to Humphrey Frederic, a civil servant, and Charlotte Mathilde J68 9 May who, although born in England, had no English blood. Her father J68 10 was Sir William Schlich, who came from Darmstadt in Germany, and J68 11 Lady Schlich was of French, Belgian and Italian descent. Humphrey's J68 12 paternal grandfather trained as an organist and was part of a J68 13 musical family that lived in Devon.

J68 14 At school, Humphrey had piano lessons but was never proficient J68 15 as a pianist, although, in later life, he could be persuaded to J68 16 play a Bach Prelude and Fugue at parties. What he did become, J68 17 however, among many other things, was a competent conductor with an J68 18 impeccable ear for meticulous sound that earned him respect as a J68 19 record reviewer, writer and commentator. In all his critical J68 20 writings he was always constructive; he was never harsh or scathing J68 21 - which is some indication of the fundamental goodness of his J68 22 character. In fact, some have said that this was one cause of the J68 23 neglect of his music, particularly from the 1970s onwards. He J68 24 championed many musicians and young composers to the extreme of his J68 25 unfailing generosity, and his own career suffered. He was reticent J68 26 and self-effacing - often to the point of being embarrassing. If J68 27 the present writer commented on one of the many touches of genius J68 28 in his work he would invariably reply by demonstrating some passage J68 29 in Bach, Beethoven or Liszt that he thought to be marvellous.

J68 30 Searle's real awakening to music came in 1928 when he went to J68 31 Winchester School, where he met Robert Irving and James Robertson. J68 32 Irving was to become musical director of the Sadler's Wells Opera J68 33 at the same time that Humphrey Searle was on its advisory panel in J68 34 the 1950s. The availability of a gramophone at school enabled the J68 35 three students to familiarize themselves with the classical J68 36 repertoire and with such modern works as were recorded. Searle J68 37 began to take harmony lessons with George Dyson, who was J68 38 "profoundly impressed" by him. Having won a J68 39 Classical scholarship to Oxford University, Searle went up in 1933 J68 40 and, while he was absorbing the appropriate material for his J68 41 degree, he pursued musical studies with Sydney Watson, the organist J68 42 of New College.

J68 43 If interest in music germinated in 1928, then it blossomed six J68 44 years later when Searle heard the first English performance of J68 45 Berg's Wozzeck under Sir Adrian Boult, broadcast in March, J68 46 1934. "It knocked me sideways" admitted Searle, J68 47 who, consequently, put his energies into finding out about the J68 48 serial style of composition as advocated by Arnold J68 49 Sch<*_>o-umlaut<*/>nberg, the leader of the Second Viennese School, J68 50 which included Webern and Berg, two of Sch<*_>o-umlaut<*/>nberg's J68 51 distinguished pupils. Today, there is no doubt about the eminence J68 52 of Berg's incredible score but it says a very great deal for Searle J68 53 that he recognized its greatness at once. Fifty years or so later, J68 54 musical opinion has realized that twelve-note music is an original J68 55 means of composition without the restrictions of traditionalism. In J68 56 the hands of brilliant composers such as Sch<*_>o-umlaut<*/>nberg, J68 57 Webern and Berg and, indeed Searle, this musical language has J68 58 proved itself to be wide-ranging and varied in expression and J68 59 emotion - yet embodying the strictest discipline, as is J68 60 demonstrated, for example, in the language of Bach.

J68 61 The history of music has produced many enigmas. One such is J68 62 vacillating fashion. A composer's work can be subjected to savage J68 63 hostility and, later, with a change in musical trends, it is J68 64 reassessed as acceptable and sometimes admired. Yet the composer, J68 65 if he is still living, cannot claim recompense for the injustice he J68 66 has suffered. There is some truth in Arthur Honegger's remark that J68 67 the only qualification for the possibility of being a great J68 68 composer is being dead.

J68 69 Searle was helped in his research by Theodor Adorno, who had J68 70 studied with both Sch<*_>o-umlaut<*/>nberg and Webern and had come J68 71 to Oxford as a refugee from Nazi Germany. We may not be able fully J68 72 to grasp what it meant in those days to espouse the cause of serial J68 73 music. It was, to say the least, unfashionable. After the war, J68 74 Searle was to make an equally courageous stand for the music of J68 75 Liszt, serving as secretary of the Liszt Society for twenty-two J68 76 years from 1950. At that time, Liszt was absurdly claimed to J68 77 "compose music of empty grand gesture", which, it J68 78 was alleged, stemmed from "the reckless extroversion of his J68 79 oscillating character". Humphrey Searle's heroic position J68 80 has, subsequently, been vindicated but, at the time, he stood alone J68 81 in this country and its moribund musical establishment, whose J68 82 members thought him to be mad and, furthermore, to be shunned at J68 83 all costs. It cannot be emphasized enough that he took the brunt of J68 84 acrimonious criticism of serial composition in this country. Later, J68 85 with the partial rebirth of the musical establishment, younger J68 86 composers gained some success with their twelve-note works and it J68 87 is to Humphrey Searle that they owe a considerable debt of J68 88 gratitude. Their respective attainments, however, brought no J68 89 recognition for Searle. He was still ignored and, while he did not J68 90 complain, inwardly he was hurt. He used to say that the music J68 91 business was a savage jungle and some music lovers are fickle, J68 92 subscribing to the immature notion that what cannot be whistled or J68 93 immediately understood is rubbish. This denotes a parochial J68 94 dependence on the security of predictability and obviousness to the J68 95 ennui that may accompany it.

J68 96 Professor of Music at Oxford, while Searle was there, was Sir J68 97 Hugh Allen, who saw the enormous potential in him and, J68 98 consequently, promised a travelling scholarship when he had taken J68 99 his Classics degree. It was not only Allen and Dyson who recognized J68 100 the gifts latent in this young man. William Walton saw some of J68 101 Searle's compositions and urged him to pursue his musical studies J68 102 without delay, recommending lessons with John Ireland at the Royal J68 103 College of Music. On his entering the College Searle's parents J68 104 withdrew their financial support since they were opposed to a J68 105 musical career, wanting their son to sit the Civil Service J68 106 examination. Searle was undeterred and made the first of many J68 107 stands showing the admirable courage and tenacity that never J68 108 deserted him.

J68 109 Searle studied with Ireland only for a short while and liked J68 110 him very much. As for Ireland, he always said of Humphrey Searle: J68 111 "He is the cleverest musician I have ever met". J68 112 That is some accolade, particularly when it is remembered that J68 113 Ireland met and knew many world-famous musicians - for example, he J68 114 used to recall encounters with Ravel.

J68 115 His allowance from his parents having ceased, Searle had to J68 116 earn his living, which partly came from teaching logic to a J68 117 clergyman in Mornington Crescent. He took his B.A. degree in 1937 J68 118 and, with the promised scholarship, went to Vienna that autumn to J68 119 study with Webern, the lessons having been arranged by Adorno. With J68 120 Webern, Searle developed his dependable ear, but his individuality J68 121 meant that he did not fall into the trap of emulating so J68 122 illustrious a teacher. "Imitation, I suppose, is a sincere J68 123 form of flattery" he told me, adding:

J68 124 I want to write what I want to write. Anyone can copy J68 125 someone else and dress it up in the mendacious disguise of J68 126 originality. Lesser composers have done this and some have made a J68 127 name for themselves in this mercenary way but it says absolutely J68 128 nothing for their integrity or the advancement of music.

J68 129 That originality is an essential ingredient in music will be, J68 130 it is hoped, readily admitted. This problem greatly troubeld Walton J68 131 in the late 1940s. He was undergoing a crisis of his own musical J68 132 identity and deploring the suggestion that he was the successor to J68 133 Elgar. Walton enjoyed "a few moments of Elgar" J68 134 whereas, at the other end of the spectrum, he considered J68 135 Shostakovich to be "the greatest composer of the twentieth J68 136 century". Sir Adrian Boult used to say how very J68 137 disappointing Elgar's music was and that the better it was played J68 138 the less he liked it - by which he meant that if Elgar's music were J68 139 badly played you could blame the orchestra but if it were played J68 140 well the only person who could take the blame was Elgar himself.

J68 141 Walton, being a discerning musician, went to Humphrey Searle J68 142 for advice and lessons, which is the highest compliment a composer J68 143 with many successes like Fa<*_>c-cedille<*/>ade, Belshazzar's J68 144 Feast and a truly memorable viola Concerto behind him could J68 145 pay to another. Yet that is the compliment Walton paid to Searle J68 146 and, as a consequence, Walton's later works have a greater J68 147 originality and an admirable texture, as is seen in the J68 148 Johannesburg Festival Overture, the masterly cello J68 149 Concerto and the Symphony no. 2, which continues to gain in J68 150 appeal.

J68 151 In Vienna during the years 1937-38 Searle attended the J68 152 Conservatoire and, among other things, took extensive tuition in J68 153 conducting. He visited the opera almost every night and these J68 154 months were among the happiest in his life. Such contentment was J68 155 further enhanced by the fact that Webern was a brilliant teacher. J68 156 Searle once told the present writer:

J68 157 It is still wrongly assumed in some quarters that J68 158 Webern was only interested in contemporary music. Nothing could be J68 159 further from the truth. It may surprise some to know that he did J68 160 not give me one lesson in serialism although he did suggest that I J68 161 analyse his Variations for piano in my private studies. His J68 162 lessons with me were on the properties of the triad, traditional J68 163 harmony and counterpoint. Webern was an outstanding conductor of J68 164 the classics. The myth that modern composers are narrow-minded J68 165 should be dispelled once and for all. It simply is not so. Yet it J68 166 is very sad that musicians today who strongly object to the J68 167 slightest hint of criticism of Bach, Beethoven or Mozart can, and J68 168 do, censure music of the twentieth-century.

J68 169 Searle returned to London just before Hitler's troops took J68 170 Vienna in March, 1938. Back at the Royal College of Music, he J68 171 continued lessons in orchestration, conducting and counterpoint; J68 172 among his teachers were R. O. Morris and Gordon Jacob. His money J68 173 having run out, he had to find work and became chorus librarian J68 174 with the B.B.C. "It was nothing more than being a J68 175 pack-horse carting music around London" he would say. That J68 176 was one of his typically over-modest remarks.

J68 177 In 1939 he conducted the London String Orchestra in one of his J68 178 now discarded pieces as well as in Liszt's Malediction, his J68 179 own arrangement of three pieces by Thomas Roseingrave and the first J68 180 British performance of Webern's Five Pieces, op. 5 - a J68 181 very brave choice indeed. The concert also included Searle's J68 182 arrangement of the Adagio cantando from Bernard van J68 183 Dieren's string Quartet no. 5. Searle had wanted to study with van J68 184 Dieren but the Dutch-born composer was far too ill and had, in J68 185 fact, died in 1936.

J68 186 When war broke out Humphrey Searle, who had by now written a J68 187 set of piano variations and a string quartet, moved with the B.B.C. J68 188 to Bristol and spent time in the stimulating company of such people J68 189 as William Glock, Lennox Berkeley, Arnold Cooke, the poets Dylan J68 190 Thomas and William Empson and the critic Henry Boys. In March, 1940 J68 191 Searle joined the Army and was with the Gloucestershire Regiment J68 192 for a short while. He was then transferred to the Intelligence J68 193 Corps and to a remote part of Scotland, which inspired his highly J68 194 engaging Highland Reel. It was in Scotland that he J68 195 composed the first work of his to bear an opus number - the J68 196 Suite no. 1 for string orchestra, which he acknowledged to be the J68 197 first work of his maturity. It was first perfomed in London in 1943 J68 198 and conducted by Walter Goehr. The composer admitted that the work J68 199 owes something to Bart<*_>o-acute<*/>k. This was followed by J68 200 Night Music, op. 2, for chamber orchestra, written in J68 201 honour of Webern's 60th birthday, a work of great character and J68 202 lucidity. The Vigil for piano, op. 3, deliberately J68 203 recalls Satie's Gymnop<*_>e-acute<*/>dies, for it was J68 204 dedicated to the French Fighting Forces. J68 205 J69 1 <#FLOB:J69\>As with the decoder, the multiplexer circuit is not J69 2 inherently hazard free, but, with a maximum of one inverter signal J69 3 path difference, no glitches should be seen and the advantages of J69 4 flexibility and design speed still apply.

J69 5 For a logic system where many outputs are formed from the same J69 6 set of input signals, a decoder-based system requires one OR gate J69 7 for each output and one decoder to supply all the basic addresses. J69 8 With a multiplexer circuit, a separate MUX is needed for each J69 9 function.

J69 10 3.6 PARITY

J69 11 The parity of a binary word is defined as even if the J69 12 number of bits set to 1 is even, and odd if the number of bits J69 13 set to 1 is odd. A parity bit is a bit added to a binary word to J69 14 make the parity of the word plus the bit the same for all words in J69 15 the data set. Parity bits are added to data before storage or J69 16 transmission so that the integrity of the data can be checked and J69 17 corrected on retrieval or reception; if the parity has changed then J69 18 an error has occurred. For systems which transfer or store data as J69 19 parallel words it is normal to add a parity bit which makes the J69 20 combined word odd parity as this ensures that at least one bit is J69 21 1, and it is easy to detect the difference between a data value of J69 22 0 and no data. For serial transmission or storage, start and stop J69 23 bits are used to delimit a data word, and so either parity type can J69 24 be employed. Table 3.1 lists the parity of binary numbers from 0 to J69 25 9 and the codes for the numbers plus parity bits to produce odd and J69 26 even parity words.

J69 27 Most systems which use parity assume that any errors are J69 28 single-bit errors and that they occur only rarely. In the simplest J69 29 scheme, one parity bit is added to each word, so that if an error J69 30 occurs (a 1 is changed to a 0, or a 0 to a 1), the parity of the J69 31 word is changed and this can be detected; no correction is possible J69 32 in this case. To allow the correction of single errors and the J69 33 detection of multiple errors, several parity bits are added to each J69 34 word where each parity bit checks the parity of part of the word; J69 35 such a scheme is known as a Hamming code.

J69 36 table&figure&captions

J69 37 Where error rates are expected to be very low, a simple, J69 38 single-error correction method is possible for a block of data J69 39 containing m words of n bits. One parity bit is generated J69 40 for each of the m words, so that an error indicates which word J69 41 is faulty, and one parity bit is generated for each bit position of J69 42 the n bit words, so that an error indicates which bit position J69 43 is faulty. An error is therefore pinpointed and may be corrected by J69 44 inverting the faulty bit.

J69 45 A circuit to measure or check the parity of an eight-bit word J69 46 is shown in Fig. 3.11; the output is 1 if the parity of the input J69 47 word is odd and 0 if the parity is even. The output of an J69 48 XOR gate is 1 if the inputs are different, i.e. odd parity. If the J69 49 outputs of two XOR gates are used as inputs to a third XOR, then an J69 50 odd parity input produces a 1 at the output and an even parity J69 51 input results in a 0 output. This XOR tree structure may be J69 52 expanded (or reduced) to accommodate any number of bits. As a J69 53 parity generating circuit, the combined output of the eight-bit J69 54 input word and the parity checking circuit is a nine-bit even J69 55 parity word. To obtain odd parity for the combined word the J69 56 output of the parity checking circuit should be inverted.

J69 57 3.7 MEMORIES AS LOGIC ELEMENTS

J69 58 A memory consists of an array of elements in which single bits J69 59 of digital information can be stored. To access a particular J69 60 element, the position (or address) of that element is required, and J69 61 then data may be written into or read from it. (Memories are J69 62 discussed in more detail in Chapter 6.) Each row of a truth table J69 63 has a unique address given by the state of the input signals. If, J69 64 therefore, the input signals are connected to the address lines of J69 65 the memory, and the memory elements already contain the truth table J69 66 for the function to be generated, reading the memory at the J69 67 selected address results in an output of the required state. For J69 68 the table in Fig. 3.9, addresses 0, 3, 4 and 5 must contain 0 and J69 69 addresses 1, 2, 6 and 7 must contain 1 for the memory to produce J69 70 the correct function.

J69 71 To change the logic function performed by the memory it is only J69 72 necessary to change the contents of the memory. As no J69 73 interconnection changes are required to do this a memory seems J69 74 ideal in providing a flexible logic system, but a number of J69 75 problems arise in trying to provide this flexibility. If the memory J69 76 is volatile, which means that when the power is switched off the J69 77 information in the memory is lost, the data must be reloaded into J69 78 the memory every time the system is switched on. This is tedious by J69 79 hand and complex by computer. Non-volatile memories are either J69 80 permanent (read-only memories or ROMs), in which case J69 81 they must be replaced to change the logic function, or relatively J69 82 slow (erasable programmable read-only memories or J69 83 EPROMs). In most cases non-volatile devices require special J69 84 circuits or systems to alter the data stored in them.

J69 85 Memories have significant advantages over the devices described J69 86 earlier in systems with a large number of inputs or where the logic J69 87 must be adaptable during use. The size of modern memories means J69 88 that single-bit logic functions of 16 to 20 signals can be produced J69 89 by a single chip with access times in the range 25 to 150 ns and J69 90 eight-bit output functions of 12 to 14 inputs in similar times. As J69 91 such functions correspond to truth tables with between 4096 (12 J69 92 bits) and approximately one million (20 bits) rows, construction of J69 93 any but the most trivial with basic gates, decoders or multiplexers J69 94 could not be contemplated because of the vast number of chips they J69 95 would need. The ability to change the logic function can be a J69 96 requirement of the system and, in that case, there is no J69 97 alternative but to use a device that can be reprogrammed, e.g. a J69 98 memory. An example of this is where, if one input signal stops J69 99 working, it may be necessary to stop the system unless the logic J69 100 can be reconfigured to allow for that signal failure.

J69 101 3.8 PROGRAMMABLE LOGIC

J69 102 The increasing complexity of integrated circuits has allowed J69 103 manufacturers to produce circuits which can be programmed to J69 104 perform random logic functions. By using these devices to replace J69 105 basic gates, the number of integrated circuits used to produce the J69 106 random logic can be reduced by about a factor of ten. A number of J69 107 designs exist for these devices with different names for different J69 108 designs. Most names are a selection from the list: array; device; J69 109 gate; logic; programmable. Although some devices are capable of J69 110 performing sequential as well as combinational logic, this section J69 111 will concentrate on programmable devices for combinational J69 112 logic.

J69 113 To make the device programmable, the manufacturer includes J69 114 fusible links into the signal paths. If the link is left intact, J69 115 then the signal is transmitted from one stage to the next; if the J69 116 fuse is blown, then the signal path is broken. To program one of J69 117 these devices it is necessary to retain all the links on the signal J69 118 paths required and to blow all the other fuses. Special equipment J69 119 is needed to provide the correct waveforms to blow the fuses J69 120 without destroying the device.

J69 121 In Chapter 2 it was seen that combinational logic functions J69 122 could be produced by generating only the terms corresponding to J69 123 ones in the truth table (or by generating the zero terms and J69 124 inverting the output). As described above, a memory generates an J69 125 output (0 or 1) for every row of the truth table and, therefore, at J69 126 least half the memory is being used inefficiently. For the example J69 127 based on Fig. 3.9, a circuit which outputs a one for addresses 1, J69 128 2, 6 and 7 is all that is needed since the other input combinations J69 129 will then automatically generate a zero. For sparse truth tables J69 130 (those with only a few rows set to one logic state and the vast J69 131 majority set to the other state), or tables with many don't J69 132 care/can't happen states, a memory is very inefficient. This is not J69 133 necessarily a problem for circuits made from individual chips but, J69 134 for the design of very large-scale integration (VLSI) devices, J69 135 where all the logic for a system has to fit on to one chip, it is J69 136 important to use space effectively.

J69 137 Basically, programmable logic implements a sum-of-products form J69 138 for the logic function or functions required and the programming of J69 139 these devices can be considered as logic minimization on a grand J69 140 scale. The most general design is the programmable logic J69 141 array (PLA) where both the connections to the AND gates, to J69 142 form the product terms, and the AND gate to OR gate connections, to J69 143 form the sum of products, contain fuses and can, therefore, be J69 144 programmed. Note that a programmable read-only memory J69 145 (PROM) fully decodes all inputs and therefore contains a fixed AND J69 146 array (the input decoder). The data in a PROM is stored by J69 147 programming the OR array.

J69 148 In a PLA each input signal is buffered and connected to every J69 149 AND gate in both normal and inverted form through a fusible link J69 150 (Fig. 3.12). For each AND gate, fuses are blown so that it receives J69 151 only the signals for one of the product terms in the minimized J69 152 expression. Each AND gate output is connected to every OR gate J69 153 through a link so that the product terms to be summed by an OR gate J69 154 can be selected. The internal design of the chip ensures that the J69 155 input to a gate with the fuse blown is set to the appropriate logic J69 156 level; high for an AND gate and low for an OR gate input.

J69 157 For the example given, the top AND gate remains connected only J69 158 to A, <*_>unch<*/> and <*_>unch<*/>; the second to J69 159 <*_>unch<*/> and B; the third to B and C; the bottom J69 160 gate is connected only to <*_>unch<*/> and C. The top OR gate J69 161 sums the top three product terms and the lower gate sums the bottom J69 162 two terms. Note that the term B<*_>dot<*/>C is common to J69 163 both OR gates. The OR gate output is connected to the output pin of J69 164 the PLA through an XOR gate so that the output can be produced in J69 165 normal or complemented form. This circuit would occupy only a small J69 166 fraction of a standard PLA compared with three packages if simple J69 167 gates were used.

J69 168 A typical field programmable device (one that can be J69 169 programmed by the user) has 16 inputs, forms 48 product terms and J69 170 produces eight outputs. It becomes impossible to draw a circuit J69 171 diagram in the way used up to now as the device contains 32 J69 172 inverters, 48 32-in AND gates, eight 48-in OR gates and eight XOR J69 173 gates. A shorthand method is needed to show which fuses are to J69 174 remain intact. A common method of doing this is shown in the lower J69 175 half of Fig. 3.12. One line is used to represent all the (separate) J69 176 inputs to a gate and crosses are used to show the connections to J69 177 those inputs (i.e. a cross marks where a fuse is to be left J69 178 intact). It is emphasized that there is no direct connection J69 179 between the signals marked with a cross on that line. The AND gate J69 180 inputs line can be termed a product line and the OR gate J69 181 inputs line a sum line. To keep the diagram as simple as J69 182 possible, the input buffers/inverters are not shown. The XOR gates J69 183 on the outputs are shown explicitly as they must be programmed to J69 184 give a normal or complemented output.

J69 185 J70 1 <#FLOB:J70\>Another difference is that the analogue plan allows a J70 2 small number of 2-wire links to be used in the connection.

J70 3 13.6 THE CONTROL OF SIDETONE

J70 4 'Sidetone' is the name given to an effect on 2-wire systems J70 5 where the speaker hears his own voice in his own earphone while he J70 6 is speaking. (We first introduced it in Chapter 2.) Too little J70 7 sidetone can make speakers think their telephone is dead, while too J70 8 much leads them to lower their voices. CCITT recommends 'sidetone J70 9 reference equivalents' of at least 17 dB.

J70 10 13.7 THE CONTROL OF ECHO

J70 11 By ensuring that the CCITT recommended loss of 6 to 7 dB at J70 12 least is encountered between the two reference points at either end J70 13 of the 4-wire part of a connection, we not only safeguard the J70 14 circuit against the effects of instability; we also make it J70 15 unlikely that any signal echo is generated by the 2-to-4-wire J70 16 converter (the so-called 'hybrid' converter). These echoes are J70 17 caused by reflection of the speaker's voice back from the distant J70 18 receiving end, and Figure 13.8 shows a 'hybrid' converter causing a J70 19 signal echo of this kind. The problem of echo arises whenever a J70 20 hybrid converter is used, and it results from the non-ideal J70 21 electrical performance (i.e. the 'mismatch') of the device. No J70 22 matter whether the 4-wire line is digital or analogue, the result J70 23 is an echo.

J70 24 To understand the cause of the echo we have first to consider J70 25 composition of the hybrid itself. It consists of a 'bridge' of four J70 26 pairs of wires; one pair correspond to the 2-wire circuit, two more J70 27 pairs correspond to the 'receiver' and 'transmit' pairs of the J70 28 4-wire circuit, and the final pair is a balance circuit, the J70 29 function of which is explained below. The wires are connected to J70 30 the 'bridge' in such a way as to create a separation of the receive J70 31 and transmit signals on the 2-wire circuit from or on to the J70 32 corresponding receive and transmit pairs of the 4-wire circuit. It J70 33 is easiest to understand the type of hybrid which uses a pair of J70 34 'cross-coupled' transformers as a bridge. This is shown in Figure J70 35 13.9.

J70 36 Any signal generated at the telephone in Figure 13.9 appears on J70 37 the transmit pair (but not on the receive pair), while any incoming J70 38 signal on the receive pair appears on the telephone (but not on the J70 39 transmit pair). It works as follows.

J70 40 figures&captions

J70 41 Electrical signal output from the telephone produces equal J70 42 magnetic fields round windings W1 and W2. Now J70 43 the resistance in the balance circuit is set up to be equal to that J70 44 of the telephone so as to induce equal fields around windings J70 45 W3 and W4, but the cross-coupling gives them J70 46 opposite polarity. The fields of windings W1 and J70 47 W3 tend to act together and to induce an output in J70 48 winding W5. Conversely, the fields of windings J70 49 W2 and W4 cancel one another (due to the J70 50 cross-coupling of W4), with the result that no output is J70 51 induced in winding W6. This gives an output on the J70 52 transmit pair as desired, but not on the receive pair.

J70 53 In the receive direction, the field around winding W6 J70 54 induces fields in W2 and W4. This produces J70 55 cancelling fields in windings W1 and W3 because J70 56 of the cross-coupling of windings W3 and W4. An J70 57 output signal is induced in the 2-wire telephone circuit but not in J70 58 the 'transmit' pair, winding W5.

J70 59 Unfortunately, the balance resistance of practical networks is J70 60 rarely matched to the resistance of the telephone. For one thing, J70 61 this is because the tolerance on workmanship in real networks is J70 62 much greater. Also, the use of exchanges in the 2-wire part of the J70 63 circuit (if relevant) means that it is impossible to match the J70 64 balance resistance to the resistance of all the individual J70 65 telephones to which the hybrid may be connected. The fields in the J70 66 windings therefore do not always cancel out entirely as intended. J70 67 So for example, when receiving a signal via the receive pair and J70 68 winding W6, the fields produced in windings W1 J70 69 and W3 may not quite cancel, and a small electric current J70 70 may be induced in winding W5. This manifests itself to J70 71 the speaker as an echo. The strength of the echo is usually denoted J70 72 in terms of its decibel rating relative to the incoming signal. J70 73 This is a value called the 'balance return loss', or sometimes, the J70 74 'echo return loss'. The more efficient the hybrid, the greater the J70 75 balance return loss (the isolation between receive and transmit J70 76 circuits). A variety of other problems can be caused by echo, the J70 77 two most important of which are:

J70 78 <*_>bullet<*/>electrical circuit instability (and possible J70 79 'feedback')

J70 80 <*_>bullet<*/>talker distraction.

J70 81 If the returned echo is nearly equal in volume to that of the J70 82 original signal, and if a rebounding 'echo' effect is taking place J70 83 at both ends of the connection, then the volume of the signal can J70 84 increase with each successive echo, leading to distortion and J70 85 circuit overload. This is circuit instability, and as we already J70 86 know, the chance of it occurring is considerably reduced by J70 87 adjusting the 4-wire circuit to include more signal attenuation. J70 88 The total round-loop loss in the UK digital network shown in Figure J70 89 13.7 is at least 14 dB, probably inflicting at least 30 dB J70 90 attenuation on echoes even if the hybrid has only a modest J70 91 isolating performance. Talker distraction (or data corruption) is J70 92 another effect of echo, but if the time delay of the echo is not J70 93 too long, then distraction is unlikely, because all talkers hear J70 94 their own voices anyway while they are talking.

J70 95 The echo delay time is equal to the time taken for propagation J70 96 over the transmission link and back again, and is thus related to J70 97 the length of the line itself. The longer the line, the greater the J70 98 delay. Should the one-way signal propagation time exceed around 8 J70 99 ms, giving an echo delay of 15 ms or more, then corrective action J70 100 is necessary to eliminate the echo which most telephone users find J70 101 obtrusive. A one-way propagation time of 8 ms is inevitable on all J70 102 long lines over 2500 km, so that undersea cables of this length and J70 103 all satellite circuits usually require echo suppression. Further J70 104 propagation delay can also be caused by certain types of switching J70 105 and transmission equipment. Indeed, a significant problem J70 106 encountered with digital transmission media is that the time J70 107 required for intermediate signal regeneration (detection and J70 108 waveform reshaping) means that the overall speed of propagation is J70 109 actually reduced to only 0.6 times the speed of light. This means J70 110 that even quite short digital lines require echo suppression. Two J70 111 methods of controlling echoes on long-distance transmission links J70 112 are common. These are termed 'echo suppression' and 'echo J70 113 cancellation'.

J70 114 An echo suppressor is a device inserted into the transmit path J70 115 of a circuit. It acts to suppress retransmission of incoming J70 116 'receive path' signals by inserting a very large attenuation into J70 117 the transmit path whenever a signal is detected in the receive J70 118 path. Figure 13.10 illustrates the principle.

J70 119 The device in Figure 13.10 is called a 'half-echo suppressor' J70 120 since it acts to suppress only the transmit path. A full echo J70 121 suppressor would suppress echoes in both transmit and receive J70 122 paths.

J70 123 It is normal for a long connection to be equipped with two J70 124 half-echo suppressors, one at each end, the actual position being J70 125 specified by the formal transmission plan. Ideally half-echo J70 126 suppressors should be located as near to the source of the echo as J70 127 possible (i.e. as near to the 2-to-4-wire conversion point as J70 128 possible), and work best when near the ends of the 4-wire part of J70 129 the connection. In practice it may not be economic to provide echo J70 130 suppressors at all exchanges in the lower levels of the hierarchy, J70 131 and so they are most commonly provided on the long lines which J70 132 terminate at regional (class 1 of Figure 13.6) and international J70 133 exchanges.

J70 134 Sophisticated inner-exchange signalling is used to control the J70 135 use of half-echo suppressors. Such signalling ensures that on J70 136 tandem connections of long-haul links intermediate echo suppressors J70 137 are 'turned off' in the manner illustrated by Figure 13.11. This J70 138 ensures that a maximum of two half-echo suppressors (one at each J70 139 end of the 4-wire part of the connection) are active at any one J70 140 time.

J70 141 figures&captions

J70 142 The amount of suppression (i.e. attenuation) required to reduce J70 143 the subjective disturbance of echo depends upon the number of echo J70 144 paths available, the echo path propagation time, and upon the J70 145 tolerance of the telephone users (or data terminal devices). CCITT J70 146 recommends that echo suppression should exceed (15 + n) dB, J70 147 where n is the number of links on the connection.

J70 148 Unfortunately echo suppressors cannot be used in circuits J70 149 carrying data, because the switching time between attenuation-on J70 150 and attenuation-off states is too slow and can itself cause loss or J70 151 corruption of data. Most data modems designed for use on telephone J70 152 circuits are therefore programmed to send an initiating 2100 Hz J70 153 tone over the circuit, in order to disable the echo suppressors.

J70 154 Another form of echo control device, called an 'echo J70 155 canceller', can be used on either voice or data circuits. Like an J70 156 echo suppressor, an echo canceller has a signal detector unit in J70 157 the receive path. However, instead of using it to switch on a large J70 158 attenuator, it predicts the likely echo signal and literally J70 159 subtracts this prediction from the transmit signal, thereby largely J70 160 'cancelling' out the real echo signal. Other signals in the J70 161 transmit path should be unaffected. Listeners rate the performance J70 162 of echo cancellers to be better than that of echo suppressors. J70 163 This, coupled with the fact that they do not corrupt data signals, J70 164 is making them a popular alternative to suppressors, common on J70 165 digital line systems and exchanges, and standard equipment in some J70 166 countries (e.g. USA).

J70 167 13.8 SIGNAL (OR 'PROPAGATION') DELAY

J70 168 An important consideration of the network transmission plan is J70 169 the overall signal delay or 'propagation time'. Excessive delay J70 170 brings with it not only the risk of echo, but also a number of J70 171 other impairments. In conversation, for example, long propagation J70 172 times between talker and listener can lead to confusion. In the J70 173 course of a conversation, when we have said what we want to say, we J70 174 expect a fairly prompt response. If we are met with a silent pause, J70 175 caused by a propagation delay, then we may well be tempted to speak J70 176 again, to check that we have been heard. Inevitably, as soon as we J70 177 do that, the other party starts speaking, and everyone is talking J70 178 at once.

J70 179 On video the effect of signal delay is even more revealing. For J70 180 example, on live satellite television broadcasts, whoever is at the J70 181 far end always gives the impression of pausing unduly before J70 182 answering any question.

J70 183 Nothing can be done to reduce the delay incurred on a physical J70 184 cable or satellite transmission link. Thus intercontinental J70 185 telephone conversations via satellite are bound to experience a J70 186 one-way propagation delay of about 1/4 second, giving a pause J70 187 between talking and response of 1/2 second. Furthermore, the J70 188 extremely rapid bit speeds and response times that computer and J70 189 data circuitry is capable of can be affected by line lengths of J70 190 only a few centimetres or metres. Line lengths should therefore be J70 191 minimized and circuitous routings avoided as far as possible. It is J70 192 common for maximum physical line lengths to be quoted for data J70 193 networks. Similarly, in telephone networks, rigid guidelines demand J70 194 that double or treble satellite hops or other excessive delay paths J70 195 (i.e. those of 400 ms one-way propagation time or longer) are J70 196 avoided whenever possible. Excessive delays can be kept in check by J70 197 appropriate network routing algorithms, as we shall see in Chapter J70 198 14.

J70 199 13.9 NOISE AND CROSSTALK

J70 200 Noise and crosstalk are unwanted signals induced on to the J70 201 transmission system by adjacent power lines, electrostatic J70 202 interference, or other telecommunication lines. The only reliable J70 203 way of controlling them is by careful initial planning and design J70 204 of the transmission system and the route. One source of noise J70 205 results from the induction of signals on to telecommunications J70 206 cables which pass too close to high-power lines. Another source of J70 207 noise is poorly soldered connections or component failures. J70 208 J71 1 <#FLOB:J71\>The deformation efficiency concept is J71 2 attractive and has demonstrably led to a better understanding of J71 3 the important parameters to be developed to give a meaningful J71 4 data-base for any given material of interest. Lyapunov J71 5 function stability criteria seem to provide powerful tools for J71 6 delineating stable and unstable regions on these mappings of J71 7 deformation behaviour. Apparently, these regions so discovered do J71 8 actually correspond to material instabilities well-known to J71 9 materials scientists dealing with both metals and non-metals. One J71 10 remaining problem, however, is that of adequately accounting for J71 11 the history of the deformation.

J71 12 Modelling of dynamic material behaviour in hot deformation can J71 13 most easily be included by using this mapping approach. The J71 14 data-base so developed could then be used to provide data as a J71 15 function of temperature, strain, strain rate and time (to include J71 16 the history of deformation ) which can then be fed into existing J71 17 comprehensive computer solutions for rolling, for example.

J71 18 However, many computer solutions are of the slab type (see J71 19 later discussion of plastic flow) and do not include either J71 20 temperature or strain rate effects. Much work has been done in J71 21 industry on this problem, but it has not been published widely.

J71 22 5.4 PHYSICAL PROPERTIES OF IMPORTANCE IN MANUFACTURE AND J71 23 USE

J71 24 Analysis of processes such as hot forming, casting or others J71 25 used at high temperatures requires a knowledge of thermal J71 26 conductivity <*_>kappa<*/>, specific heat J71 27 c, and thermal expansion <*_>alpha<*/>.

J71 28 The last is especially important in resistance to thermal J71 29 shock. A material such as silicon-aluminium-oxy-nitride has a much J71 30 lower thermal expansion coefficient than, say, aluminium oxide and J71 31 it is much more resistant to cracking when suddenly heated or J71 32 cooled. A high thermal conductivity is also valuable in reducing J71 33 any large temperature gradients which might otherwise J71 34 occur in a material or composite.

J71 35 When composites are used it is necessary to consider the J71 36 compatibility of expansion coefficients, otherwise high stresses J71 37 and even plastic deformation may occur.

J71 38 5.5 CHEMICAL PROPERTIES

J71 39 Oxidation is clearly important in hot working operations J71 40 although in an enclosed process like extrusion, for example, there J71 41 is very little access of air, so oxidation is substantially J71 42 avoided.

J71 43 The book edited by Braithwaite explains how chemical J71 44 reaction with lubricants plays a major role in the J71 45 effectiveness of lubricants. Titanium, for example, is very J71 46 difficult to lubricate because of its thin inert oxide skin. Schey J71 47 points out that stainless steels present the same problem.

J71 48 Corrosion is also a critical feature. A high proportion of J71 49 all metallic scrap arises from rusting and related J71 50 electrochemical degradation processes.

J71 51 Much attention has been given to plating and other J71 52 protective coatings to avoid corrosion and also to J71 53 improve other surface properties such as hardness and wear J71 54 resistance (see the handbooks by Swann, Ford and Westwood and J71 55 also by Peterson and Winer.

J71 56 5.6 WORKABILITY

J71 57 Although industrial producers will readily recognize some J71 58 alloys as being easier to work than others, there is no simple test J71 59 or clearly defined set of properties.

J71 60 Ductility is a related property, and it is generally J71 61 considered that a greater reduction of area to failure in a tensile J71 62 test indicates better workability. While this is true in general J71 63 terms, the actual fracture in a tensile test depends upon the J71 64 complex stress state in the neck region for all but the most J71 65 brittle alloys.

J71 66 In the comprehensive book by Atkins and Mai, various criteria J71 67 are mentioned which have often been proposed on the supposition J71 68 that fracture in tension depends upon the total plastic work or the J71 69 tensile plastic work expended in the deformation. It is very J71 70 difficult to obtain accurate experimental data because the final J71 71 stress state before fracture is frequently ill-defined. Recent J71 72 analytical work using finite-element plasticity has given J71 73 a better insight into these problems.

J71 74 Hot workability is often described in terms of the J71 75 number of revolutions to failure in a torsion test, but again the J71 76 analytical significance is obscure. Torsion testing is also a J71 77 valuable indicator of cold workability, but the results J71 78 depend strongly upon the axial constraint, possibly because J71 79 fracture and rewelding occur during the test, so the conditions J71 80 must be carefully controlled.

J71 81 5.7 ELEMENTS OF THE THEORY OF PLASTICITY

J71 82 5.7.1 General Theory

J71 83 To understand in more detail the mechanics of the forming J71 84 processes which are being discussed, it is necessary to understand J71 85 the fundamentals underlying the theory of the plastic flow of J71 86 materials. This is the one unifying theme that underlies all J71 87 forming processes. A complete treatment of the mathematical J71 88 theory of plasticity is beyond the scope or intent of this J71 89 book; comprehensive early texts have been written by Nadai, Hill, J71 90 and Prager and Hodge. Later treatments of the theory of plasticity, J71 91 discussed specifically with the needs of engineers in mind, are J71 92 given by Ford with Alexander, Johnson and Mellor, and Alexander. A J71 93 simpler, more practical account is given by Rowe. What it is J71 94 intended to do here is to consider the basic fundamentals on which J71 95 the theory rests and thereby to show how a better understanding of J71 96 the behaviour of materials subjected to these processes can be J71 97 achieved.

J71 98 In the first place the question may be asked: 'What is the main J71 99 difference between the behaviour of metal when subjected to large J71 100 deformation in metal-working processes and its behaviour when J71 101 stressed within the elastic range?' The simple answer is that in J71 102 metal-working processes the amount of straining or deformation is J71 103 many times larger. It is this essential difference between J71 104 metal-working and elastic straining which is all-important in J71 105 formulating a workable theory. The solution of a problem in J71 106 elasticity is often achieved by making the assumption that the J71 107 external shape is unaltered during elastic straining. That this is J71 108 justified for most metals can be realized when it is remembered J71 109 that their elastic range never exceeds about 0.4 per cent strain, J71 110 corresponding with a change in external dimensions of 0.4 per cent, J71 111 and is often only 0.1 per cent.

J71 112 By their very nature the processes of metal-working demand that J71 113 strains of more than one hundred times this value be imposed. J71 114 Indeed, in the case of forging or extrusion, reductions in J71 115 cross-sectional area of the order of 95 per cent are found. For J71 116 such large deformations a more precise definition of strain is J71 117 necessary than simply that of '(change in length)/(original J71 118 length)'. For example, if a tensile specimen is stretched to double J71 119 its length the engineer's strain would be unity, or 100 J71 120 per cent. To achieve 100 per cent compressive strain (engineer's J71 121 strain) it would obviously be necessary to compress the specimen J71 122 until it had zero thickness. Clearly, then, this measure of strain J71 123 is unsuitable for such large deformations, since 100 per cent J71 124 compression certainly represents much more straining than does 100 J71 125 per cent extension. Thus can be realized the first result of having J71 126 such large deformations, namely the necessity for specifying some J71 127 better measure of strain. Another consequence of having large J71 128 strains of this order is that, as far as determining the changes of J71 129 shape and the forces required is concerned, it will be permissible J71 130 to neglect elastic strains. This considerably simplifies the J71 131 theoretical approach, (but it should be remembered that elastic J71 132 deformation is very important when considering residual J71 133 stresses).

J71 134 Bearing in mind the large deformations which are imposed in J71 135 metal-working, another question can be asked: 'Why is it that, in a J71 136 tensile test, the metal will rarely withstand more than about 30 J71 137 per cent elongation without fracture, whereas in J71 138 metal-working processes much larger strains can be J71 139 imposed?' Consideration of the major processes reveals the answer J71 140 to this question - they are all predominantly compressive in J71 141 nature. Any stress system can be regarded as the sum of an J71 142 all-round hydrostatic stress (usually taken as being J71 143 equal to the mean stress) and stresses equal to the J71 144 differences between the actual stresses and this mean stress. It J71 145 was shown experimentally by Bridgman, and later again by Crossland, J71 146 that there is no plastic flow occasioned by this hydrostatic J71 147 stress. The plastic flow is caused by the deviatoric stress J71 148 system, so-called because the stresses take the values by which the J71 149 actual stresses 'deviate' from this mean stress. Although the mean J71 150 stress is not responsible for any plastic flow it has a profound J71 151 influence on fracture, and the more compressive its value the more J71 152 deformation can be imposed before failure occurs owing to fracture. J71 153 The exact dependence is not properly understood, but the tests of J71 154 Bridgman and Crossland have given us much information. A corollary J71 155 of the observation that the mean stress occasions no plastic flow J71 156 is that there can be no permanent change of volume. Thus, if J71 157 elastic strains are neglected, it may reasonably be assumed that J71 158 there is no change of volume at all during plastic J71 159 flow.

J71 160 Now, almost all metal-working processes take place under J71 161 conditions of complex stressing. In other words, if uniaxial J71 162 tension or compression is regarded as a simple system of stressing, J71 163 in which there is only one principal stress acting along the axial J71 164 direction of the prismatic specimen, then most J71 165 metal-working processes involve more complicated systems of J71 166 stress. Thus, in formulating a theory, a general three-dimensional J71 167 system of stress must be introduced and yet another question must J71 168 be asked: 'How can the behaviour of metal subjected to a complex J71 169 system of stress be correlated with its behaviour in simple tension J71 170 or compression?' To answer this question it is convenient to invoke J71 171 the concepts of equivalent stress (or strain) and J71 172 effective stress (or strain). These are parameters which J71 173 are functions of the imposed complex stresses or strains which J71 174 quantify their effectiveness in causing the plastic flow of J71 175 materials.

J71 176 Having discussed the principles on which the theory rests, it J71 177 is now possible to develop the equations of plastic flow. Before so J71 178 doing, it is well to recall the equations of three-dimensional J71 179 elasticity, to compare them with those for plastic flow. Initially, J71 180 what may be called the laws of elasticity and the J71 181 laws of plasticity can be set up, as follows:

J71 182 Laws of Elasticity

J71 183 Hooke's Law

J71 184 Equations of Equilibrium (of stresses and forces)

J71 185 Equations of Compatibility (of strains and displacements)

J71 186 Laws of Plasticity

J71 187 Stress-Strain Relations

J71 188 Equations of Equilibrium

J71 189 Equations of Compatibility

J71 190 Yield Criterion (function of stresses necessary to initiate and J71 191 maintain plastic flow)

J71 192 To these must be added the boundary conditions for J71 193 both stresses and displacements.

J71 194 Because of the magnitude of the strains and also the J71 195 non-linearity of the relationships involved in J71 196 metal-working processes it is necessary to formulate the equations J71 197 in terms of incremental strains or strain rates. This difficulty J71 198 does not arise for elastic straining, since the strains are always J71 199 sufficiently small for the total strains to be used. Thus, Hooke's J71 200 Law for a three-dimensional system of stress, referred to J71 201 x,y,z cartesian coordinates, is as follows:

J71 202 formulae

J71 203 In these equations, <*_>tau<*/>xy is the stress in J71 204 the y direction acting on a plane normal to the x J71 205 direction, <*_>gamma<*/>xy is the corresponding J71 206 engineering shear strain, E is Young's Modulus, J71 207 G is the shear modulus, <*_>nu<*/> is Poisson's J71 208 ratio, and formula. Equations (5.26) are the J71 209 well-known equations of elasticity representing Hooke's law. J71 210 Considering the first one of these equations, a positive (tensile) J71 211 stress <*_>sigma<*/>xx in the x direction produces a J71 212 strain of formula in that direction, whilst the positive J71 213 tensile stress <*_>sigma<*/>yy will produce a negative J71 214 strain (contraction) of formula in the x direction, J71 215 and <*_>sigma<*/>zz similarly produces a contraction. By J71 216 adding together the first three equations of equations (5.26) the J71 217 elastic volume change is determined as:

J71 218 formula

. J71 219 (The factor formula is three times the bulk J71 220 modulus K of the material, since formula is the J71 221 mean stress, or hydrostatic component of the stress system.)

J71 222 Thus, if the volume change is to be zero in plastic flow, J71 223 Poisson's ratio <*_>nu<*/> must be replaced by the factor 1/2 in J71 224 the stress-strain relations, since the right-hand side of equation J71 225 (5.27) then becomes zero. Young's Modulus E has no meaning for J71 226 plastic flow in which the elastic strains are neglected, and must J71 227 be replaced by an analogous parameter which will be considered J71 228 later.

J71 229 In any element of material subjected to a complex system of J71 230 stress there are three mutually perpendicular directions in which J71 231 the local direct stresses attain either maximum or minimum values. J71 232 J72 1 <#FLOB:J72\>The following issues should be incorporated into J72 2 all stages of the design, as an iterative process, to ensure that J72 3 an offshore system operates safely and efficiently.

J72 4 - workplace layout

J72 5 - console design

J72 6 - control and instrumentation design

J72 7 - VDU display design

J72 8 - manning levels and work organisation

J72 9 - communications

J72 10 - environmental issues

J72 11 - procedures

J72 12 - training

J72 13 - management and organisational issues

J72 14 - evacuation system design

J72 15 Workspace Layout and Console Design

J72 16 Ergonomics techniques and principles can be used to specify an J72 17 appropriate workplace layout, not only in relation to the platform J72 18 as a whole but also within the individual work areas. There are a J72 19 number of issues to be considered eg. the equipment requirements J72 20 within this workplace, necessary access space for both operation J72 21 and maintenance activities and the number of operators it is J72 22 envisaged will work in the area. The workplace layout must also be J72 23 considered with respect to viewing and communications requirements J72 24 in relation to other work areas.

J72 25 There exist a number of design recommendations based on data, J72 26 such as for example 560 mm clearance space must be allowed for body J72 27 access if the person is wearing normal clothing (obviously this J72 28 will be larger if protective clothing will be worn). There are in J72 29 addition, other considerations based on psychological factors eg. J72 30 performance has been shown to be better if a seated workstation is J72 31 positioned at right angles to the window, if appropriate, rather J72 32 than any other orientation. In Hardies<&|>sic! paper (4) he J72 33 discusses the changes in perceptions of the offshore accommodation J72 34 module in the 1980s. There was a move towards more pleasant J72 35 surroundings which led to hidden benefits in terms of increased J72 36 motivation, attention and a consequent reduction in accident rate. J72 37 It is therefore an important psychological consideration within the J72 38 design process.

J72 39 Anthropometric data (data on bodily dimensions 5) should be J72 40 used to design both control room areas and local work areas to J72 41 ensure that operator access to all instrumentation and other J72 42 equipment is efficient and comfortable, that displays and controls J72 43 most frequently used are within easy reach and to ensure ease of J72 44 maintenance. This latter is an important, often overlooked factor, J72 45 as a greater number of accidents occur per unit of time during J72 46 maintenance than during operations. An example of poor ergonomics J72 47 is where a valve requiring frequent maintenance is located out of J72 48 reach and thus necessitates the use of a ladder/scaffolding, or a J72 49 situation where access is restricted to a piece of equipment, the J72 50 maintenance of which demands the <}_><-|>used<+|>use<}/> of a large J72 51 or intricate tool.

J72 52 Anthropometric data should also be used to design the console J72 53 which houses displays and controls. This data will assist in the J72 54 decisions regarding the location of the instrumentation on the J72 55 panels to ensure that the most frequently used and important J72 56 displays and controls are within the primary reach envelope and J72 57 optimum viewing angle of the operators.

J72 58 The displays and controls used must be tailored to meet the J72 59 information needs of the operators. This will have implications for J72 60 the design and the layout of the instrumentation eg. if there is a J72 61 logical sequence of use of controls and displays the layout should J72 62 reflect this sequence and where possible the instruments should be J72 63 functionally grouped.

J72 64 Desgin of Instrumentation

J72 65 Task Analysis should be used to obtain the information on which J72 66 to base these decisions. This will serve to identify individual J72 67 elements of the task such as the following:

J72 68 - the goal of the particular task

J72 69 - the subtasks needed to achieve that goal

J72 70 - the information required in order for the operator to J72 71 complete that task

J72 72 - the action to be implemented

J72 73 - the feedback which is required by the operator to confirm J72 74 that the task has been successfully completed

J72 75 - the relevant performance shaping factors (NB. Performance J72 76 shaping factors are those factors which affect human performance J72 77 and may be of an external nature such <}_><-|>a<+|>as<}/> noise, J72 78 lighting etc. or of an internal nature such as stress, memory J72 79 overload etc.).

J72 80 This technique will be used to select a design which is J72 81 appropriate for the task eg. if the operator is required to compare J72 82 two tank levels, this is likely to be displayed in the form of two J72 83 vertical barcharts located next to each other. If a precise value J72 84 is required it may need to be displayed by means of a digital J72 85 indicator. If a simple on/off action is required during the task it J72 86 will be implemented by means of a simple switch or on/off J72 87 pushbuttons, whereas if continuous action is required, as in the J72 88 use of a hoist or crane, than a joystick would be more J72 89 appropriate.

J72 90 Ergonomics guidelines based on data and past experimentation J72 91 exist to assist the designer/ergonomist/engineer. These guidelines J72 92 also include advice about the use of colour, labelling, e.t.c. The J72 93 principles used to design VDU displays, if appropriate, are similar J72 94 to those for control and display design. It is useful to provide a J72 95 layout of information on a VDU screen which is compatible with that J72 96 on the hardwired control panels and which uses a consistent colour J72 97 coding scheme ie. colours which are ascribed specific meanings on J72 98 the hardwired panels should adopt the same meanings on the VDU J72 99 displays.

J72 100 Environmental Issues

J72 101 The design of the man machine interface alone however, will not J72 102 ensure safe operating performance. Although the ergonomics J72 103 techniques used to design and assess offshore systems are similar J72 104 to those used in any other industry, the performance constraints J72 105 and environmental conditions are very different.

J72 106 The levels of lighting will affect the visibility of the J72 107 instrumentation and equipment and if inadequate may lead to J72 108 hazardous situations. Noise levels which are too high are hazardous J72 109 to health and will affect communications.

J72 110 The thermal environment is known to have an J72 111 <}_><-|>affect<+|>effect<}/> upon performance (5) for further J72 112 details).

J72 113 Ergonomics guidelines exist to predict the length of time J72 114 people can be expected to work attentively at certain temperatures J72 115 and noise levels, where prolonged periods of exposure can lead to J72 116 permanent damage.

J72 117 Training and Procedures

J72 118 The ergonomist must not work in isolation to examine the design J72 119 issues relating for the workspace layout, the instrumentation and J72 120 consideration of the environment. 'Man at work' combines these J72 121 factors to produce a certain level and quality of performance. The J72 122 task analysis mentioned earlier will allow the analyst to examine J72 123 the interaction between them. The task analysis can be taken a J72 124 stage further and a human error analysis can be performed. Each J72 125 task step is examined to see if errors can potentially occur and J72 126 the nature of these errors. The consequences will be identified and J72 127 recovery points, if any, will be noted; for example if it is J72 128 imperative in a particular task step that a dial is read very J72 129 precisely, it has been identified that potentially the dial could J72 130 be misread or not read at all, there is no scope within the J72 131 procedure for the operator to recover his error and the J72 132 consequences lead to safety deficiencies, then a solution must J72 133 be found.

J72 134 The analyst must examine the reasons why the dial could be J72 135 misread; eg. the operator may be experiencing performance decrement J72 136 due to the stress induced by the situation. The solution may well J72 137 be to place a flashing lamp above the dial to attract the J72 138 operator's attention and make some attempt to reduce his stress J72 139 levels.

J72 140 The resulting information provides a valuable input to training J72 141 programmes and the writing or operational procedures as it enables J72 142 the assessor to identify the following:

J72 143 - those parts of the task which may be prone to error.

J72 144 - those tasks which may be performed infrequently but it is J72 145 important are performed correctly.

J72 146 - those tasks which require specialist skills.

J72 147 - any pieces of equipment which require detailed J72 148 instruction.

J72 149 - the points in the task which require communication with other J72 150 parts of the platform, and those communications which are critical J72 151 in terms of safe operation (this will be dealt with in more detail J72 152 when the case study is discussed later up the paper).

J72 153 The training programme will thus be designed to provide J72 154 intensive training, retraining and specialist training where J72 155 necessary and methods can be identified for monitoring training J72 156 effectiveness. Procedural documents can be written using ergonomics J72 157 guidelines to ensure completeness, adequacy and clarity and will be J72 158 able to provide the necessary observations, cautions, precautions J72 159 etc. to ensure effective and safer operation.

J72 160 In addition to training and procedures input, the task and J72 161 human error analysis will contribute greatly to the quantification J72 162 of risk by incorporating the human errors into fault trees. It can J72 163 be seen that the probability of a hazardous gas release remaining J72 164 undetected may be the result of an instrumentation failure but it J72 165 may also be the result of a human being failing to read, or failing J72 166 to read correctly, the instrumentation and thus failing to make the J72 167 correct diagnosis of the circumstances.

J72 168 Workload Assessment and Job Design

J72 169 Having discussed the ways in which design can be used to J72 170 enhance human performance, assisted by correct training and J72 171 procedures, it must also be mentioned that a well trained employee J72 172 in a well designed environment may still find himself in a J72 173 situation which he cannot deal with, for example if the operator is J72 174 under stress (the issue of stress as a performance shaping factor J72 175 was touched upon earlier). It is useful to perform a workload J72 176 assessment of the tasks the operator is expected to perform. This J72 177 may make use of different task analysis techniques such as J72 178 observation techniques, walk/talk through exercises, simulation J72 179 etc. to determine the tasks to be performed and the time taken to J72 180 perform them. They can then be plotted on a time line graph. If the J72 181 time line shows the operator to be overloaded at certain points in J72 182 the shift there are certain manipulations which can be completed to J72 183 optimise performance.

J72 184 The following are examples of corrective actions which can be J72 185 taken:

J72 186 - the reallocation of some of the tasks within the shift so the J72 187 operator does not have periods of intense activity or boredom.

J72 188 - the reallocation of tasks between the workers.

J72 189 -the provision of additional workers on the platform or for J72 190 that particular work area.

J72 191 - the reallocation of tasks between men and machines ie. J72 192 automate parts of the system.

J72 193 Thus this activity will not only provide useful information J72 194 about the adequacy of manning levels but also about the J72 195 organisation and management of the work.

J72 196 Management and Organisational Issues

J72 197 It has been demonstrated by reference to the Piper Alpha J72 198 incident that a system must not only be designed efficiently but J72 199 must operate efficiently for it to be productive and the management J72 200 and organisational machinery must also be well oiled. In order for J72 201 the workers to operate in safe manner and assign safety the highest J72 202 priority, the management and organisation of the company must J72 203 create and maintain the environment in which these safe attitudes J72 204 will be reinforced. Management must demonstrate their commitment to J72 205 safety. The following are examples of organisational mechanisms J72 206 which will assist in the creation of a safety culture:

J72 207 - a no-blame method of incident reporting with consequent J72 208 remedial actions to prevent future accidents.

J72 209 - communications mechanisms to ensure that safety concerns are J72 210 expressed both upwards and downwards through the hierarchy.

J72 211 - the practice of training needs assessment (and reassessment J72 212 when new equipment is introduced), to ensure the worker is aware of J72 213 all job related safety concerns, and the monitoring of training J72 214 effectiveness.

J72 215 - an effective permit to work system.

J72 216 - auditing systems to ensure the above systems are operating as J72 217 they should.

J72 218 There are many more. Human Factors techniques should be used to J72 219 tailor the specific safety management requirements of each offshore J72 220 system.

J72 221 Evacuation System Design

J72 222 The issues the paper has discussed in relation to ergonomics J72 223 methods so far have been with respect to work activity on the J72 224 platform. The consideration of the human and its behaviour within J72 225 the surroundings are of prime importance when evacuating the work J72 226 place. Traditionally the consideration of the human element in the J72 227 design of evacuation systems has been confined to design aspects J72 228 such as the width of evacuation passageways to ensure that an J72 229 adequate number of people can pass and that the correct number and J72 230 location of lifeboats has been provided to evacuate all J72 231 employees.

J72 232 It is necessary to design the system in this way but equal J72 233 weight must be given to the human behaviourial aspects of J72 234 evacuation. J72 235 J73 1 <#FLOB:J73\>As an example, two prototype gates both 15 m wide and J73 2 about 15.5 m high, for a head of about 15.2 m, weighed 93.8 and J73 3 117.7 t, respectively. The estimated weights given by the above J73 4 formulas are about 102 and 107 t, satisfactorily splitting the two J73 5 prototype weights.

J73 6 5.6 Rising sector gate

J73 7 This type of gate is illustrated in Fig. 5.2. The most famous J73 8 application is for the gates of the Thames barrier, where the J73 9 inherent stiffness of the design allowed a 60 m opening to be J73 10 adopted for navigation, while any failure of a gate to close before J73 11 a surge tide would not be catastrophic, the remaining gates J73 12 providing enough obstruction to flows to prevent flooding J73 13 upstream.

J73 14 This type of gate was not short-listed during the studies for J73 15 the Severn Barrage Committee (Ref. 1981(1)), largely because a good J73 16 seal is required between the gate and the sill to prevent loss of J73 17 water during the generating period, and this seal would be J73 18 difficult to maintain. In addition, there was concern that the gap J73 19 underneath the gate would trap sediments and lead to difficulties J73 20 in opening the gate. These do not apply to the Thames barrier J73 21 gates, where there is a 300 mm gap between the gates and their J73 22 sills. Loss of water through this gap is of no economic J73 23 consequence.

J73 24 A slightly modified version of this gate has been proposed for J73 25 use where the sluice is over the turbine (Ref. 1987(12)). By making J73 26 one of the longitudinal chambers watertight and buoyant, the forces J73 27 needed to operate the gate would be reduced. In addition, because J73 28 the gate is located well above the seabed, there would be little J73 29 risk of sediment building up in the gap underneath the gate.

J73 30 5.7 Gate operation

Lewin (Ref. 1980(16)) has estimated that the power required to J73 32 open a 12 m square vertical-lift wheeled gate in 15 min is about 60 J73 33 kW, i.e. about 15 kWh. A 20 m wide radial gate lifted 15 m will J73 34 require about 100 kW. Assuming that similar power would be used to J73 35 close the gates, which is pessimistic, these represent only around J73 36 0.02% of the output of a barrage.

J73 37 The simplest form of motive power for either vertical-lift J73 38 wheeled gates or radial gates are chains or ropes passing over J73 39 sprockets or drums. Each sprocket or drum is driven through a J73 40 reduction gearbox by a single high speed shaft, itself driven by an J73 41 hydraulic motor. The motive power for the hydraulic motors for a J73 42 number of gates could be provided by a single power source. J73 43 Alternatively, each gate could have its own power source, i.e. a J73 44 hydraulic pump driven by an electric motor.

J73 45 The use of oil-hydraulic motors instead of direct electric J73 46 drive has much to recommend it. Firstly, the hydraulic system is J73 47 sealed against the corrosive effects of a marine atmosphere. J73 48 Secondly, protection against overload, especially at the end of J73 49 travel, can be provided simply and reliably by pressure relief J73 50 valves instead of limit switches or other mechanisms which rely on J73 51 electrical contacts which can corrode. Thirdly, safety features J73 52 such as oil-immersed disc brakes, which cannot be released J73 53 until the system oil pressure has reached a certain minimum value, J73 54 can be readily incorporated into an oil-hydraulic J73 55 system.

J73 56 An oil-hydraulic system will be more expensive than an J73 57 electrical system, but, properly designed, will be far more J73 58 reliable and have a much longer working life.

J73 59 Each of the radial gates of the Haringvleit sluices is raised J73 60 and lowered by a pair of hydraulic rams, one at each end of the J73 61 gate, operating through a mechanical linkage. This removes problems J73 62 of corrosion of lifting chains or ropes ar the expense of requiring J73 63 a sophisticated system to measure and adjust the travel of each ram J73 64 so that the gate is not twisted.

J73 65 5.8 Conclusions

J73 66 The choice of type of sluice is, in practice, limited to the J73 67 types discussed above, and will depend on site-specification J73 68 factors as well as factors such as first cost. For sites with deep J73 69 water, say 20 m or so at mean tide, the vertical lift wheeled gate J73 70 appears to be the best choice. For shallower sites, especially J73 71 those on the routes of migratory fish, the radial gated sluice is J73 72 probably the best choice.

J73 73 Chapter 6

J73 74 Embankments and plain caissons

J73 75 6.1 Introduction

J73 76 Only in the exceptional circumstances of a narrow, steep-sided J73 77 site for a barrage would the 'working' components of a barrage, J73 78 namely the turbines in their power house, the sluices and the ship J73 79 lock, occupy the full width of the estuary. The remaining gaps J73 80 would have to be closed by embankments or plain, i.e. J73 81 non-working-caissons. This chapter considers both options. J73 82 Except where access from one bank of the estuary has to be provided J73 83 early in the construction programme, the non-working parts are the J73 84 least cost items and should be least sensitive to tidal currents J73 85 during construction. Thus they would normally be built when the J73 86 expensive turbine caissons and sluices had been placed. This J73 87 aspect, loosely referred to as the 'closure' of the estuary, is J73 88 discussed in Chapter 8.

J73 89 6.2 Embankments

J73 90 When considering embankments, the main questions to be J73 91 addressed are:

J73 92 <*_>bullet<*/>What has been the experience elsewhere?

J73 93 <*_>bullet<*/>What materials would be used?

J73 94 <*_>bullet<*/>How would these materials be transported and J73 95 placed?

J73 96 <*_>bullet<*/>What are the costs?

J73 97 Each of these is addressed in the following Sections.

J73 98 6.2.1 Experience elsewhere

J73 99 The tidal barrage at La Rance includes a short length of J73 100 embankment, but this was built in the dry within the area enclosed J73 101 by the main cofferdams and so was more akin to a normal dam than an J73 102 embankment built in the sea. The 20 MW pilot plant at Annapolis J73 103 Royal in Nova Scotia was built in an existing island and so no new J73 104 embankments were required. Thus there is no direct experience of J73 105 building an embankment for a tidal power scheme. However, there are J73 106 several projects where relevant experience was gained.

J73 107 In Hong Kong in the mid 1960s a dam was built across the mouth J73 108 of an inlet, Plover Cove, in order to form a fresh water reservoir. J73 109 When complete, the water behind the dam was pumped out and the J73 110 reservoir filled with fresh water. The dam was built mainly of J73 111 dredged sand on a foundation of relatively soft seabed deposits. J73 112 Fig. 6.1 shows the completed dam, which bears quite a good J73 113 resemblance to an artist's impression of the proposed embankment J73 114 for the Severn barrage as seen from the Welsh shore at low water J73 115 (Fig. 6.2). Ref. 1965(3) describes the project in detail.

J73 116 figure&caption

J73 117 In the mid 1970s, an even more imaginative project was built in J73 118 Hong Kong to help meet the increasing demand for water. This was J73 119 the High Island reservoir and comprised the building of a large dam J73 120 at each end of High Island to link the island to the mainland. The J73 121 area enclosed was then converted to a fresh water reservoir with a J73 122 top water level some 20 m above sea level. What made this project J73 123 particularly interesting was the building of the cofferdam at the J73 124 eastern end, in water about 30 m deep, in a location exposed to J73 125 typhoon waves from the South China Sea. Fig. 6.3 shows the seaward J73 126 face of the cofferdam, which formed a permanent part of the J73 127 complete structure.

J73 128 In the Netherlands, the Delta project has recently been J73 129 completed. Apart from the Osterschelde storm surge barrier, J73 130 discussed in Chapter 4, and the Haringvleit sluices (Chapter 5), J73 131 this project has comprised a succession of large embankments built J73 132 to close off estuaries or embayments from the sea in order to J73 133 prevent a recurrence of the disastrous flooding which occurred on J73 134 31 January 1953 as a result of an exceptionally high surge tide. J73 135 These embankments have generally been based on an initial closure J73 136 structure which has allowed control to be achieved over the tidal J73 137 flows (Refs. 1965(1), 1972(1)). After closure, the embankments have J73 138 been completed by covering the initial structure with large J73 139 quantities of dredged sand with shallow slopes suitably protected J73 140 against wave attack. The experience gained has left the Dutch J73 141 dredging contractors pre-eminent in their field.

J73 142 figure&caption

J73 143 A variety of closure methods was used in the Delta project, J73 144 including: floating in a series of conrete caissons; a series of J73 145 caissons with temporary openings which were opened as soon as each J73 146 caisson had been floated into position, and then all closed at the J73 147 same time to achieve total closure; and the dropping of thousands J73 148 of 2 t concrete blocks by a specially built cable car system, thus J73 149 building up a long weir which eventually appeared above the sea and J73 150 allowed dredging to begin. A similar system was erected to enable J73 151 the Osterschelde to be closed from the sea, but this was removed J73 152 when the environmental lobby persuaded the government that an open J73 153 barrier with gates should be built instead.

J73 154 In Germany, the river Eider was closed by a clever system which J73 155 started with a series of steel frames across the gap. Perforated J73 156 steel piles were then installed progressively by sliding them down J73 157 slots in the steel frames so that they spanned 'on edge', thus J73 158 presenting a moderate obstruction to the tidal flows (Fig. 6.4). J73 159 Dredged sand was pumped into position between the frames and J73 160 settled out between the piles, some being carried through the J73 161 perforations. Additional piles were inserted as required to prevent J73 162 undue loss of sand, but at the same time to avoid large loads on J73 163 the frames. The only materials to be brought to the site were the J73 164 steel frames and piles. All other methods have involved large J73 165 quantities of rock, concrete blocks or concrete caissons.

J73 166 The projects described above are important for tidal power in J73 167 that they demonstrate a variety of engineering skills that could be J73 168 used to construct embankments for a tidal power sheme. However, J73 169 they have all been built in locations where the tidal range does J73 170 not exceed 3.5 m, one third of the typical spring tide range of a J73 171 good tidal power site. Consequently, the differential heads across J73 172 the structures, and the velocities through remaining gaps, would J73 173 have been much less than those that could occur during the final J73 174 stages of building a tidal barrage. Other problems follow. For J73 175 example, temporary (and permanent) protection against waves during J73 176 construction has to be much more extensive in area if the still J73 177 water level can rise and fall over 10 m instead of 3 m.

J73 178 In 1975, towards the end of a major study of the feasibility of J73 179 storing fresh water in a reservoir formed by enclosing a large area J73 180 of the intertidal foreshore of the Wash bay on the east coast of J73 181 England, by an embankemnt built mainly of dredged sand, a trial J73 182 embankment was built 4.5 km offshore (Fig. 6.5). This trial, J73 183 costing <*_>pounds<*/>2.4 million, was necessary to evaluate the J73 184 practicalities of handling and placing large quantities of sand, J73 185 gravel and rock in a spring tide range of over 6 m, i.e. about J73 186 double the range experienced elsewhere. Ref. 1976(5) is a detailed J73 187 report. This trial highlighted the importance of providing suitable J73 188 protection of the dredged sand against scour by tidal currents and J73 189 waves if large losses of materials or double handling were to be J73 190 avoided. It also demonstrated the feasibility of transporting slope J73 191 protection materials by sea, and of handling and placing them using J73 192 floating plant. Fig. 6.6 shows a 2000 t capacity coaster being J73 193 unloaded by floating cranes onto 500 t capacity barges, this J73 194 operation being necessary because the water depth at the site was J73 195 insufficient for the coasters. Fig. 6.7 shows the placing of filter J73 196 material on the sand. The sand was excavated at a rate exceeding 1 J73 197 t/s by a modest-sized cutter suction dredger from a nearby borrow J73 198 area and pumped directly to the site via a 700 mm diameter pipeline J73 199 laid on the seabed.

J73 200 There are many other marine engineering projects which have J73 201 been successully built and where the experience is of some J73 202 relevance to the design and construction of embankments for a tidal J73 203 barrage. To take one example, islands have been built of dredged J73 204 sand in the Beaufort Sea, north of Canada, for oil exploration. J73 205 J74 1 <#FLOB:J74\>PRESSURE COMFORT CRITERIA FOR RAIL TUNNEL J74 2 OPERATIONS

J74 3 R G GAWTHORPE

J74 4 Head of Aerodynamics, British Rail Research

J74 5 Derby, UK

J74 6 ABSTRACT J74 7 The discomfort felt on the ears by passengers when travelling J74 8 through tunnels has now reached a stage where limits of pressure J74 9 change have to be applied. The choice of limit is not J74 10 straightforward because of the variation in perception of J74 11 discomfort felt between one person and another and also on J74 12 different journeys depending on the journey characteristics. Though J74 13 the variations between people can be dealt with statistically, the J74 14 choice of a limiting pressure criterion appropriate to a particular J74 15 route depends on a number of complex factors. The Paper attempts to J74 16 categorize journeys into four different types, and to allocate a J74 17 pressure change criterion to each one. Such criteria are then used J74 18 for example by Project planners to size the tunnel cross-sections J74 19 on the route in conjunction with a route speed profile, or maybe J74 20 also to consider the need for sealed rolling stock.

J74 21 1. INTRODUCTION

J74 22 A feature of modern rail operations which can cause discomfort J74 23 to train occupants is the sensation felt on the ears from air J74 24 pressure pulses propagated by the trains' motion through tunnels. J74 25 The strength of these pressure changes (see Ref. 1) increases J74 26 approximately with the square of train speed, and multiple pulses, J74 27 of reinforced magnitude, can be produced by the interaction of J74 28 additional trains in the tunnel. Further, the size of the tunnel J74 29 cross-section area relative to that of the train has a strong J74 30 bearing on the pressure magnitude. Thus, considerable care is J74 31 needed with the design of tunnels on new routes, especially those J74 32 for high-speed, to avoid ear discomfort. At the present time, a J74 33 substantial amount of new railway building is taking place around J74 34 the world, often with considerable lengths of route in tunnel. It J74 35 is the nature of things that the public's expectation of comfort is J74 36 continuously growing and this emphasizes the importance of choosing J74 37 the right comfort criterion with which to optimize the tunnel J74 38 design.

J74 39 This Paper is concerned with the questions of how much comfort J74 40 is expected (and indeed demanded) of modern operations, how can a J74 41 limiting degree of discomfort be gauged and defined in measurable J74 42 units, and what does this mean in terms of railway operations. J74 43 These questions are difficult to answer. Sociological, technical, J74 44 and economic considerations play an important part. The passenger's J74 45 view of the appropriate standard of comfort together with a J74 46 railway's preparedness to meet it are largely dependent on the J74 47 current affluence of the former and the economic well-being of the J74 48 latter.

J74 49 Section 2 of this Paper discusses the comparatively small J74 50 number of railway pressure comfort criteria that are known to exist J74 51 at the present time.

J74 52 Section 3 describes some of the more important features of J74 53 human response to pressure change that need to be taken into J74 54 account in a comfort criterion.

J74 55 Section 4 of this Paper examines the range of railway J74 56 operations that are now demanding an expression of sufficient J74 57 comfort level, and discusses whether such a universal definition is J74 58 practicable. Section 4 goes on to propose a more flexible approach J74 59 which allows a limiting pressure specification to be chosen J74 60 according to the particular circumstances of the rail operation J74 61 concerned.

J74 62 2. EXISTING CRITERIA

J74 63 The need to fix limit values of pressure change on board rail J74 64 vehicles for reasons of passenger comfort first arose: a) for J74 65 high-speed passenger trains operating on new routes in hilly J74 66 terrain where there were frequent tunnels b) for new Metro systems J74 67 where commuter journeys were punctuated at frequent intervals by J74 68 underground station stops between sections of tight-bore tunnel. J74 69 The common factor between them was the frequent nature of the J74 70 pressure transients created in the two situations. The crucial J74 71 factors that brought the problem to a head were, in the first J74 72 instance, high train speed and, in the second, the high blockage J74 73 ratios of Metro trains within the tunnel cross-section. It is J74 74 useful to examine the form that the existing known criteria J74 75 take.

J74 76 2.1 Japan Shin Kansen Operations

J74 77 Initial service trials with the prototype 'Bullet Trains' on J74 78 the newly built Shin Kansen routes in the Sixties drew considerable J74 79 adverse reaction to the aural discomfort felt by passengers as they J74 80 passed through the numerous tunnels on these routes. By the current J74 81 standards of the day, the tunnels were built to generous J74 82 proportions (63.4 m2 cross-section) in order to alleviate J74 83 the effect of high speed. Nevertheless, the number of tunnels, and J74 84 consequently the total cost of tunnel construction, limited the J74 85 adoption of even larger sections to reduce the discomfort to more J74 86 acceptable proportions. As a result of this early experience, J74 87 Japanese National Railways decided to embark on the revolutionary J74 88 process of sealing the structure of the passenger coaches so as to J74 89 prevent, or at least largely attenuate, the pressure transients, J74 90 produced within the tunnel external to the train, propagating into J74 91 the coach interiors. In spite of the considerable cost and J74 92 complexity of the modifications to coach construction that this has J74 93 incurred, JNR and its successor JR have persisted with this design J74 94 policy right into their latest rolling stock range.

J74 95 Having taken the decision to resort to pressure-sealed J74 96 construction, then it was feasible to go for a high standard of J74 97 'pressure' comfort whilst still achieving high train speeds. It can J74 98 also be argued that some of the incurred costs can then be balanced J74 99 off against the cost savings to the infrastructure allowed by the J74 100 modest tunnel cross-sections that are feasible. As a target J74 101 standard, JNR have built their Shin Kansen rolling stock to a J74 102 pressure criterion defined as:

J74 103 Max change of pressure = 1000 Pa

J74 104 Max rate of change of pressure = 200 Pa/s

J74 105 However, it is understood that recent reviews of the J74 106 maintenance costs associated with meeting these standards, J74 107 particularly as the stock gets older, have persuaded JR to relax J74 108 their criterion to:

J74 109 Max rate of change of pressure = 300 Pa/s

J74 110 It seems that the specification of overall pressure change of 1 J74 111 kPa (for which a minimum time scale is only implied) is left J74 112 unaltered, although it would appear that this could be infringed in J74 113 long tunnels.

J74 114 The above criteria have been adopted for 'sealed' Shin-Kansen J74 115 stock, but it is believed that no corresponding criterion exists J74 116 for Japanese conventional 'unsealed' stock.

J74 117 2.2 US Underground Rapid Transit Systems

J74 118 The American Dept of Transportation (UMTA) have produced J74 119 recommendations (Ref.2) for subway train conditions. Subway tunnels J74 120 tend to be of tight bore and a typical journey involves several J74 121 tunnel transits as trains pass from tunnel into station and later J74 122 back into tunnel. Thus, the resulting pressure changes are J74 123 experienced in rather quick succession and all too often by J74 124 commuter travellers who may be users on a regular daily basis. J74 125 Thus, such travel may demand tight control over pressure J74 126 variation.

J74 127 UMTA recommend a pressure standard which may be expressed J74 128 as:

J74 129 Max pressure change = 700 Pa within a 1.7 s period

J74 130 Max rate of change of pressure = 410 Pa/s

J74 131 (over periods longer than 1.7 s)

J74 132 2.3 British Railways Board

J74 133 In the early-Seventies, a need arose in the UK for a pressure J74 134 criterion to cater for the new generation of electric trains which J74 135 were to operate on the newly electrified West Coast Main Line at J74 136 significantly higher speeds. Although this route had relatively few J74 137 tunnels and these were of the short-to-medium length category, J74 138 their cross-sectional areas were small (many less than 40 J74 139 m2) and thus pressures could be substantial. A study of J74 140 existing conditions, which were apparently acceptable to the J74 141 public, and of the results of pressure chamber tests with volunteer J74 142 subjects (Ref. 3) resulted in a limiting criterion defined as:

J74 143 Max pressure change = 3000 Pa within any 3 s period

J74 144 At the same time, BR adopted a further criterion which was J74 145 specifically chosen for operation through the Anglo-French Channel J74 146 Tunnel that was being actively pursued in the early 1970s. The J74 147 configuration of the Tunnel being considered at that time was very J74 148 similar to the present day Tunnel and concern was expressed at the J74 149 pressure pulses that would be experienced on the Shuttle Trains as J74 150 they passed the cross-connecting ducts linking the two separate J74 151 running tunnels. Train operation at 120 km/h meant that these ducts J74 152 were passed at intervals of approximately 7 seconds, and the J74 153 British Railways Board (one of the members of the British Channel J74 154 Tunnel Company) adopted a standard for a desirable comfort J74 155 level during normal operation through the Tunnel which stated:

J74 156 Max pressure change = 450 Pa

J74 157 (when repeated at 7s intervals for the 25 min tunnel J74 158 journey)

J74 159 A corresponding standard of 700 Pa was considered to be the J74 160 upper limit of acceptability for such tunnel operations. A rare J74 161 extreme case was not identified.

J74 162 In 1986, BRB revised their former criterion for inter-city J74 163 route operation and effectively relaxed the standard to allow:

J74 164 Max pressure change = 4000 Pa within any 4 s period

J74 165 Thus, the new criterion allows the maximum short duration J74 166 pressure change to be 4000 Pa rather than the previous 3000 Pa. The J74 167 basis for the decision was favourable operating experience with the J74 168 3000 Pa criterion together with the results of international J74 169 studies (Ref. 4).

J74 170 The 4000 Pa within 4 s criterion remains the BRB comfort limit J74 171 for all BR routes except for the proposed Rail Link between London J74 172 and the Channel Tunnel. This route is exceptional within the BR J74 173 network as it is proposed to have several tunnels (amounting to J74 174 approximately 30% of total route length). A recommended pressure J74 175 criterion for the route has been proposed to be:

J74 176 2.5 kPa within 4 s for single-track tunnels

J74 177 3.0 kPa within 4 s for double-track tunnels.

J74 178 The higher pressure permitted in a double-track tunnel allows J74 179 for the fact that this limit will only occur when two trains pass J74 180 at maximum speed within the tunnel and even then only when they J74 181 pass in a particular place within the tunnel. For single-track J74 182 tunnels, the highest pressure changes felt on each occasion will J74 183 always be the same (for a given operating speed). At the time of J74 184 writing, project studies show that relatively few tunnels are J74 185 single-track and that therefore this increased level is reasonable, J74 186 and consistent with route infrastructure economies. Notional J74 187 maximum speed for the route with 'non-sealed' stock is 225 km/h.

J74 188 2.4 The Deutsche Bundesbahn Neubaustrecke Routes

J74 189 In the early-eighties, DB decided to build a network of new J74 190 high-speed routes which could also be used for mixed traffic J74 191 operation. The first of these routes, from Hannover to Wuerzburg, J74 192 involved approximately 30% of the route length in tunnel due to J74 193 topographical and environmental restrictions. DB designed the J74 194 railway for operation at 300 km/h with an ultimate goal in mind of J74 195 350 km/h. Consequently, large diameter tunnels were planned J74 196 (cross-section 82 m2) to limit the pressure problem and J74 197 to ease traction energy costs associated with aerodynamic drag.

J74 198 During pressure comfort tests held in the Derby pressure J74 199 chamber simulating future NBS operation (Ref. 5) with unsealed J74 200 stock, the senior management of DB came to the decision that they J74 201 should adopt sealed rolling stock for future high-speed operation. J74 202 It was felt that the superior standards of comfort designed into J74 203 the NBS operation, with the ICE train in particular, necessitated, J74 204 at the same time, an improved pressure comfort standard. Because of J74 205 the similarities between NBS operation and Japanese Shin Kansen J74 206 services, DB decided to adopt the JNR criteria:

J74 207 Max change of pressure = 1000 Pa

J74 208 Max rate of change of pressure = 200 Pa/s

J74 209 although it is understood that the 200 Pa/s limit may be J74 210 relaxed to 300-400 Pa/s.

J74 211 2.5 Others

J74 212 Other high-speed routes such as the Italian Direttissima and J74 213 French TGV-Atlantique line have multiple tunnels, and presently use J74 214 'non-sealed' rolling stock, but FS and SNCF have not declared their J74 215 limiting pressure criteria. However, it is known that the next J74 216 generation of coaching stock to be used with the new Italian ETR500 J74 217 train and the proposed TGV-R train will be pressure-sealed.

J74 218 There is no standard definition of 'pressure-sealed' stock. J74 219 However pressure sealing has been quantified for test purposes (see J74 220 for example Ref. 6) in relation to the degree of sealing of a J74 221 railway coach by a time constant associated with the time decay of J74 222 internal pressure of the coach (initially at an over-pressure) J74 223 after the pressurizing means (eg ventilation fan) is removed, J74 224 according to the law formula

J74 225 where Pi = coach internal pressure

J74 226 PI = coach initial internal pressure

J74 227 t = time in seconds from Pi = PI

J74 228 <*_>tau<*/> = time constant in seconds

J74 229 J75 1 <#FLOB:J75\>2.2 Two Phase Structures, One Three-dimensional J75 2 and One Two-dimensional Grain Boundary Phase

J75 3 2.2.1 Binary alloys

J75 4 Most metallic alloying elements purposely added to improve J75 5 properties are highly soluble and do not show a marked tendency to J75 6 segregate on grain boundaries. Those of interest here are mainly J75 7 non-metallic or metalloid elements present as impurities. The grain J75 8 boundary is here treated as a separate phase which, being very J75 9 thin, is designated two dimensional (2D). Solute elements and J75 10 vacancies are partitioned between the bulk grain, i.e. the 3D J75 11 phase, and the grain boundary.

J75 12 In an alloy consisting of two elements and two bulk phases, J75 13 such as <*_>alpha<*/><*_>beta<*/> brass, addition of extra solute J75 14 element would merely change the relative volumes of the two phases J75 15 without any change in composition of either. This cannot happen in J75 16 the grain - grain boundary case since the volume (area) of the J75 17 latter is fixed. Adding more solute, therefore, increases the J75 18 concentration in both the 3D and the 2D phase. In his original J75 19 analysis of the thermodynamics of the partition McLean supposed the J75 20 grain boundary to consist of a fixed number of exactly equivalent J75 21 sites where solute atoms can exchange with solvent atoms with a J75 22 drop in energy of the system by an amount <*_>delta<*/>G per mole J75 23 so exchanged. He arrived at a relationship which has since been J75 24 widely used:

J75 25 formula

J75 26 where X refers to the mole fraction of the segregated solute J75 27 species on the grain boundary and x its mole fraction in the bulk J75 28 grain, R is the molar gas constant and T the absolute temperature. J75 29 The value of Xo, the concentration of X at which the boundary J75 30 appears to be saturated, can be obtained for each system by J75 31 experiment. For some systems saturation occurs at a monolayer of J75 32 atoms. For other systems Xo is a fraction of a monolayer. For atoms J75 33 which segregate on the boundary <*_>delta<*/>G is negative - the J75 34 total free energy of the system is reduced when solute atoms fall J75 35 into grain boundary traps and the term exp(-<*_>delta<*/>G/RT) is J75 36 large and positive. X/(Xo-X) is the ratio of solute to solvent J75 37 atoms on the grain boundary and x/(1-x) the ratio in the bulk J75 38 grain. The term exp(-<*_>delta<*/>G/RT) is therefore the grain J75 39 boundary enrichment ratio which can take large positive values up J75 40 to about 104 for values of <*_>delta<*/>G that are found J75 41 in practice.

J75 42 The equation has the same form as the earlier Langmuir equation J75 43 for adsorption on the free surface of a solid, but the analogy J75 44 should not be taken too far, since the environment and constraints J75 45 on atoms at a grain boundary are very different from those of a J75 46 free surface. The latter is an empty half-

space and the J75 47 co-ordination number of atoms on a surface ranges from 1 to about J75 48 8, while for atoms on a grain boundary there is only a small or J75 49 zero reduction in co-ordination number compared with the bulk grain J75 50 value and can be in the range 8 to 11 for close-packed metals.

J75 51 McLean assumed that the grain boundary was 3 atoms thick and J75 52 that the substitutional atomic sites were of three types: oversize, J75 53 undersize and average. The sites of each type were assumed to be J75 54 exactly equivalent so that exchange of a solvent atom for a solute J75 55 atom always involved the same energy change. The sites were J75 56 therefore assumed to be pre-existing and unchanged by the J75 57 segregation process. Since next nearest neighbour interactions are J75 58 significant when exchanging a metal atom for a non-metal or J75 59 metalloid having also a large size difference, the McLean model is J75 60 valid only for low occupancy, i.e. about 1 to 2%. For fracture J75 61 studies much higher levels of occupancy are of interest, usually in J75 62 the range of 10 to 100% of a monolayer. Experimental results even J75 63 here have been 'fitted' roughly to the McLean equation although the J75 64 atomic model is no longer tenable because exchanging a layer of J75 65 metal atoms for non-metal or metalloid at the boundary must J75 66 severely change its structure. Many examples quoted later show a J75 67 far from linear relationship between bulk concentration and grain J75 68 boundary occupancy. In spite of the range of powerful experimental J75 69 techniques available, however, it has not been possible to reveal J75 70 directly the detailed way in which segregated elements are J75 71 distributed on grain boundaries and how the structures evolve. They J75 72 may be randomly distributed or in small rafts or even in multilayer J75 73 plates. Also, in areas of strong segregation of two elemental J75 74 species, they may or may not occupy nearest neighbour sites.

J75 75 Preferred models of the boundary and for build up of segregated J75 76 layers begin in each case with a boundary in which the most J75 77 disturbed sites are confined to a single mono-atomic layer. On the J75 78 centre layer there are only a few sites for substitution of solute J75 79 atoms where exchange would involve an energy change of J75 80 <*_>delta<*/>G max. These sites can be arbitrarily chosen to be the J75 81 large sites which acquire large solute atoms in place of solvent J75 82 atoms. When the solute atom arrives the neighbouring atoms relax J75 83 outwards, i.e. the two grains are prized part, so that what were J75 84 the next largest sites become as large as the pre-existing large J75 85 sites. These are now, in turn, occupied by solute atoms, so that J75 86 the boundary is further prized apart, thus creating new large J75 87 sites. Thus the value of <*_>delta<*/>G remains constant.

J75 88 If 'large' sites are created immediately adjacent only to J75 89 segregated solute atoms, then the layer will develop by nucleation J75 90 and growth of platelets. Otherwise, for more general grain boundary J75 91 widening the development might go through a series of ordered J75 92 arrangements. Large atoms might find large sites by dropping into J75 93 vacant sites on the boundary and diffusing within the boundary.

J75 94 It is obviously unsatisfactory to discuss the segregation up to J75 95 high occupancy entirely in terms of the size of the grain boundary J75 96 sites, since each newly arriving impurity atom will alter the J75 97 electron distribution over many interatomic distances. For example, J75 98 in iron the clean boundary has metallic bonding while the boundary J75 99 with a high sulphur content has weak bonding across the boundary J75 100 with ionic and/or covalent bonding in the boundary. The fact that J75 101 <*_>delta<*/>G remains constant therefore signifies only that, as J75 102 the boundary fills and adopts a series of new structures and types J75 103 of bonding, the energy drop for each newly arrived atom remains J75 104 sensibly constant.

J75 105 From the previous discussion it is clear that the J75 106 <*_>delta<*/>G = const. as the grain boundary fills with impurity J75 107 is not a special case; but rather is accidental. For phosphorous in J75 108 steel the constancy has been well established as phosphorous J75 109 substitutes for iron atoms. The same model can be used for the J75 110 cases where <*_>delta<*/>G increases or decreases with occupancy, J75 111 with the modification that as each atom arrives it creates sites J75 112 respectively greater than or not so great as those which were first J75 113 occupied or offer a more or a less attractive electron J75 114 configuration. Alternatively, as the boundary fills, the solute J75 115 atoms take up a series of ordered structures, each more stable than J75 116 the preceding one, i.e. <*_>delta<*/>G increases, or it may be less J75 117 stable, i.e. <*_>delta<*/>G decreasing. These changes may not be J75 118 smooth but are, perhaps, difficult to detect, except in certain J75 119 cases where saturation is reached and <*_>delta<*/>G suffers a J75 120 large step change in value. In the exceptional case of tin in pure J75 121 iron, saturation does not occur until over two monolayers have been J75 122 formed. The same may be true for iron containing antimony.

J75 123 There are three important consequences for systems where J75 124 <*_>delta<*/>G increases as the concentration of solute in the J75 125 boundary increases:

J75 126 1. At constant temperature over a certain range the grain J75 127 boundary equilibrium content of solute is a step function of the J75 128 concentration in the bulk grains. In an extreme case this would be J75 129 a step function over part of the range as shown in the diagram, J75 130 Fig. 6.

J75 131 2. The temperature range <}_><-|>betwen<+|>between<}/> J75 132 segregation and desegregation is very narrow. Again in an extreme J75 133 case the gap would almost disappear and the grain boundary J75 134 transition would approach first order kinetics.

J75 135 3. Since the conditions for near complete segregation or J75 136 desegregation are so critical, the quality of the grain boundary J75 137 structure is also critical. In the critical range, therefore, some J75 138 grain boundary facets are near saturation while others remain J75 139 unsegregated. This third effect tends to obscure the effects of 1. J75 140 and 2. since measurements of grain boundary composition are usually J75 141 taken over many facets and give an average value for heavily and J75 142 lightly segregated facets. The build up of segregation, therefore, J75 143 is smoothed out, both as a function of temperature and as a J75 144 function of concentration within the grains, even when very sharp J75 145 transitions occur on individual facets.

J75 146 Figure 6 shows grain boundary concentration versus bulk J75 147 concentration for selenium and tellurium in pure iron. In each case J75 148 a factor 3 to 4 in bulk concentration is sufficient to cover the J75 149 range from low grain boundary occupancy to saturation. But even J75 150 here the measurements are the average of many grain facets, some J75 151 heavily segregated, others almost free of segregate. Therefore the J75 152 possibility of near first order kinetics is obscured.

J75 153 A drop in the <*_>delta<*/>G value could arise simply from the J75 154 filling of pre-existing sites, starting with the deepest traps and J75 155 then filling shallower traps. Such an explanation has been given J75 156 for the segregation of sulphur in Ni3Al and J75 157 Ni3(Ti, Al) alloys. The grain boundary occupancy changed J75 158 with temperature increase more slowly than was consistent with a J75 159 single value of <*_>delta<*/>G. Thus at the higher temperature only J75 160 the deeper traps were filled. Over the temperature range 977 to J75 161 1045K there was only a factor 2 decrease in boundary sulphur J75 162 concentration and the maximum concentration was estimated to be J75 163 only 5 at.%. The pre-existing site model could here be accepted. J75 164 For higher concentrations, however, the interaction of solute atoms J75 165 on the boundary would have to be conceded. The ordered structure of J75 166 the Ni3Al type intermetallic compounds does, J75 167 nevertheless, give rise to lack of adaptability of atomic positions J75 168 to the development of close-packed grain boundary structures.

J75 169 figure&caption

J75 170 2.2.2 Ternary alloys

J75 171 Still restricting discussion to 'two phase' systems, i.e. a J75 172 single bulk solid solution phase plus a grain boundary, the first J75 173 set of examples include those where both solute species would J75 174 strongly segregate to grain boundaries if present individually. J75 175 When present together they might either:

J75 176 1. show no interaction and each segregate as though the other J75 177 were not present, or

J75 178 2. strongly compete for sites, or

J75 179 3. segregate co-operatively so that each enhances the J75 180 segregation of the other. An example of this type has not been J75 181 reported.

J75 182 In a typical commercial steel there are many elements on grain J75 183 boundaries in small amounts which may belong to the first type, but J75 184 the best example is from the classical work of Seah and Hondros who J75 185 demonstrated in pure iron that tin up to a monolayer and sulphur up J75 186 to 50 at.% coverage each segregated as though the other were not J75 187 present, i.e. the level of each was not affected by the presence of J75 188 the other. Whether the two elements segregate on different parts of J75 189 the boundary with tin forming a double layer, whether sulphur is at J75 190 the tin-iron interface or whether the two elements form a solid J75 191 solution or compound layer is open to speculation.

J75 192 The interstitial elements carbon, nitrogen and boron, so J75 193 important in ferrous metallurgy, segregate strongly to ferrite J75 194 grain boundaries. Unlike the substitutional segregants their J75 195 presence in steels has a number of beneficial effects, including J75 196 their ability to strengthen grain boundaries and so counteract the J75 197 effect of embrittling species. These elements in ferrite exchange J75 198 interstitial sites in the grains for interstitial sites in the J75 199 boundary. The enrichment ratio can be very high since interstitial J75 200 sites in the b.c.c. structure are small. F.c.c austenite has larger J75 201 interstitial sites and grain boundary enrichment is consequently J75 202 less.

J75 203 Competition between carbon and phosphorus on the boundaries of J75 204 ferrite is of great practical importance since carbon strengthens J75 205 and phosphorus embrittles boundaries. Of the two, carbon has the J75 206 greater affinity for the grain boundary sites and reduces the J75 207 amount of phosphorus segregated. J75 208 J76 1 <#FLOB:J76\>POWERGEN'S APPROACH TO THE IMPLEMENTATION OF J76 2 F.G.D.

J76 3 J Evans and S K Reynolds

J76 4 The paper outlines PowerGen's requirement for FGD plant and J76 5 discusses the selection of station site for retro-fitting, the J76 6 process type adopted and the main design requirements for the J76 7 plant. The paper concentrates on the project, contract and J76 8 commercial strategies for its implementation.

J76 9 Details are also given of PowerGen's first FGD plant at J76 10 Ratcliffe-on-Soar power station near Nottingham.

J76 11 INTRODUCTION J76 12 This paper discusses the sulphur abatement programme proposed J76 13 by PowerGen and the necessity to install Flue Gas Desulphurisation J76 14 (FGD) plants. The paper concentrates on PowerGen's technical & J76 15 commercial approach to FGD and identifies the key features of the J76 16 first plant proposed for Ratcliffe on Soar power station.

J76 17 SULPHUR ABATEMENT REQUIREMENTS

J76 18 The effects of acid rain and the need for control on S02 J76 19 emissions have been debated within the EEC for several years. These J76 20 discussions culminated in the production of the 'Large Combustion J76 21 Plant (LCP) Directive', proposed in draft in December 1983 and J76 22 agreed in principle in June 1988.

J76 23 The Directive covers two distinct areas, namely:-

J76 24 (a) New combustion plant The limits for S02 emissions J76 25 etc are based on best available technology.

J76 26 (b) Existing combustion plant The limits are covered J76 27 in terms of national emission limits, i.e a 'bubble' limit. These J76 28 bubbles are set for each of the EEC member states and have three J76 29 target implementation dates, 1993, 1998, and 2003. Adoption of J76 30 these bubble limits will necessitate reductions in S02 emission J76 31 levels of 20%, 40% and 60% respectively for the UK (based on the J76 32 1980 emissions).

J76 33 THE REQUIREMENT FOR FGD RETROFITS

J76 34 In order to meet its own sulphur abatement targets, PowerGen, J76 35 as a power utility, now has a number of options available to it. J76 36 These include the following:-

J76 37 <*_>bullet<*/> Increasing thermal efficiency of existing coal J76 38 fired plant

J76 39 <*_>bullet<*/> Increased use of low sulphur fuel imports

J76 40 <*_>bullet<*/> Increased use of natural gas

J76 41 <*_>bullet<*/> Increased use of alternative energy supplies

J76 42 <*_>bullet<*/> Use of 'clean coal' technology

J76 43 <*_>bullet<*/> Retrofitting FGD plant to existing stations

J76 44 PowerGen has examined the most economic and environmentally J76 45 acceptable way of achieving the sulphur abatement targets. The J76 46 process is however inextricably linked to almost all the major J76 47 investment decisions currently facing the company. The commitment J76 48 to reduce sulphur emissions is not just a question of assessing the J76 49 amount of FGD required to reduce emissions from current emitters, J76 50 it involves assessing both the current company emissions and the J76 51 long term future emissions from any future plant. This is J76 52 complicated by a number of factors:-

J76 53 <*_>bullet<*/> PowerGen has to compete for market share within J76 54 the UK market which may change generation capacity within the J76 55 company.

J76 56 <*_>bullet<*/> Numerous gas turbine schemes are under J76 57 consideration.

J76 58 <*_>bullet<*/> Prediction of future coal sources (and their J76 59 sulphur levels) is difficult.

J76 60 Taking the above into account, PowerGen applied to the J76 61 Secretary of State for consent to retro-fit 4000MW of Flue Gas J76 62 Desulphurisation Plant.

J76 63 SELECTION OF F.G.D. PROCESS TYPE

J76 64 There are over 100 different types of FGD systems available J76 65 worldwide, ranging from pilot plant processes to systems installed J76 66 on several thousand megawatts of combustion plant. Detailed J76 67 engineering studies on a number of different FGD process systems J76 68 were carried out under the auspices of the C.E.G.B. and have been J76 69 progressed further by PowerGen. Great emphasis has also been made J76 70 on the operating experience of overseas FGD installations.

J76 71 In selecting the type of system to be used, PowerGen has taken J76 72 into account the following points:-

J76 73 <*_>bullet<*/> The stations will operate on medium to high J76 74 sulphur, high chloride coals.

J76 75 <*_>bullet<*/> The main generation stations are large (- J76 76 2000MW) sites

J76 77 <*_>bullet<*/> The boiler units are large (500MW) and are all J76 78 20-25 years old

J76 79 <*_>bullet<*/> The boiler units are required to operate for 38 J76 80 months between major overhauls.

J76 81 <*_>bullet<*/> The stations operate on a high load factor J76 82 (>80%)

J76 83 <*_>bullet<*/> Production of a 'socially acceptable' by-product J76 84 required

J76 85 <*_>bullet<*/> Minimisation of other effluents/waste J76 86 products

J76 87 <*_>bullet<*/> Low lifetime cost required

J76 88 <*_>bullet<*/> Minimum maintenance requirement

J76 89 <*_>bullet<*/> Proven reference list of FGD process required

J76 90 Considering the above factors, the Limestone/Gypsum process is J76 91 considered the most suitable for PowerGen's retrofit FGD J76 92 systems.

J76 93 This process is the most widely used FGD type worldwide and its J76 94 selection is in line with world FGD trends.

J76 95 SELECTION OF STATION SITES FOR FGD RETROFITTING

J76 96 In identifying which stations would be suitable for J76 97 retro-fitting an FGD plant, a number <}_><-|>a<+|>of<}/> factors J76 98 require consideration:-

J76 99 (a) Commercial Aspects As FGD represents a major J76 100 capital investment, maximum benefit (in terms of the mass of J76 101 sulphur removed) must be gained from its installation. This entails J76 102 considering only large stations, operating on high load factors J76 103 (i.e. base load operation) and also firing high sulphur fuels. J76 104 Coastal stations which may import low sulphur overseas coals are J76 105 not considered at present. Remaining station life also needs to be J76 106 taken into account in assessing the lifetime costs of the FGD J76 107 system.

J76 108 (b) Environmental Aspects Given the size of the above J76 109 stations (2000MW) and the relatively high fuel sulphur content, an J76 110 FGD system will require large quantities of reagent feedstock and J76 111 will produce large quantities of by-products. Transport of J76 112 feedstock and by-product at the power station is a major design J76 113 consideration. The process will also produce large quantities of J76 114 liquid effluent.

J76 115 (c) Station Factors Most of the large coal fired J76 116 stations within PowerGen are now over 20 years old.

J76 117 Taking into account the above factors, the most suitable power J76 118 stations identified within PowerGen were Ratcliffe-on-soar near J76 119 Nottingham and Ferrybridge C near Leeds. Both are 2000MW base load J76 120 stations.

J76 121 DESIGN REQUIREMENTS OF THE F.G.D. PLANT

J76 122 Stack Emissions

J76 123 The control of gaseous emissions are monitored and regulated in J76 124 the UK by HM Inspectorate of Pollution (HMIP). They in turn report J76 125 to the Department of the Environment (DOE). Regulations concerning J76 126 the emissions from large combustion works (>500MW thermal) are J76 127 covered in their 'Best Practicable Means Note No.27'. This note J76 128 however, only applies to new combustion plant<&|>sic!, there are no J76 129 current regulations concerning the design of a retrofit FGD plant. J76 130 In view of this, PowerGen proposed its own design parameters for J76 131 the FGD plant which were discussed and agreed with HMIP and other J76 132 relevant bodies. The principal parameters proposed were as J76 133 follows:-

J76 134 (a) Sulphur Dioxide Emissions The FGD plant will be J76 135 designed to achieve 90% SO2 removal at the mean coal sulphur level J76 136 with one absorber spray bank redundant. This redundant bank will J76 137 serve as an operational spare and will also be brought into service J76 138 if a high sulphur coal is fired. The plant is designed to handle J76 139 coal sulphur levels up to 3% (by wt) and coal chloride levels up to J76 140 0.6% (by wt).

J76 141 (b) Stack Discharge Temperature The temperature of J76 142 the emitted gases entering the chimney would be maintained at not J76 143 less than 80 degC. This means that reheating of the flue gas is J76 144 required.

J76 145 (c) Dust Emissions Dust emission from the process J76 146 would not exceed 85 mg/m3 (@15 degC, 1 bar, dry, 6%02).

J76 147 Liquid Effluent Emissions

J76 148 Due to the relatively high chloride content of UK coal, a J76 149 liquid purge stream will be required to control the level of J76 150 soluble chloride in the process. This stream will discharge calcium J76 151 chloride into the local watercourse.

J76 152 The purge stream however contains not only calcium chloride but J76 153 also flyash not removed in the ESPs, and contaminant trace metals J76 154 from the coal, limestone and process water feeds. This stream J76 155 requires treatment prior to discharge in the local water course to J76 156 reduce these associated contaminants to acceptable levels. J76 157 Treatment will be carried out in 2 x 100% water treatment plants J76 158 and will result in a waste sludge being produced. This will require J76 159 disposal, probably by landfill.

J76 160 The control of discharges into the rivers is controlled in the J76 161 UK by the National Rivers Authority (NRA). The large size of J76 162 PowerGen's stations, the high chloride content of UK coals and the J76 163 relatively low flowrates of UK rivers result in FGD discharges J76 164 being significant. Consequently the discharge limits applied by NRA J76 165 are very tight. Discharges are normally applied in terms of maximum J76 166 outfall concentrations, total mass discharges of selected species J76 167 and total mass of chlorides discharged.

J76 168 Plant Availability

J76 169 As a result of the Environmental Protection Act (1990) all J76 170 large combustion plant<&|>sic! will be required to operate under an J76 171 authorisation granted by HMIP. Whilst planned maintenance or J76 172 occasional outages from a well designed and operated FGD plant may J76 173 be acceptable, frequent outages from a poorly designed and operated J76 174 plant will not. Thus there is requirement to design and operate the J76 175 plant to high availability and reliability standards.

J76 176 Whilst HMIP have not stated an acceptable design availability, J76 177 they are aware of values adopted by overseas utilities, in J76 178 particular West Germany. In West Germany the legislation allows the J76 179 combustion plant to be operated for a maximum of 10 days per year J76 180 without the FGD operating (there is a further proviso that each J76 181 outage must not last longer than 72 hours). This 10 day outage J76 182 allowance means the FGD must be available for 355 days per year J76 183 (assuming continuous boiler operation) - ie a plant availability of J76 184 97.5%. Thus this availability figure, in the absence of specific UK J76 185 values, has been selected by PowerGen as a design figure. This J76 186 represents a very high target value, particularly in relation to J76 187 the PowerGen requirement for 38 months continual operation between J76 188 overhauls and the boiler availability levels of approximately 90% J76 189 currently being achieved.

J76 190 Quality of By Product

J76 191 It is the stated aim of PowerGen to produce a socially J76 192 beneficial by-product from its FGD systems, which with the J76 193 Limestone/Gypsum process results in the production of a commercial J76 194 grade gypsum.

J76 195 Local Planning Restrictions

J76 196 Other restrictions on the process are imposed by local J76 197 regulations such as visual aspects, noise levels (construction and J76 198 operating), road & rail traffic movements etc. These all have to be J76 199 catered for in the plant design.

J76 200 All aspects associated with the design of the FGD plant are J76 201 outlined in the PowerGen's Environmental Statement (2) which is J76 202 produced for the Local Authorities, County Councils and other J76 203 Regulatory Bodies, local groups, and the General J76 204 Public<}_><-|>,<+|>.<}/>

J76 205 PROJECT & CONTRACT STRATEGY

J76 206 Type of Contract

J76 207 Traditionally within the Electrical Supply Industry the project J76 208 management of a large capital plant such as an FGD plant would have J76 209 been achieved by the letting of several contracts. A main J76 210 'mechanical' contract would have covered the unit plant and most of J76 211 the common plant, with separate contracts for the preliminary J76 212 civils works, main civils works, mechanical handling equipment, J76 213 electrical and C & I works. The contracts would also have been let J76 214 against tight, comprehensive specifications produced in-house.

J76 215 Whilst this form of tendering procedure produces a clearly J76 216 defined plant scope, it has a number of disadvantages:-

J76 217 <*_>bullet<*/> a large in-house team is required to produce the J76 218 detailed specification and to monitor compliance of the tenders.

J76 219 <*_>bullet<*/> the tenderers are forced into offering a plant J76 220 design which may be slightly outside their experience and J76 221 consequently carry a risk element.

J76 222 <*_>bullet<*/> the tenderer has little room to J76 223 <}_><-|>optimised<+|>optimise<}/> the plant design

J76 224 <*_>bullet<*/> conventional specifications tend to concentrate J76 225 on component rather than system designs

J76 226 <*_>bullet<*/> there is little scope for fully evaluating all J76 227 the plant options

J76 228 <*_>bullet<*/> there is a tendency to 'over-specify' the J76 229 plant

J76 230 <*_>bullet<*/> numerous contracts would pose interface risks to J76 231 PowerGen

J76 232 In view of this, PowerGen decided to opt for a different J76 233 contract strategy which would involve a central, lump sum, turn-key J76 234 contract for each station, let against a performance type J76 235 specification. The contract covers the main FGD plant, all the J76 236 materials handling plant, the water treatment plant and all J76 237 associated plant. It also encompassed mechanical, electrical, C & J76 238 I, and all civil engineering and building works. PowerGen will only J76 239 be placing separate contracts for the following minor J76 240 contracts:-

J76 241 <*_>bullet<*/> Pre-application studies

J76 242 <*_>bullet<*/> Preliminary works design

J76 243 <*_>bullet<*/> Preliminary works

J76 244 <*_>bullet<*/> Unit transformers

J76 245 <*_>bullet<*/> Railway signalling

J76 246 <*_>bullet<*/> 11kv and 3.3kv switchgear mods

J76 247 <*_>bullet<*/> Landscaping

J76 248 <*_>bullet<*/> Finishing works

J76 249 An important feature of the contract strategy is that the J76 250 enquiry was issued simultaneously for both the Ratcliffe and J76 251 Ferrybridge plants, with the contracts being linked together. This J76 252 enables a high degree of standardisation to be achieved and the J76 253 process designs of both plants will be identical.

J76 254 Selection of Tenderers

J76 255 Considerable effort was made in the selection of suitable J76 256 tenderers for the FGD project. Generally the tenderers consisted of J76 257 a UK engineering company linked to an overseas group providing the J76 258 process technology. J76 259 J77 1 <#FLOB:J77\>AIRCRAFT DESIGNER'S VIEWPOINT OF RELIABILITY J77 2 AND MAINTAINABILITY

J77 3 A J Hill

J77 4 Westland Helicopters Limited

J77 5 Yeovil, England

J77 6 Abstract

J77 7 Papers on Reliability and Maintainability (R&M) stress the J77 8 importance of R&M design to optimise Life Cycle Costs and J77 9 Availability. This view is reflected in the requirements placed J77 10 upon aircraft manufacturers by Civil and MoD customers to the J77 11 extent that R&M is frequently given equal priority with, for J77 12 example, aircraft performance requirements.

J77 13 During the initial aircraft design phase and subsequent J77 14 development phase, R&M engineers advise the designer on R&M J77 15 requirements but these requirements are frequently in direct J77 16 conflict with other requirements such as weight, aircraft J77 17 performance and cost. Although the idealised view is to treat these J77 18 requirements as equal, the reality is that aircraft performance and J77 19 weight are frequently perceived to be more equal than other J77 20 specialist requirements, including R&M, particularly if the J77 21 aircraft experiences weight problems.

J77 22 The aircraft designer has to reconcile these conflicting J77 23 requirements to achieve the overall aircraft design. Frequently, J77 24 this results in rejection of the design recommendations of J77 25 specialist departments and adopting alternative design solutions to J77 26 meet the aircraft design objectives.

J77 27 This paper discusses an aircraft designer's view of R&M design, J77 28 its relationship to other design requirements and the problems J77 29 raised by the often conflicting requirements.

J77 30 Discussion

J77 31 Introduction

J77 32 Ladies and gentlemen, you will note from the Chairman's J77 33 introduction that my background is in aircraft maintenance, R&M and J77 34 Design to Cost. Therefore, I am only too aware that R&M is J77 35 extremely important in aircraft design and the subsequent success J77 36 of the aircraft in the market place, either in the military market J77 37 or in the civil market.

J77 38 The necessity for Failure Mode Effect and Criticality Analyses, J77 39 Direct Maintenance Cost Analyses and the many other R&M tools used J77 40 in the design and development of an aircraft is acknowledged: J77 41 however, these tools will not be discussed in this paper. I will J77 42 put forward some views of an aircraft designer on R&M issues with J77 43 an emphasis on maintainability. I have held the job of an assistant J77 44 to a Chief Designer for some years and I am, therefore, familiar J77 45 with R&M and other design requirements.

J77 46 Most papers on R&M are presented by R&M specialists who J77 47 frequently exclude other disciplines. This paper will attempt to J77 48 show that the aircraft designer has to reconcile the requirements J77 49 of R&M with other, often conflicting, requirements.

J77 50 R&M Requirements

J77 51 R&M has become increasingly important over the years with the J77 52 MoD and civil customers requiring good R&M features to minimise J77 53 Life Cycle Costs and to maximise the Availability of the aircraft. J77 54 This is a typical statement which is included in aircraft J77 55 specifications from the earliest stage of an aircraft project:

J77 56 Reliability and Maintainability shall be given equal priority J77 57 to aircraft performance

J77 58 To be given equal priority to aircraft performance means that J77 59 R&M requirements must be clearly defined before the design begins J77 60 and closely monitored during the design and development phases. J77 61 Figure 1 shows a simple bar chart defining the phases and the J77 62 activities addressed in each phase.

J77 63 In my view, the most important phase of a project is the J77 64 definition of the R&M requirements. In addition to requirements J77 65 expressed in rates, such as failures per flying hour and J77 66 maintenance man hours per flying hour, the R&M design features must J77 67 also be clearly defined. Loosely defined or ambiguous requirements J77 68 frequently result in very expensive re-design activities when J77 69 problems arise during the development phase. However, and this is a J77 70 very important point, the requirements must not be excessive for J77 71 the task.

J77 72 The next important activity is the initial design phase which J77 73 must be given sufficient time for R&M specialists to advise J77 74 designers and draughtsmen on the required R&M design features. J77 75 Programme Managers must ensure that the initial design timescale J77 76 allows the iteration of all designs before the drawings are J77 77 required for manufacture. Appreciable cost savings are made by J77 78 avoiding the re-design activities during the development phase J77 79 which result from designing to tight project timescales in the J77 80 initial design phase.

J77 81 The necessity for R&M specialist knowledge in the initial and J77 82 subsequent design phases is considered essential. Careful personnel J77 83 selection can ensure that R&M specialists provide valuable advice J77 84 to designers. For example, R engineers with an extensive knowledge J77 85 of past reliability problems, and M engineers with first hand J77 86 knowledge of problems unfamiliar to the designer, such as J77 87 maintaining naval aircraft at sea in cramped hangar conditions, J77 88 servicing aircraft on flight decks in adverse sea and weather J77 89 conditions and operating military aircraft from forward bases. J77 90 Thus, the R&M specialists can make important contributions to the J77 91 design process due to their on-the-job experience which the J77 92 designer or draughtsman may not possess. However, it should be J77 93 noted that, whereas maintainability features can be assessed with J77 94 reasonable confidence from drawings, reliability assessment is J77 95 often overturned during development due to unforeseen problems, J77 96 such as higher than expected local vibration levels.

J77 97 At Westland, we have a very active R&M department influencing J77 98 the design process. The result is large numbers of recommendations J77 99 for initial design and subsequent design improvements throughout J77 100 the design, manufacture/build and development phases in order to J77 101 achieve optimum R&M design. This is only to be expected, as any R J77 102 or M engineer worth his salt will strive to introduce the best J77 103 features relating to his own discipline. However, R&M engineers are J77 104 often not concerned with the adverse effects that the J77 105 recommendations have on other design requirements and may be J77 106 unaware of any adverse effects. The aircraft designer is, however, J77 107 concerned with all design parameters and has to reconcile the often J77 108 conflicting requirements of aircraft performance, weight and cost J77 109 as shown in Figure 2.

J77 110 Selecting the maintainability discipline as an example, many J77 111 maintainability features lead to higher acquisition costs and J77 112 increased weight. Whilst the cost aspect is important, an increase J77 113 in weight beyond the aircraft specification dry weight has serious J77 114 consequences to aircraft performance such as reductions in payload, J77 115 range, endurance and time-on-station. Any recommendation, from J77 116 whatever source, which compromises the aircraft specification dry J77 117 weight and performance will not be looked upon favourably. In other J77 118 words, aircraft performance requirements will take precedence over J77 119 other requirements, excluding safety, as any approach to the J77 120 customer on a reduction in aircraft performance would be rejected. J77 121 In the final analysis, the fighting capability of a military J77 122 aircraft or the commercial capability of a civil aircraft will be J77 123 more important.

J77 124 Examples of Weight Increases

J77 125 Examples of recommended maintainability features which have J77 126 weight and cost penalties are as follows:

J77 127 Reposition hydraulic ground servicing points from the cabin J77 128 roof to a position 5 feet from ground level

J77 129 The maintainability advantages of such a feature are J77 130 self-evident. Maintenance personnel would have easy access to the J77 131 servicing points, connection and disconnection of hoses would be J77 132 less risky to personnel, particularly on a ship rolling at sea, and J77 133 the hydraulic hoses would not abrade the aircraft skin. However, J77 134 such an improvement increases the hydraulic pipe lengths and J77 135 introduces an additional panel, both of which increase weight and J77 136 cost. The final judgement, taking into account such features as an J77 137 on-board electro-hydraulic servicing pump, may be to reject the J77 138 proposal.

J77 139 Introduce built-in steps for access to maintenance areas

J77 140 Again, the maintainability advantages are obvious, but built-in J77 141 steps also cause increased weight due to the steps and the required J77 142 strengthening of the surrounding structure. An alternative and much J77 143 lighter option would be to provide hard points on the aircraft for J77 144 GSE ladders to access the maintenance areas. However, this J77 145 alternative is always vetoed by customers on the grounds that GSE J77 146 must be minimised and that the ladder would have to be carried in J77 147 the aircraft, therefore providing no weight advantage. In my view, J77 148 the requirement to carry ladders in the aircraft on all flights is J77 149 questionable. Hard points and lightweight GSE ladders could provide J77 150 greater safety for maintainers when accessing high areas of the J77 151 aircraft and could be provided in more maintenance areas at a J77 152 reduced weight penalty. In any event, if built-in steps are chosen, J77 153 the number and position will be very carefully scrutinised to J77 154 minimise the weight and drag impact.

J77 155 Modify the fuel pump installation to allow the pump to be J77 156 changed without draining the fuel from the tank

J77 157 Modern fuel pump installations provide the capability to change J77 158 the fuel pump without entering the tank. Therefore, the advantage J77 159 of not draining the tank is a reduction in the Time To Repair. J77 160 However, this has to be weighed against the reliability of the pump J77 161 and the time to drain the fuel. If the pump is highly reliable and J77 162 the time to drain the tank is small, the additional weight and cost J77 163 of such an improvement is questionable.

J77 164 Provide a single point opening/closing mechanism on the engine J77 165 and gearbox panels

J77 166 This recommendation reflects the customer's wish to ensure the J77 167 safety of a maintainer when opening and closing panels, J77 168 particularly in adverse weather conditions ashore and at sea. J77 169 However, interconnected latches increase weight and cost and raise J77 170 questions of safety. The loss of a panel, particularly on a J77 171 helicopter, can be catastrophic and any design which interconnects J77 172 panel latches must always incorporate a separate safety latch to J77 173 guard against failure of the mechanism and loss of the panel.

J77 174 Provide sealed bearings in the landing gear to avoid routine J77 175 greasing operations

J77 176 Experience has shown that sealed bearings are not a 'fit and J77 177 forget' maintenance feature when sealed bearings are used in areas J77 178 exposed to the weather. Water ingress into the bearing, J77 179 particularly salt water, causes the grease to deteriorate, damaging J77 180 the bearings and reducing reliability. There have been reports that J77 181 some operators have modified sealed bearings to introduce grease J77 182 nipples.

J77 183 Provide a single point engine compressor washing facility with J77 184 a motorised ball valve operated from the cockpit

J77 185 On a multi-engined aircraft, this recommendation avoids the J77 186 need for maintainers to move the compressor washing rig from one J77 187 wash point to the other and reduces the time to wash engines. J77 188 However, once again the penalties are the additional weight and J77 189 cost of the motorised ball valve, electrical wiring and controls J77 190 and the addition to the overall aircraft failure rate caused by the J77 191 new system.

J77 192 The foregoing are typical examples of maintainability features J77 193 which impact on weight and cost. Other examples are:

J77 194 Fit dummy plug stowages on the main rotor head for the blade J77 195 heater mat plugs.

J77 196 Fit a cross spirit level in lieu of a plumb bob for aircraft J77 197 levelling.

J77 198 Fit a tie-down socket in the main rotor blade for blade J77 199 tethering.

J77 200 Fit lifting lugs to the transmission gearboxes.

J77 201 Lower the engine compressor wash panel to improve access.

J77 202 Introduce a maintenance 'bridge' to traverse the roof deck.

J77 203 Provide an in-flight compressor washing facility.

J77 204 Provide a drain pipe from each engine oil sump.

J77 205 Exclude locking wire from all aspects of design.

J77 206 Introduce quick release/self sealing couplings on hydraulic and J77 207 fuel lines.

J77 208 Introduce remote oil level sensing and replenishment.

J77 209 Introduce Built In Test facilities for all avionic J77 210 equipments.

J77 211 All the examples illustrated are accepted as good features from J77 212 a maintainability point of view, but each of the recommendations J77 213 has to be studied to assess its effect on weight and cost. If the J77 214 features are essential, they will be introduced but non essential J77 215 design features will be modified or rejected, particularly if the J77 216 specification dry weight and the aircraft performance is J77 217 compromised.

J77 218 Summary J77 219 To summarise the points made:

J77 220 R&M requirements must be clearly defined before the initial J77 221 design phase commences. There must be no ambiguity in the stated J77 222 requirements and excessive design features should be avoided.

J77 223 R&M expertise is essential to the designer. Careful personnel J77 224 selection is important to provide on-the-job experience to assist J77 225 designers and draughtsmen.

J77 226 R&M must be an integral part of the design process. It must not J77 227 be regarded as a fringe activity.

J77 228 Project Managers must ensure adequate time for iteration of the J77 229 design in the initial design phase to avoid expensive re-design J77 230 activities during development.

J77 231 R&M specialists must consider the cost and weight implications J77 232 of R&M design improvements. Can the improvements be justified or J77 233 are they excessive for the task? The R&M disciplines lose J77 234 credibility if excessive demands are placed on the aircraft J77 235 designer.

J77 236 When aircraft specification dry weight is exceeded, a J77 237 compromise on all design features is necessary to maintain weight J77 238 and performance levels without adversely affecting aircraft J77 239 safety.

J77 240 J78 1 <#FLOB:J78\>Technology transfer: the BTG model

J78 2 by Peter Tanner

J78 3 Although there may be some who still think of technology J78 4 transfer in terms of assistance to Third World countries, most now J78 5 understand it to mean the process in a developed economy by which a J78 6 business can benefit from technology transferred to it. Even so the J78 7 significance of intellectual property is still not fully J78 8 appreciated and often undervalued, particularly in the engineering J78 9 field. This article draws attention to the opportunities which J78 10 exist, the changes of attitude necessary to take advantage of them J78 11 and some of the ways in which technology transfer can be achieved. J78 12 It is suggested that, as the world's principal exponent of J78 13 technology transfer, the British Technology Group (BTG) is a good J78 14 model for all similar activities in this field.

J78 15 The paradigm shift

J78 16 With a few notable exceptions, for too long technology transfer J78 17 has had too low a profile in the UK and been an undervalued J78 18 resource here. This is not generally so elsewhere. Increasingly, in J78 19 all other industrial countries the awareness is growing of the J78 20 importance and value of intellectual property as a real asset (just J78 21 like real estate) and awareness is also growing of the significance J78 22 to be attached to the transfer of it if it is to be used to the J78 23 maximum commercial advantage. We must ensure that we redress this J78 24 in the UK rather than allowing ourselves to fall further behind. To J78 25 do so, however, will require a radical revision of attitudes, a J78 26 paradigm shift in the current jargon, away from the deeply J78 27 ingrained 'not invented here' culture which many companies still J78 28 find it uncomfortable to make. In the power engineering field this J78 29 is particularly so. Without doubt, however, technology transfer is J78 30 worth taking seriously. The value of trade in IPR (intellectual J78 31 property rights) is considerable, amounting annually to around J78 32 pounds10 billion worldwide, corresponding (at a notional 5% J78 33 royalty) to products valued at pounds200 billion per annum and J78 34 growing rapidly.

J78 35 The whole topic of the exploitation of technology was aired and J78 36 given fresh impetus in the leading article in IEE News J78 37 (December 1990) entitled 'Not invented here or implemented here'. J78 38 As that article notes, "While the British see themselves as J78 39 inventors par excellence, our track record of exploitation is far J78 40 less encouraging". Again, it points out, "Although J78 41 a reputation for inventiveness is exciting, a competence in J78 42 profitable exploitation is more important for maintaining our J78 43 standard of living". We must take heed of this and in doing J78 44 so abandon not only the NIH syndrome but also any vestiges of a J78 45 nationalistic attitude in technology transfer. We need not only to J78 46 exploit our own ideas abroad but also to look abroad for ideas J78 47 which we can manufacture here, always planning for a substantial J78 48 financial benefit from the flow in either direction.

J78 49 Clearly we can ill afford to be parochial about technology J78 50 transfer. In the UK we carry out only 5% of the world's research J78 51 and have only around 5% of the world's industry. It is unlikely, J78 52 therefore, that we will find a close match between our own R&D and J78 53 our own industry. We must be prepared to license our ideas to J78 54 companies overseas (we have, after all, provided over 40% of the J78 55 basic research ideas over the last 30 years) and correspondingly J78 56 look in those countries for ideas which our industry may find J78 57 valuable. This is the true formula for the manufacturing prosperity J78 58 which we must urgently restore, rather than standing by to witness J78 59 its further rapid decline.

J78 60 As Lord Caldecote pointed out in the fifth Lord Nelson of J78 61 Stafford Lecture to the IEE in 1988, the only way the UK will J78 62 achieve its primary objective of providing satisfying lives for all J78 63 will be by the creation of wealth as measured by profitability. J78 64 This, he reminds us, requires the nurturing of a prosperous J78 65 manufacturing industry, the key to which must be greater investment J78 66 in innovation and an increasing commitment to it. He might have J78 67 gone on to add that a highly important and significant element of J78 68 this is the professional handling of intellectual property J78 69 resulting from innovation in the UK and elsewhere.

J78 70 The substantive model

J78 71 The IEE News article referred to earlier points out J78 72 in relation to technology transfer "the need to add a J78 73 missionary zeal to convert the internal sceptics". BTG has J78 74 just such a zeal. Its mission is "The profitable J78 75 exploitation of technology" and consequently as a J78 76 contribution to the debate it might be instructive to spend a J78 77 little time looking at how it operates and how it has firmly J78 78 established itself as a leading technology transfer J78 79 organisation.

J78 80 BTG is currently still a public corporation but well on the way J78 81 to being privatised, which should be achieved before the end of the J78 82 year. To all intents and purposes it has operated as a private J78 83 company for many years. Although when set up as a public J78 84 corporation it borrowed money from the Treasury, the loans were J78 85 repaid long ago and it has been self-financing and J78 86 profitable since the early 1970s.

J78 87 The success achieved by BTG is evidence of the benefits to be J78 88 obtained from the efficient and professional handling of technology J78 89 transfer. During the last financial year BTG increased its total J78 90 revenue from licensing by 30% to well over pounds25 million and J78 91 profit before tax by over 25% to close on pounds10 million. BTG's J78 92 sources of innovation benefited substantially from this success; J78 93 over the year a total of pounds13.1 million was returned to the J78 94 academic sector in revenue sharing, project investment and J78 95 investment in patents for university inventions. The J78 96 professionalism and success of BTG in the handling of technology is J78 97 widely recognised, both in the UK and internationally. As a result J78 98 the number of inventions continues to rise rapidly; last year the J78 99 number offered to BTG increased 50% and accepted for further J78 100 exploitation by 35%.

J78 101 Once an invention is accepted for exploitation BTG, if J78 102 appropriate, takes responsibility for patenting it at its own J78 103 expense, funds further development if this is considered necessary J78 104 to improve the prospects of successful international licensing, J78 105 negotiates licence agreements, shares the licence income with the J78 106 owner of the invention (so that the institution and the inventor J78 107 receive 50% of the proceeds between them) and enforces the patent J78 108 protection if necessary. The acceptance of an invention offered to J78 109 BTG is the result of a favourable assessment of its commercial J78 110 potential by an executive in one of the Operations Divisions which J78 111 are the core of the business. All these executives have not only a J78 112 first class academic background at research and development level J78 113 but also a managerial commercial experience in the particular J78 114 sector of technology for which they are responsible.

J78 115 If an invention is one for which it is thought that useful J78 116 patent protection might be obtained, then through BTG's Patents J78 117 Department and at BTG's expense the patent cover necessary to J78 118 protect it will be provided. The Patents Department is amongst the J78 119 largest in the country, and highly regarded internationally. It is J78 120 certainly a considerable advantage to have a resource of this J78 121 calibre in-house, working closely with executives and inventors to J78 122 produce intellectual property of maximum value.

J78 123 It is of crucial importance not only to have the skills to get J78 124 a proper, well written patent application (leaving the researcher J78 125 thereafter free to publish if he wishes to do so) but also to be J78 126 able to build a stockade of patents round the key idea to achieve J78 127 the maximum protection. These skills have ensured that in many of J78 128 the areas in which the organisation operates BTG has secured such a J78 129 strong patent position for it to be able to take successful patent J78 130 action when necessary against even the most powerful multinationals J78 131 or even governments. Thus major cases have been won involving J78 132 Magnetic Resonance Imaging patents against Johnson and Johnson and J78 133 against the Pentagon in infringement of patents on Hovercraft, to J78 134 name but two. It is not only executives and patent agents working J78 135 together but these collaborating in one team with legal, financial, J78 136 commercial, marketing and other skills which give the organisation J78 137 its unique strength. This extends also to licensing software and J78 138 knowhow which, although generally not protectable by patents, still J78 139 constitute valuable intellectual property.

J78 140 Of course as already mentioned, it is not always the case that J78 141 intellectual property acquired by BTG is in a condition to be J78 142 immediately licensable; it may need further development before it J78 143 is likely to be of commercial interest. BTG has the money to invest J78 144 in such projects and last year this increased by 32% to a total J78 145 project expenditure of nearly pounds10 million. In developing J78 146 inventions BTG now has pounds33 million committed to 480 projects J78 147 as well as investments also in a number of projects with companies. J78 148 In these projects at universities, polytechnics and elsewhere the J78 149 costs supported include materials, components, apparatus and the J78 150 salaries of staff engaged on the project, plus a reasonable J78 151 allocation of overheads. It can also include the cost of market J78 152 research work under subcontract and other activities that enhance J78 153 the commercial value of the product. A feature of a number of these J78 154 projects is to appoint, when appropriate and with the agreement of J78 155 their institutions, selected post-doctoral research staff engaged J78 156 on the project as BTG Fellows.

J78 157 This brings us to the important stage of licensing the J78 158 technology that has been acquired and developed. Clearly this J78 159 requires a detailed knowledge of the industry involved on a J78 160 worldwide basis. Some of this information can be gathered from the J78 161 various databases increasingly becoming available. These are useful J78 162 but no substitute for personal contacts with the principal players J78 163 in the field. Contacts, levels of confidence and trust need to be J78 164 built up over many years so that a direct personal approach can be J78 165 made. All this demands well developed marketing skills, including J78 166 product development, the placing in the marketplace, pricing and J78 167 promotion. These four tenets of marketing are all critical and the J78 168 executive responsible for the licensing must be experienced in them J78 169 all. With the marketing resolved, selling will then be the J78 170 important activity and all will then centre on negotiating an J78 171 appropriate licence agreement.

J78 172 The Licence Agreement, like the Patent Specification, is a key J78 173 document since once the intellectual property has been made J78 174 available to the licensee the continuing payment of royalties and a J78 175 satisfactory ongoing relationship may be dependent on it. In such a J78 176 licence it is usual to have a down payment which has the merit of J78 177 identifying those seriously interested and likely to pursue the J78 178 activity with vigour. In certain circumstances this down payment, J78 179 or at least some part of it, might be treated as an advance on J78 180 royalties.

J78 181 The royalty level itself needs careful negotiation. In many J78 182 areas this works out to be around 5% for patented work although J78 183 this is dependent on the technology, and for software, where J78 184 margins are high, a level of 50% is not uncommon. As well as fixing J78 185 a royalty level it is standard practice also to establish a J78 186 guaranteed annual royalty level and to write this into the J78 187 agreement. This is particularly so where a degree of exclusivity is J78 188 concerned. Exclusivity either to manufacture or sell in certain J78 189 territories is not an uncommon feature, normally only over a J78 190 limited period but long enough to give the licensee a chance to J78 191 become established in the market.

J78 192 As well as having a guaranteed royalty during the exclusive J78 193 period it is also sensible to establish break minima for each year J78 194 so that if this level has not been reached the licensor can J78 195 terminate the agreement. This ensures that licensees do not cocoon J78 196 the work and prevent more serious operators taking a licence.

J78 197 However, it is not intended here to go into all the J78 198 ramifications of the licence deal which could well be the subject J78 199 of a subsequent article, but only to emphasise, by touching on the J78 200 detail involved, the complexity of concluding a suitable J78 201 arrangement and the need for involving expert advice in doing J78 202 so.

J78 203 The success of power

J78 204 Although mention was made at the beginning of this article that J78 205 power engineering is one of the areas where there might be a J78 206 reluctance to become involved with technology transfer, this was J78 207 the engineering area in which BTG achieved an early notable J78 208 success, due to the vision of the late Prof. Gordon Rawcliffe. J78 209 J79 1 <#FLOB:J79\>NON-ISOTHERMAL ADSORPTION IN A PELLET

J79 2 J. H. HILLS

J79 3 Abstract - An analytical solution is presented for the J79 4 temperature rise in an adsorbent pellet in the presence of both J79 5 internal and external mass transfer resistance, and external heat J79 6 transfer resistance. A linear, temperature-independent isotherm and J79 7 constant diffusion coefficient are assumed. Predictions are J79 8 compared with literature data.

J79 9 <}_><-|>INTROUCTION<+|>INTRODUCTION<}/>

J79 10 In the design of packed bed adsorbers both equilibrium data J79 11 (adsorption isotherms) and rate data (effect of internal and J79 12 external heat and mass transfer resistances) are needed to predict J79 13 breakthrough curves. While there are a great deal of data on the J79 14 former, it is much harder to obtain reliable mass transfer data, J79 15 which are consequently less common in the literature. These data J79 16 may be obtained from experimental breakthrough curves, but a number J79 17 of simplifying assumptions must be made whose validity is hard to J79 18 check. A safer method is to subject a small quantity of adsorbent J79 19 to a large excess of fluid containing various fixed concentrations J79 20 of adsorbate, and measure the rate of uptake as a function of J79 21 time.

J79 22 Mass transfer resistance can arise from a number of factors: J79 23 diffusion through the fluid film surrounding the adsorbent pellet, J79 24 transport along the macropores leading from the pellet surface to J79 25 the interior, and transport along the micropores of individual J79 26 crystals. A popular simple method of analysing data is in terms of J79 27 an overall effective diffusion coefficient, obtained from the J79 28 well-known equation for unsteady-state diffusion in a sphere J79 29 (Crank, 1975a) or an overall mass transfer coefficient obtained J79 30 from the simpler linear driving force approximation (Glueckauf and J79 31 Coates, 1947). These methods are only valid when the effective J79 32 intraparticle diffusion coefficient is independent of J79 33 concentration, and when there is no temperature rise in the pellet. J79 34 If either of these constraints is violated, a time-consuming J79 35 numerical integration will be needed, which is particularly onerous J79 36 in the case of temperature effects, since the heat and mass J79 37 balances are strongly coupled. While a temperature rise in the J79 38 fluid is easily detected, it is the solid temperature which is J79 39 important; this is hard to measure, and must normally be inferred J79 40 on theoretical grounds.

J79 41 The most general solution of this problem would allow for both J79 42 concentration and temperature gradients within the adsorbent pellet J79 43 and across the fluid film, for temperature- and J79 44 concentration-dependent diffusion coefficient within the pellet, J79 45 and for a general, non-linear adsorption isotherm, formula. J79 46 The heat of adsorption and other physical properties may also be J79 47 temperature- and concentration-dependent.

J79 48 Such a complex system would require a very long numerical J79 49 procedure to solve it, and the general case would involve so many J79 50 parameters that only particular solutions could be given. Marcussen J79 51 (1982) included most of the elements, and his paper illustrates the J79 52 problems involved. Most other workers introduced considerable J79 53 simplification in order to make the problem tractable.

J79 54 Kondis and Dranoff (1971) used a relatively complete version of J79 55 the system in their study of the adsorption of ethane on a 4 J79 56 <*_>A-circlet<*/> molecular sieve. Their numerical solution J79 57 considered a concentration- and temperature-dependent diffusion J79 58 coefficient in the micropores, and a Langmuir isotherm with J79 59 temperature-dependent equilibrium constant. However, they assumed J79 60 micropore diffusion control and constant temperature throughout the J79 61 pellet. Their predicted temperature variations were much greater J79 62 than those measured, but they point out that this could have been J79 63 due to experimental difficulties.

J79 64 Brunovska et al. (1978) gave numerical solutions for J79 65 the case of negligible external heat and mass transfer resistance, J79 66 internal mass transfer by pore diffusion with a coefficient J79 67 independent of temperature and concentration, and instantaneous J79 68 gas-solid equilibrium at all points within the pellet according to J79 69 a Langmuir isotherm with temperature-dependent coefficients. Later, J79 70 the same workers gave a very restricted set of numerical solutions J79 71 which included an external heat transfer resistance (Brunovska J79 72 et al., 1980), and a simplified version in which internal J79 73 heat transfer resistance was neglected (Brunovska et al., J79 74 1981).

J79 75 Chihara et al. (1976) assumed the isotherm to be J79 76 linear in both concentration and temperature, and represented the J79 77 overall mass tranfer resistance by the linear driving force model J79 78 with a coefficient independent of temperature. With these J79 79 assumptions, they were able to obtain a simple analytical solution J79 80 to the problem which they used to propose a criterion for the J79 81 validity of the isothermal adsoprtion model. However, the linear J79 82 driving force model is only exact when external mass transfer J79 83 controls the rate, or when the concentration profile within the J79 84 particle is parabolic; the latter condition holds approximately at J79 85 long contact times (Rice, 1982), but in most practical cases J79 86 internal mass transfer resistance is important, and J79 87 non-isothermality is most likely at short contact times.

J79 88 Ruthven et al. (1980) studied the case of internal J79 89 mass transfer control with a diffusion coefficient independent of J79 90 temperature and concentration, and an isotherm varying linearly J79 91 with temperature. Their analytical solution is in terms of the sum J79 92 of an infinite series, and while it was developed for the micropore J79 93 diffusion-controlled adsorption in molecular sieves, it can also be J79 94 applied, with minor modifications, to macropore diffusion control, J79 95 as shown later.

J79 96 Bowen and Rimmer (1973) assumed the isotherm to be J79 97 temperature-independent when expressed in terms of relative J79 98 humidity. They used the quadratic driving force approximation for J79 99 mass transfer, with external heat transfer resistance. The J79 100 equations can be solved numerically apart from the very early J79 101 stages, when the rate of adsorption is infinite; in this region, J79 102 Bowen and Rimmer proposed using external mass transfer control, J79 103 which leads to a simple analytical solution.

J79 104 In the present paper the effects of simultaneous internal and J79 105 external mass transfer resistance are combined with external heat J79 106 transfer resistance to give a simplified analytical solution. Both J79 107 the diffusion coefficient and the linear adsorption isotherm are J79 108 assumed independent of temperature, which can be justified if the J79 109 calculated temperature rise is, in fact, small. The solution is J79 110 presented for short adsorption times as a simple graphical J79 111 correlation which can be used to estimate the maximum solid J79 112 temperature rise, and hence to decide whether the data can fairly J79 113 be analysed by an isothermal model.

J79 114 DERIVATION OF THE ESTIMATE

J79 115 The following assumptions are made.

J79 116 (a) Uniform, spherical pellets of adsorbent.

J79 117 (b) Mass transfer resistance can be represented externally by a J79 118 fluid-to-particle mass transfer coefficient, k, and internally J79 119 by a constant effective diffusion coefficient, D.

J79 120 (c) The isotherm is linear with distribution coefficient J79 121 K.

J79 122 (d) There is no internal heat transfer resistance, so the J79 123 pellet is at uniform temperature. Externally, fluid-to-particle J79 124 heat transfer is governed by a heat transfer coefficient, J79 125 h.

J79 126 (e) Bulk fluid concentration cB and temperature J79 127 TB are constant.

J79 128 (f) K, D, k and h are constants, independent of J79 129 the pellet temperature.This assumption, which will only be true for J79 130 very small temperature rises, effectively uncouples the heat and J79 131 mass balances, and means that a standard isothermal mass transfer J79 132 equation can be used, with the temperature rise calculated from the J79 133 rate of adsorption.

J79 134 Crank (1975b) gives the analytical solution for isothermal J79 135 adsorption in a sphere with both internal and external mass J79 136 transfer resistance:

J79 137 formula

J79 138 where formulae, and <*_>beta<*/>n are the J79 139 roots of formula, the equilibrium solid phase concentration J79 140 of adsorbate, is given by formula.

J79 141 Equation (1) is exact, but difficult to use at short exposure J79 142 times (<*_>tau<*/><*_>unch<*/>0) because an excessive number of J79 143 terms are needed for convergence. In this case, an asymptotically J79 144 correct solution can be derived from the solution for penetration J79 145 into a semi-infinite solid, also given by Crank (1975c). For small J79 146 times, adsorbate only penetrates the surface layers of the sphere, J79 147 so the effects of curvature can be neglected.

J79 148 Expressed in terms of <*_>nu<*/> and <*_>tau<*/>, the result J79 149 is:

J79 150 formula.

J79 151 The relative error involved in using eq. (2) rather than (1) is J79 152 less than 1% for formula.

J79 153 The heat balance for a single pellet can be written:

J79 154 formula.

J79 155 Now, writing formula and replacing t by the J79 156 dimensionless <*_>tau<*/>:

J79 157 formula

J79 158 where formula.

J79 159 Differentiating eq. (2) and substituting for J79 160 dq/d<*_>tau<*/> in (3) gives:

J79 161 formula.

J79 162 To integrate eq. (4), use exp (<*_>alpha<*/><*_>tau<*/>) as J79 163 integrating factor, and integrate the right hand side by parts. The J79 164 result is:

J79 165 formula.

J79 166 The remaining integral in eq. (5) can be transformed by writing J79 167 formula. The result is:

J79 168 formula

J79 169 where formula is known as Dawson's integral, and can be J79 170 found in standard tables and computer libraries.

J79 171 A final transformation of eq. (6) can be made by the J79 172 introduction of the dimensionless groups:

J79 173 formulae

J79 174 The resulting expression is:

J79 175 formula.

J79 176 Figure 1 shows a graph of J79 177 <*_>theta<*/>/<*_>theta<*/>h as a function of x J79 178 for various values of the parameter <*_>mu<*/>. The curves pass J79 179 through a maximum at a value of dimensionless time x given by J79 180 the implicit equation:

J79 181 formula.

J79 182 figures&captions

J79 183 The corresponding maximum temperature rise is given by:

J79 184 formula.

J79 185 xm and formula are functions of J79 186 <*_>mu<*/> alone, and are plotted in Fig. 2. Thus the effects of J79 187 internal and external mass transfer resistance on particle J79 188 temperature can be combined in a single parameter <*_>mu<*/>, and J79 189 Fig. 2 can be used directly to estimate the magnitude of any J79 190 temperature rise in the adsorbent. All the parameters involved in J79 191 <*_>theta<*/>h and <*_>mu<*/> can be determined J79 192 experimentally: a, <*_>rho<*/>s and J79 193 Cps can be measured for the solid particles, k J79 194 and h estimated from standard correlations, and K and J79 195 D calculated from the adsorption experiments themselves by J79 196 assuming a negligible temperature rise.

J79 197 DISCUSSION J79 198 Because of experimental difficulties, very few workers have J79 199 actually measured the temperature of the adsorbent during an J79 200 adsorption experiment, and the majority of those who have measured J79 201 it have done so in conditions where there is no external mass J79 202 transfer resistance (pure adsorbent gas). Marcussen (1970) worked J79 203 with water vapour in air adsorbing onto alumina pellets and made J79 204 some temperature measurements, but does not record the results in J79 205 enough detail to check the present theory.

J79 206 Bowen and Rimmer (1973) measured the rates of adsorption and J79 207 the temperature rise when 2.83 mm pellets of activated alumina were J79 208 used to adsorb water vapour from a stream of nitrogen at J79 209 30<*_>degree<*/>C. As explained above, they compared their J79 210 experimental data with the results of a numerical integration using J79 211 the quadratic driving force approximation, but for short exposure J79 212 times they replaced this with external mass transfer control, J79 213 leading to a very simple analytical result which, expressed in our J79 214 nomenclature, is:

J79 215 formulae

J79 216 Comparison with Fig. 2 shows that eq. (10) underestimates J79 217 xm for <*_>mu<*/><20, while eq. (11) grossly J79 218 overestimates <*_>theta<*/>m except at very low values of J79 219 <*_>mu<*/>. Both the simple equations (10) and (11) and the more J79 220 complex (8) and (9) predict xm to be independent of J79 221 gas-phase concentration of adsorbent and <*_>theta<*/>m J79 222 to be proportional to this concentration when K and D are J79 223 concentration-independent.

J79 224 The experimental data presented in Table 1 of Bowen and J79 225 Rimmer's paper can be used to estimate K and D at J79 226 different humidities. Since the temperature measurements (although J79 227 not the mass transfer ones) were made on a single pellet in the J79 228 flowing gas stream, k and h can be estimated from the J79 229 Fr<*_>o-umlaut<*/>ssling equation as given by Rowe et al. J79 230 (1965):

J79 231 formulae

J79 232 which gives formula and formula at a Reynolds J79 233 number of 48.9. (The authors quote Re = 40, but this is J79 234 inconsistent with their other data.)

J79 235 From the literature, formula and formula, and J79 236 in another paper (Rimmer and Bowen, 1972) the authors give a value J79 237 of formula for this experiment.

J79 238 Table 1 presents the results of the calculations for the time, J79 239 tm, at which the peak temperature occurs, and the J79 240 maximum temperature rise, <*_>theta<*/>m. Since the J79 241 isotherm is not completely linear, both K and D vary J79 242 somewhat with the relative humidity of the inlet gas, which leads J79 243 to a small predicted variation of tm as shown in the J79 244 table.

J79 245 Figure 3 compares the measured temperature rise at different J79 246 relative humidities with the predictions of Table 1, and of Bowen J79 247 and Rimmer's simple theory [eq. (11)]. It can be seen that eq. (11) J79 248 gives a large over-prediction in all cases, whereas eq. (9) (Table J79 249 1) gives good predictions for relative humidities up to about 50%. J79 250 For higher RH, the temperature rise exceeds 10<*_>degree<*/>C, J79 251 which almost certainly affects both K and D J79 252 significantly, so the whole basis of the present theory breaks J79 253 down.

J79 254 figure&caption

J79 255 The actual values of <*_>nu<*/> and <*_>tau<*/>m for J79 256 these experiments are such that the use of eq. (2) in place of eq. J79 257 (1) introduces an error of about 1%.

J79 258 J79 259 J80 1 <#FLOB:J80\>Shop Floor Data Captures Top Flight J80 2 Maintenance

J80 3 LAST year international aerospace maintenance engineers FFV J80 4 Aerotech acquired one of the best equipped maintenance facilities J80 5 in the UK at Stansted and set about becoming the first company in J80 6 the industry worldwide to control Job Costing via computerised Shop J80 7 Floor Data Capture (SFCD).

J80 8 The requirements of SFDC for aircraft maintenance are different J80 9 from the standard production control systems in manufacturing. A J80 10 full maintenance overhaul on an aircraft can involve 16,000 J80 11 different jobs with the customer charged on a labour plus parts J80 12 basis.

J80 13 FFV Aerotech employs over 700 permanent staff and 120 J80 14 sub-contractors. The 'Heavy Check' in which the company specialises J80 15 can take between 30,000 and 40,000 man hours at a cost to the J80 16 customer of pounds1.2 million with the labour content accounting J80 17 for 80%.

J80 18 FFV's aim was to pull in quickly accurate records of the time J80 19 spent on each job and thus reduce the 35 days taken between J80 20 completing a maintenance overhaul and invoicing the customer. The J80 21 company's aim was to achieve good financial management, rapid job J80 22 costing and accurate job tracking.

J80 23 Capturing the data on the shop floor was the first key step. J80 24 All job sheets are now bar coded and each engineer has a personal J80 25 bar code number. SFDC terminals manufactured by Source Computer J80 26 Systems collect the data.

J80 27 When each job is issued the engineer takes the job card to the J80 28 terminal and passes a wand over the bar code and over his personal J80 29 bar code. The Source terminal logs the time and the number; when J80 30 the job is finished, the engineer repeats the process and the J80 31 terminal clocks him off. FFV Aerotech thus has an accurate record J80 32 of the time to be booked to each job which can be used for job J80 33 costing.

J80 34 A further benefit is the saving in skilled man hours. With the J80 35 previous manual system booking on and off each job took an average J80 36 of ten seconds per man. The control clerk entered the engineer's J80 37 name and the time against the job on a tabloid size record J80 38 sheet.

J80 39 At shift change, with 120 engineers booking in at the same J80 40 time, delays inevitably mounted. Using the Source terminals it J80 41 takes just three seconds for each man to enter his personal bar J80 42 code and the bar code on his job card before he is contributing to J80 43 FFV's productive man hours.

J80 44 Major advantage

J80 45 FFV chose Source terminals after evaluating the market in J80 46 depth. A major advantage is their flexibility, enabling FFV to J80 47 expend or adapt their use once the system was up and running. J80 48 Secondly, the robust construction enables them to withstand the J80 49 engineering environment of the engine service bays. FFV also wanted J80 50 terminals with the ability to continue collecting data if the main J80 51 frame went down. The Source 400 series terminals can store up to J80 52 2,000 transactions if necessary before they need to be downloaded. J80 53 Source also offer both fixed and hand held versions; in aircraft J80 54 maintenance it is often impractical to wire up terminals on the J80 55 hangar floor as this interferes with the movement of aircraft.

J80 56 The VAS report writing facility generates an initial report on J80 57 the booking of jobs. A second <}_><-|> reports <+|>report<}/> J80 58 throws up errors and anomalies for investigation. Amendments J80 59 necessitated because, for example, an engineer omits to book off J80 60 before leaving the premises, are made by the supervisors of the J80 61 appropriate departments. The system also has the flexibility to J80 62 cope with a group of people working on one job or an engineer J80 63 working on another job while waiting for the issue of spare J80 64 parts.

J80 65 The job record is transmitted to the Accounts Department for J80 66 job costing so that delays between completion of work and invoicing J80 67 are kept to a minimum. Previously, collating, totalling and J80 68 checking this information manually took ten man hours a day. The J80 69 customer also receives a copy of the job record on his aircraft so J80 70 that he has a detailed, factual analysis of the time spent and the J80 71 work carried out. If necessary, daily reports can be produced for J80 72 the customer and the work signed off as it progresses.

J80 73 In the longer term the data gathered for the shop floor will J80 74 prove an important marketing tool in improving estimates, J80 75 turnaround times and customer service levels, says group project J80 76 manager Mike Gray.

J80 77 At present there is no detailed record of the work content J80 78 required on an ageing jet aircraft. The production of a labour and J80 79 parts data base will enable FFV to estimate the cost and turnaround J80 80 time of heavy maintenance checks. "Such a system will be J80 81 invaluable in optimising the productivity of our maintenance J80 82 facilities and give us the information we need to offer fixed price J80 83 quotations," explains Mike Gray.

J80 84 J80 85 Plastic Media Paint Stripping in Aircraft J80 86 Maintenance

J80 87 THE world is becoming so environmentally conscious that any J80 88 technique that eliminates or reduces pollution is being strictly J80 89 monitored. Recent trials indicate that plastic media paint J80 90 stripping may represent one such 'green' technique.

J80 91 All aircraft, civil or military, fixed wing or rotary have to J80 92 be painted several times in their lives to enhance their appearance J80 93 and invariably, to update their livery. Typically, an aircraft is J80 94 paint stripped and then repainted, as several coats of paint may J80 95 mean reduced revenue because of weight limitations; particularly J80 96 relevant in the case of larger, commercial aircraft.

J80 97 Traditionally, the most commonly practicised technique for J80 98 paint stripping of aircraft has been chemical stripping, J80 99 predominantly used methylene chloride. Apart from the problems of J80 100 safe disposal of materials, operator working conditions are J80 101 hazardous and the overall technique is costly. An alternative J80 102 method of shot blasting the surface has also been used and although J80 103 paint stripping is achieved, the substrate is invariably J80 104 damaged.

J80 105 In the last few years, in the light of the problems associated J80 106 with the above techniques, many aerospace repair and maintenance J80 107 organisations have conducted trials on a new technique of plastic J80 108 media paint stripping which in environmental terms is non-hazardous J80 109 and if conducted correctly, achieves removal of paint coatings J80 110 without damaging the substrate material beneath.

J80 111 Although several aerospace organisations have given approval J80 112 for the use of plastic media, trials are still in progress to J80 113 clearly define the advantages and disadvantages of this technique J80 114 over traditional methods.

J80 115 Several types of plastic media are available, generally made of J80 116 a thermoset polyester, urea formaldehyde or melamine formaldehyde. J80 117 Plastic media vary in size from 80-8 mesh (approximately 0.13 - 2.4 J80 118 mm in size) and in hardness from 3 - 4 on the mohs scale. As a J80 119 general guide, harder plastic media are faster cutting and are more J80 120 suitable for hard surfaces and tenacious coatings, whilst software J80 121 plastic media are used on more delicate parts. Plastic media also J80 122 vary in terms of breakdown. Those plastic media with limited J80 123 breakdown rondate in operation, thus losing their cutting action J80 124 and developing abrasive peening effects.

J80 125 Several distinctive advances over traditional J80 126 techniques

J80 127 Plastic media paint stripping has several distinctive J80 128 advantages over traditional techniques. It is non-hazardous in J80 129 environmental terms. Dependent on the specific application, plastic J80 130 media paint stripping is between 2 - 10 times cheaper than chemical J80 131 stripping. Unlike chemical stripping, with selected plastic media, J80 132 it is possible to remove paint coat by coat, without damaging the J80 133 substrate material beneath.

J80 134 Plastic media can be used on metallic and composite surfaces J80 135 which is contrary to chemical stripping which is limited to J80 136 metallic surfaces only. To date, plastic media have been used J80 137 successfully to strip graphite/epoxy, kevlor and J80 138 fibre-glass composite surfaces. Presently, aerospace repair J80 139 and maintenance organisations are only permitted to remove paint J80 140 from composite surfaces using mechanical means such as sanding. J80 141 Although sanding has been successful over many years, it is a time J80 142 consuming technique and one in which damage to the substrate may J80 143 easily occur.

J80 144 Controlled, dry plastic media paint stripping is best achieved J80 145 manually, using skilled personnel as coat thicknesses vary J80 146 considerably and it is essential not to dwell on the surface and to J80 147 aim at multiple passes rather than a continued attack on one area. J80 148 Excessive dwelling results in overheating, with the risk of warpage J80 149 and even material loss.

J80 150 As most aircraft structures and therefore assemblies are J80 151 mixtures of metallic and composite materials which always vary in J80 152 thickness, it is essential prior to stripping that an appropriate J80 153 stripping technique is designated for each area.

J80 154 The equipment used for plastic media paint stripping must J80 155 satisfy the separation requirements necessary to ensure safe J80 156 operator working conditions. Thus, it is necessary to identify what J80 157 materials are being removed, how they are likely to appear as a mix J80 158 in the plastic media and whether they can be removed prior to J80 159 re-blasting by air wash, sieve or mechanical separator.

J80 160 The equipment must control delivery sufficiently to enable very J80 161 low air pressures to be achieved at consistently high feed rates. J80 162 Consistency here is of paramount importance as small variations of J80 163 pressure and feed rate will vary particle velocity considerably J80 164 which may result in substantial damage, if delicate structures are J80 165 being treated.

J80 166 Metal Improvement Company have been active for over 40 years in J80 167 the standard aircraft maintenance operation of controlled shot J80 168 peening and shot blasting of aerospace structures to remove J80 169 corrosion, expose exfoliation corrosion, reduce fretting fatigue J80 170 and restore fatigue life and stress corrosion resistance. They now J80 171 include plastic media paint stripping in the services they offer on J80 172 site, or at least 3 U.K. facilities. Metal Improvement Company are J80 173 an international corporation with 34 divisions worldwide. They J80 174 believe that plastic media paint stripping can achieve excellent J80 175 results providing caution, care and control are exercised in the J80 176 selection of both the type of plastic media and the equipment to be J80 177 used for each specific allocation.

J80 178 J80 179 Kardex Carousel Storage Keeps British Airways Aircraft on J80 180 the Move

J80 181 REVIEWING spares storage and maintenance supply systems for J80 182 their multi-million pound aircraft fleets at Heathrow, J80 183 British Airways with their reputation for efficiency, could not J80 184 afford to have their heads in the clouds!

J80 185 Moves from multiple locations to a more central, but smaller J80 186 Engineering Stores facility at Heathrow created the potential for J80 187 greater operational efficiency, but presented physical problems of J80 188 storage capacity.

J80 189 While high density automated storage and retrieval held the J80 190 promise of solving the equation, the company were aware that J80 191 success or failure would turn on system reliability.

J80 192 Aircraft service schedules, being pre-planned, demand J80 193 instant spare parts availability, with aircraft allocated specific J80 194 time slots, location and positions in the workshop.

J80 195 Delays in supplying spares for aircraft servicing and J80 196 maintenance, which may hold up an aircraft being released into J80 197 service are costly. Additionally, there is a domino effect on later J80 198 scheduling.

J80 199 Absolute reliability of any automated storage system was, J80 200 therefore, crucial and the lynch pin of success.

J80 201 Today, 10 Industriever rotary carousel automated storage and J80 202 retrieval systems from Kardex Systems (UK) Limited are holding a J80 203 wide variety of spares for the British Airways fleets of 34, 1-11s; J80 204 43, 737's; 17 Tristars and 7 A.320 Airbuses.

J80 205 As part of Phase 1 of the installation, four Industrievers were J80 206 installed in June, 1989, with a further three in Phase 2, followed J80 207 shortly after by three more in Phase 3.

J80 208 The interest of British Airways in rotary carousel storage J80 209 technology was first stirred in 1986 when methods of concentrating J80 210 aircraft spares into a more centralised location were examined. At J80 211 this time, parts were held both at Heathrow and Gatwick, entailing J80 212 considerable transhipment. Inside large stores areas, conventional J80 213 racking was employed both at floor level and on mezzanines. The J80 214 system suffered from demanding a high level of physical effort, J80 215 much walking to and from different locations with the bending, J80 216 stooping and reaching associated with conventional racks and J80 217 shelves.

J80 218 Appraisal J80 219 As part of their appraisal, managers from BA conducted J80 220 extensive studies into the concept of rotary carousel systems and J80 221 the relative qualities of specific manufacturers. Several J80 222 Industriever sites were visited, to examine systems in operation, J80 223 including installations in other aviation-related engineering J80 224 establishments. A visit to Hanover Fair, Europe's shop window for J80 225 industrial goods, was followed by a visit to Kardex's manufacturing J80 226 facility at Belheim in West Germany.

J80 227 Feasibility studies established essential criteria and examined J80 228 the capability of carousel systems to meet them; for example, J80 229 bottlenecks, hold-ups at access points were not acceptable. Picking J80 230 and replenishment rates were established and the new installation J80 231 planned jointly by BA managers and Kardex personnel.

J80 232