[Paper for International Congress of Anthropological and Ethnological Sciences. Vancouver. 1983]

Language cannot be arbitrary, either in the syntactic processes or in the forms of individual words observed in different languages. Because language is not arbitrary, there can be an inherent predisposition to learn some one language out of a wide range of particular languages (selection from a range of possible lexicons and possible syntaxes) in very much the same way as a bird is innately programmed to learn the particular song of its species (if it is exposed to that song) or a cat is innately prepared, through the structuring of its visual system at birth, to perceive certain visual patternings in its environment (if it is exposed to patternings of that kind). The process by which a particular language is selected and learnt is completely analogous to the plasticity in a critical period or imprinting found in many species of animals: that is, a readiness during a critical learning period for neurological structure to be modified and permanently 'set' by the characteristics of the environment encountered by the young creature. The array of words which form the lexicon of any language and of the processes which constitute its syntax derive from the complete and persisting integration of the neurological and physiological systems underlying language with those underlying visual perception (and other forms of perception) and bodily action. Differences in lexicon and syntax between languages are comparable to differences between bird-song in different species and differences in behavioural patterns between creatures of the same species in different environmental settings. Between human individuals, there can be wide differences in cognitive strategies and specifically in modes of perception or action and these differences between individuals, generalised to a language community, provide the basis for differences between the lexicons and syntaxes of different languages. In every human being, language, visual perception and bodily action are and must be fully integrated for effective behaviour (interaction with the external world, physical and social) but this integration can be achieved in different ways in different language-communities.

To arrive at plausible views on the origin and mode of development of language in the course of human evolution, we can only proceed from what we know, that is from our present knowledge of language, the researches of comparative linguistics, the study of the acquisition of language by children, the study of the capacity of sub-human creatures to communicate with each other in ways resembling (at some distance) human language, the results of research into the physiological and neurological bases of speech and its comprehension, the palaeontological and anthropological evidence for general human evolution (the evolution of mentality as well as of bodily skill). Clearly we shall never be able to say with much certainty that human language originated in precisely this way, in these stages, any more than we can say that the body and brain of man evolved from this precursor in precisely these stages.

But our problems in discussing the origin and development of language are much greater than those encountered in most other fields of scientific research. Before speculating about the origin of language we ought to recognise the huge gaps that exist in our understanding of language as we use it now (quite apart from the special complexities of the process by which children acquire the language of their community). Unless we attempt to have a clear understanding of what language is now - and recognise where the difficulties lie - we shall be scratching around in the dark in attempting to discuss its origin and development. We have no prehistoric language fossils and the records of written language that have survived are relatively so recent that any structure of theory built on them is bound to be unacceptably speculative.

If we could look to a science of linguistics to tell us how language functions, we should be better placed to talk about language origins and development. But at this point one must recognise what the state of affairs in linguistics now is. The comments made here may be contentious and almost certainly will be resented and dismissed by those professionally committed to linguistics but a negative judgment of the worth of linguistics is justified, even though the terms used may seem severe. Linguistics, as a modern science, has failed in its objective of understanding the functioning of language, and this failure extends both to syntax and to lexicon. If we are profitably to explore the origins and development of language, we need a new linguistics as our starting point. There is no use in searching for the prehistoric antecedents of the unclear and unsettled account of language given by modern linguistics. Chomsky led the study of syntax into a quagmire from which it has not yet emerged; the lexicon and semantics of language were for many years neglected and damagingly separated from the study of syntax, to which logically they should be prior.

The new linguistics must be founded on quite different principles from those of the transformational-generative school and its successors, and must move in a quite different direction. The new principles might be these:

1. Linguistics must be integrated with biology, physiology, neurology and psychology. It must become a real part of the biological and behavioural sciences. The evolution of language, as well as the functioning of language, must be treated as an integral aspect of general human evolution.

2. Nothing in the experience of language should be treated as chance, arbitrary or 'purely cultural'. Language is an instrument which has undergone the same pressures and served the same objectives of human adaptation as other aspects of human behaviour and human physical abilities.

3. The fact that there are different languages, with different words and different syntactic processes is not to be treated as a boundary which linguistics must not cross but as the starting point for research into the underlying processes by which the different languages achieve the same end. Linguistics should again become a genuinely comparative science.

4. The basic task of the new linguistics should be to give a verified account of how language, a modulated stream of sound, communicates experience from one human to another, from one brain to another. The essential concern should be not syntax, word-formation or phonology but the transmission of meaning. The old linguistics of Chomsky in evading meaning as a problem and at the same time treating individual words, the stuff of language, as wholly arbitrary, destroyed any chance that it might have of achieving valuable results.

5. Because those professionally committed to linguistics for the most part deny themselves the larger outlook, it seems inevitable that the biologists, the anthropologists, the neurologists and the experts in artificial intelligence must take over the problem of the functioning of language. Language, one might say, is too important a subject to be left any longer to the linguists. Language will have to be tackled by the methods of the biological and physical sciences. We must be ready to form hypotheses about its mode of functioning and to test their implications to destruction or confirmation.

It may help to give some direction to discussion in this symposium if we consider what it is we hope to find by elaborating theories about the origin and development of language. What surely must drive us is our wish to understand better what language is now and so what we are, as predominantly language-animals. A theory of the origin of language which fits on to and explains how language has come to function as it does could now give us 'power' in the same way as a successful scientific theory can give power. From the tested hypotheses of science come new perceptions, new directions of enquiry, new links between separated areas of enquiry. But what should we look for from our hypotheses on the origin and development of language? At the most banal level, we get a picture, perhaps a fantasy picture (like a good deal of historical study) of the harmonious development of mankind with all man's nascent abilities for language, for technology, for socialisation and so on, elegantly integrated with each other. What satisfying or superficially satisfying answers can we find to the question: What for? in relation to the study of the origin and development of language? Possibly:

- so that we come to understand better the nature of man - to increase our effectiveness in the present: to give new life to the current study of language, to allow us to understand and in some sense master human language, as physics has allowed us to master electricity, flight etc. etc. - to satisfy our curiosity about ourselves - to make progress in our study of man as a whole, to separate out the operation of language from other aspects of the human mind, to come with new eyes to research into brain-function, visual perception, the organisation of human action - to help us in tackling the pressing problems of human society, since language is a major component in the socialisation of man and in the functioning of any society

There may be other more satisfying answers but what seems clear is that it is hard to maintain the drive towards investigating the origin and development of language merely as a historical or prehistorical study. What gives life to it is the link between the distant past and the powerful present. But even if we find our motive for speculating about the origin of language in some such terms as those listed above, the inadequacies of the account of language given by modern linguistics force us to attempt some clarification of our own views about language as it at present exists before we speak about language origins. There are obvious points that can be made:

• Every human has a language and it is a spoken language, that is, it consists in a varying pattern of sounds made by a stream of air flowing from the lungs through the mouth.
• The patternings of the sounds thus made are formed into a large number of different systems (the many distinct languages of the world).
• The significance of language as a system of sounds is that it can be transmitted across space and the patterning can be received, decoded, by other human beings belonging to the same speech-community.
• The value of language is that the whole range of physical and mental experience of a human being can be transduced into the pattern of transmitted sound. The experience of one human being can become operative in the functioning of another human being, receiving and interpreting the pattern of sounds.
• Language is the medium for brain to brain contact between human beings and for the preservation (through writing, recording etc.) of previous patterns of brain organisation and so of human experience.
• The elements of language are matched specifically to shared aspects of experience of the world of the speech-community. What is heard in speech can be related to what is seen in the shared world, and to what is felt in the internal experience of the individual.

All of this is of course totally familiar as an account of what language is and what language does but at the same time every item in the list is something of a mystery. These commonplace operations of language are unexplained and no outline of any satisfactory explanation has as yet emerged. To list these features is to give a very inadequate account of language as a huge present fact. It is far too general and complacent. What we need to say is how language presently achieves what it does. How do I understand what you say? or write? There are large problems in explaining how I hear as language the sounds you make, never mind whether I understand what it is you are saying. At least with written language, a good part of the intermediate processing has already been done, the multi-channel complexity of speech has been reduced to the single channel of the visual patterning of the writing, the words have been separated and formed into groups, sentences and paragraphs. To understand how we extract the meaning from written language ought to be simpler than to understand how we decode spoken language (except of course that written language introduces the complexity of the relation between the visual pattern of the letter or word and its conversion into an internally-recognised speech-sound pattern).

If it is in some respects easier to interpret written language than spoken language, let us start by a small mental experiment. If I write on a piece of paper, which I hand to you, 'Give me the cup' , how exactly do you proceed in interpreting this, and then acting or not acting on what is written? In brief, how do you know what each word (GIVE ME THE CUP) separately means and then combine the meanings of these words into an instruction for an action by you ie. into a proposed pattern of action which you are being urged to make actual? The combination of the words into an instruction involves a second stage of complexity beyond the understanding of each individual word: at this point it will be enough to attempt to deal with the first stage, the four individual words, the lexical aspect.

To simplify: suppose initially that there are several pieces of paper on which each word is written separately. If I show you the piece on which GIVE is written, what operation do you perform, what change does this produce in you? (This rather resembles Miller and Johnson-Laird's attempt to explain how LAMP means 'lamp'). But, you say, 'I know what GIVE means: it means Give' - some idea of me having something which I hold, stretch out towards you so that you in your turn hold it, and I release my hold on it' (and internally I may count the thing given as no longer mine but yours). The dictionary certainly does not help in any way to understand what GIVE means: it says that GIVE means 'to bestow, impart, yield, grant, donate, permit, afford, furnish, pay or render, allow or admit, grow soft, begin to melt, open'. All these meanings of GIVE are much less clear than GIVE itself and a great deal more knowledge is required to judge when GIVE means one of these rather than just GIVE.

So there is a mystery of how we know what GIVE means. It is probably one of the earliest words we use as children. No one then told us that GIVE meant 'bestow' or 'impart'; no one even told us that GIVE meant 'Hold something, then stretch out, let you grasp it, then release it so that now you hold it and it is yours. All we had as children was the word and the action - but how did we know that this action or succession of actions was what GIVE means? If someone said 'Give me the towel', how did we know what aspect of the situation GIVE referred to. Yet we learnt to use the word GIVE very easily and frequently, quite correctly.

There is then this puzzle at the very beginning of language. Would any noise (word, pattern of speech-sounds) have done equally well with GIVE to mean the action involved in giving? If I now said 'Carunk me the cup', could you immediately deduce that this was just another word for GIVE? Might it mean BREAK or FILL or EMPTY or WASH or something quite other such as 'Decorate the cup with a thin gold line'? We obviously cannot take our grasp of the lexicon, this very small part of the lexicon, for granted.

So far we have found great difficulty with understanding one simple word GIVE. 'Give' is such a simple action that there can be no complication about how this pattern of sound relates to this pattern of bodily action. Now if I show you the piece of paper on which ME is written, what reaction does this produce in you? How is your functioning changed? A piece of paper with ME on it: I hold it and hand it to you. ME is very familiar to you except that for you, it means you yourself, NOT 'me', the person handing it to you! Some children, autistic children for example, have difficulty in realising that I or ME refers to themselves when they use the words whilst the words are also used by others to refer to the other selves. Great intellectual ability is required to switch from my point of view to your point of view, so that I can understand that ME or I when used by you refers to you. What then does ME mean? How do I know what ME means? I feel that ME is some sensation in my own body, perhaps in my chest, mouth or forehead, somewhere in here. What mother ever explained to a child how it should use the word ME correctly? 'Don't say - GIVE YOU THE CUP - Johnny! Say: GIVE ME THE CUP'. It becomes totally impossible. The child would have to be a young Bertrand Russell. One cannot explain what ME means; what ME means can only be felt in response to the sound (and beyond that a sophisticated ability to swap one's own and another's point of view has to develop). The dictionary for ME has: 'Accusative and Dative of I' - which does not help in any way, simply referring you to the equal mystery of what 'I' means. For 'I', the dictionary has: 'The nominative singular of the first personal pronoun' (which is only an account of the grammatical classification of the word) and 'The word used in referring to oneself' - but what, Johnny asks, is 'oneself'?

The difficulty seems no less if I show you the scrap of paper with THE written on it. How do you react to that? What change in your functioning or organisation occurs when you see the word THE in isolation? What does THE mean to a Russian or the speaker of another language which has no equivalent for THE? Faced with THE by itself, one's reaction is complete puzzlement. You know the word but can say no more about it than that THE means 'the'. Perhaps the word vaguely predisposes you for something else: maybe it predisposes you for reference to a thing, but of course it might be THE moon, THE sky, THE song, THE word. If we could explain the physiological and neurological impact of the recognition of THE, then we should have advanced a great deal further in understanding how language, and particularly how elements in the lexicon, function. The dictionary says of THE: 'Called the definite article, used to denote a particular person or thing' though whether 'definite article' is a particular person or thing seems doubtful, to say the least, and what in any case does 'A' mean or PARTICULAR or THING?

Finally I show you a piece of paper on which CUP is written. Probably this produces the most definite reaction in you of any of the words shown. CUP: you see 'in your mind's eye' a familiar object, probably white, rounded, concave, possibly with a handle (if you are not Chinese!).You have a feeling of something pretty specific, though you are ready to be shown a small brown coffee-cup or a large yellow breakfast cup. What is a CUP? What then does CUP mean? It is certainly less precise than, say, CROSS or CIRCLE. CUP seems to be a whole family of possibilities, a container of liquid, though perhaps a cup of sugar, a container of liquid but not a kettle, a soup-plate, a vase, a teapot. A smallish object - do you think of an egg-cup on hearing CUP? The dictionary offers something apparently more precise: 'A drinking-vessel usual hemispherical; an ornamental vessel offered as a prize; a hollow'. From the dictionary definition, could you correctly identify a 'cup' if you understood the words of the definition but did not know the word CUP? Is a glass for beer a cup? or a glass for wine? Is a saucer that the cat drinks from? Is a mug a cup and if not, how do you distinguish a MUG from a CUP?

In the case of all these 'difficult words'- CUP ME THE GIVE - one feels inclined to say: 'Don't tell me what each word means. I know' but how do you know? Only one of the words normally refers to a physical object of which one could say: 'That is (one sample of what we call) a CUP'. Even to understand CUP, we need to make use of the object, to manipulate it, to see other examples of CUP. If you are shown only one CUP, the word CUP might be a proper name like Fido. How do we generalise CUP to use the word in referring to a whole range of objects which broadly share certain physical properties or are normally found in certain circumstances such as a tea-table or in a china cabinet?

If we cannot learn the meaning of a word by using other words (and this seems to be the case for the words we most commonly use) and if for many or most words there is no possibility of learning by ostension, by pointing to what the word means (eg. point to THE), how do we come to know what words mean, how each word should 'properly' be used? How incidentally do I know how to use a word such as HOW? Our understanding of the functioning of language, and more specifically of our mastery of the lexicon, the individual words, thus falls at the first hurdle. If we cannot understand how we know and make use of the meanings of individual words, how can we tackle the much greater problem of how we grasp the precise meaning of combinations of these strange things, words? How on earth do we understand what it means to say 'The implications of monetary policy for the functioning of the economy in a period of cyclical depression are radically different from those in conditions of steady growth and near-full employment' ? Explain how we are able to understand 'Give me the cup' and we might then be able to go on to explain the harder sentence.

The total set of words going to form a language and each individual word in the lexicon as has just been discussed constitute the major problem in getting a clear idea of how language functions. The terms 'microstructure' and 'macrostructure' are introduced here as a way of organising the discussion of the problems that have been seen to exist. The microstructure of the lexicon refers to the basis of the patterning of each individual word; the macrostructure of the lexicon refers to the resemblances, uniformities and tendencies to system and grouping that one can observe in the lexicon as a whole. This paper concentrates on the microstructure but it is in the macrostructure of the lexicon that one must look for experimental verification of what is proposed.

It is an important, and indeed essential, step before we attempt to frame hypotheses about the origin and development of language to provide a plausible account of how we understand, or come to understand, individual words in the lexicon. Understanding a word means linking it with something else in our experience eg. as just discussed, with the pattern of action in GIVING, the visual perception of a CUP and perhaps the action of drinking from it, the awareness of myself in ME, the peculiar functioning of THE when it used along with other words. So far we have found no way of explaining how we know, now and without hesitation, what CUP or GIVE means. But maybe this is because we have ignored, taken as totally unimportant, the most obvious feature of the two words: each of the words is formed of a distinct pattern of speech-sounds. Is the formation of the word GIVE from the sounds G I V totally insignificant and accidental? Could any other collection of sounds have been used, such as X A W ? If GIVE clearly seems to refer to a broad pattern of bodily action (movement of the hand and arm in relation to the body), so also the enunciation of the sounds G I V involves a distinct pattern of bodily action for each sound as it is articulated and a more complex pattern as the three sounds are formed into the one word. Why should there not be a relation between the pattern of action in articulating G I V and the pattern of action in stretching out and handing something to somebody? There seems no reason to think that the action of giving involves much more complex muscular activity than the action of saying GIVE. It would seem physiologically uneconomical in the single human organism to operate with two completely unrelated patterns of muscular activity when the need is to link the word GIVE as closely as possible with the action GIVE, which is what would be involved if the articulation GIVE had no structural similarity to the bodily action of giving.

Assume the contrary as more probable: that the patterns of muscular activity for the word and for the action are related. Immediately a whole new area of possibilities for understanding the relation between words and what they mean opens up. If every word (or at least every 'primary' word not definable in terms of other words) contains in its structure a clue to its meaning, then the acquisition of language by children becomes a much simpler and more comprehensible matter and the rapidity and assurance with which we understand the meaning of any individual word which we hear or read also becomes comprehensible. Each word as we hear it would induce in us a muscular patterning which structurally resembles in some way an action or a perception to which the word refers. In the same way as we feel our arm move in sympathy with the arm of the tennis-player preparing to serve, so as we hear a word there would be a preliminary organisation of our own muscles in a way corresponding to the muscular activity of the speaker. The speaker, on this view, has, through the speech-code, transferred to us part of his own immediate physiological and neurological structuring. Of course, apart from language, we are all the time accepting transfers of this kind, both when we obviously imitate what another is doing but also when we respond in our own expression to the facial expression of another, when we smile or yawn when we see another smile or yawn. We understand others, one might propose, by momentarily altering ourselves to resemble them. Understanding language would then be seen as a rather specialised form of this empathy or unconscious mimesis. We would know what the other meant because momentarily we would feel what the other was feeling as he formulates the words he speaks to us.

All very well, one may say, as an interesting hypothesis but what does it imply in terms of the underlying relation of speech, action and perception and how could such a hypothesis be given a specific content? Take first the question of specific content. If in some way the patterning in the articulation of the word GIVE resembles the patterning in the action of GIVING, note that the word GIVE is formed from three sound-elements, each of which may take part in the formation of many other words. Similarly, the action of giving can be analysed into a number of separate elements, each of which can take part in many other types of bodily action. One possible approach is to equate the elements of articulation in the word GIVE with the elements of action in GIVING and see whether one thus arrives at detailed equivalences of speech-sound elements and action-elements which could form part of a larger system, in the sense that in forming other words the same equivalences would be found. If for example the articulation of G involves a raising of the hand and arm to a particular position, the articulation of I involves some direction and duration of forward movement and the articulation of V involves some mode of termination of the action, then one can start to establish testable correspondences.

Obviously, many, particularly linguists but no doubt also experts in other disciplines, will feel great reluctance to contemplate this kind of approach to the understanding of words and the functioning of language. So long as it remains a general proposal, the approach can be met with equally general arguments about its plausibility. Only if the hypothesis can be made more specific will it deserve, and perhaps attract, closer examination. It needs to be made more specific in two ways: first, a sharper account of how there can be a relation or parallelism between articulation and the systems of perception and bodily action (which are the source of the real experience to which language refers) and secondly, some detailed specification of the possible relation between individual speech-sounds and individual elements of action or perception to which particular words refer.

The propositions advanced to explain the parallelism between articulation and action are these:

• Every physical action involves some previous mental (brain) patterning, the formation of a program to be executed by that part of the muscular system appropriate to the action
• Every program of action is formulated against the background of the persisting state of bodily organisation of the individual. A particular action is a modification of the current postural system (which is itself actively maintained); it will require compensating changes in the postural system eg. if we lift a heavy weight, there have to be compensating adjustments in the muscular system to maintain balance and control
• A program of action can be formulated in general terms and executed in a wide variety of ways allowing for differences in the external situation eg. variations in the weight of the object to be raised, in its position and so on. We can execute an action program with the left hand rather than the right, we can divert a program of action normally implemented through one set of muscles to a different set of muscles
• In formulating a particular action program and executing it, because the postural system is a unity, we can get the phenomenon of sympathetic, linked or parallel action. At the simplest level, a particular action may produce with it a grunt, a cry, a movement of the tongue (as Darwin observed of children attempting to write)
• Every distinct speech-sound as a neuromuscular pattern (normally expressed in articulation) can be expressed in muscular patterning other than the articulatory patterning. In particular, it can, by a shift of attention, constitute a muscular pattern for controlling the position and movement of the hand and arm.
• Every element in the movement and positioning of the hand and arm (that is every element in the neuromuscular program for such movement) can be transferred to the muscular system involved in articulation.
• Given the range of neuromuscular patterns for the range of distinct speech-sounds in a language, there is a corresponding range of hand and arm positions and movements.
• Without speech, we can convey a meaning to another by moving our hand or arm, either by imitating a particular action (or in fact performing the particular action) such as HITTING, GIVING, HOLDING or by using a movement of the hand and arm to draw in the air a visual contour, a picture of what we want to refer to eg. a circling movement to represent something circular, or we can indicate our meaning by pointing to an object, perhaps some part of ourselves, our mouth if we are hungry or thirsty, our eye or ear to show that we can see or hear something and so on. Each of these actions or imitations of actions involves a particular neuromuscular patterning which can be transferred to any other set of joints and muscles in the body which is sufficiently complex to allow the muscular program to be executed (as many have pointed out, you can make an attempt to write your name without using your hand, using your elbow or your foot).
• With a shift of attention, ie. a redirection of the program of action, to the mouth and throat, and given a stream of air on which changes in the positions and movements of the articulatory muscles and cartilages can operate, the step between action and language is crossed. The word initially is the by-product of the action because it transduces the action into a patterning of the muscles of the chest, throat, mouth and tongue. Modulating the stream of air, the action-pattern produces a succession of sounds structurally related to the original action-patterning.
• Correspondingly, the action-program involved in the production of a certain pattern of speech-sound can be transferred (overspill perhaps) into patterning controlling the position and movement of the hand and arm and cam manifest itself as gesture accompanying speech.

The above propositions constitute a first attempt to set out how there can be a direct relation or parallelism between articulation and the systems of perception and bodily action (the contents of the real experience of the individual). The other way in which the hypothesis presented in this paper on the relation between the structures of words (the microstructure of the lexicon) and the structure of the actions and percepts to which words refer can be made more specific is by presenting a more definite proposal for the possible relation between individual speech-sounds and individual elements of action or perception. This more definite proposal takes the form of the following Table of correlations between speech-sound elements and action elements, in terms of the positions and movements of the hand and arm.

The Table presents the correlations between speech-sounds and arm movements and positions in a very compressed way. Some explanation is required. Taking the first row in the Table, the Vowel Line, this comprises the principal short vowels. As will be seen, the arm positions from left to right match the vowels AEIOU. The vowels correspond to a range of positions - not movements - of the arm (directed forwards). The vowel A as in 'PAT' corresponds to the lowest position, E to the next higher position and so on, with U with the highest position of the short vowels. The long vowels correspond to systematically higher positions with long U (UU) as in 'POOL' at the highest position of all.

The second row in the Table is for what is called the Projective group. Though at first sight the elements in this group seem to resemble those in the Vowel group, there is an important difference. Whilst the Vowel Group is composed of varying positions of the arm, in the Projective Group what is associated with each speech-sound is a (ballistic) forward movement of the arm (hence the name Projective). There are seven levels at which the forward movement occurs; at the lowest level there is V and at the highest level Y. Each element in the group contributes to any word in which it appears the same type of vigorous forward movement. This can be most easily observed in words where the first speech-sound is an element belonging to the Projective Group: words such as VAULT VENT CHARGE CHUCK JET JUT JUMP POINT PUSH TAP TOUCH THROW THRUST YANK.

The third row contains elements belonging to what is called the Main Consonantal Line. Elements in this group are associated with positions of the bent arm in front of the body, at successively higher levels. The first element B is at the lowest level, in front of the chest, and the last element H is at the highest level;, with the hand above the head. Elements in this group, unlike those in the Projective Group, relate to positions and not movements of the arm. It is more difficult to list words directly illustrative of the effects of elements in this group but there are words such as BASE BOTTOM FRONT HEAD HIGH which show the correspondences reasonably well.

For elements in the fourth row, the Lateral group, the associated arm-movement is a side-to-side movement across the body, the least extent of movement being associated with the speech-sound S and the greatest with the speech-sound Z. It is because of the characteristic sideways movement that these elements have been formed into the Lateral group. Words where the first speech-sound is an element belonging to this group often seem to show the characteristic movement for example, SIDE SHAKE SHOVE WAG WAVE SWEEP SWAY EXTEND ZIGZAG(Humboldt drew attention to the correspondence between sound and meaning in many similar words in German).

For the Circular Group, in the fifth row, the movements associated with the speech-sounds are small but very significant rotations of the arm on its long axis. The Table shows in enlargement the differing positions of the hand as a result of rotation of the arm. The least degree of rotation is for the speech-sound L, rather more for the speech-sound M (resulting in the back of the hand facing forward), still more for N and the greatest degree of rotation for the speech-sound R (more than a half-turn of the hand, to the limit of what is anatomically possible). The effect of the rotatory movement is most marked when speech-sounds in this group are combined with speech-sounds from other groups - by way of illustration the Table shows the effect of combining L with the speech-sound A in LA, with the whole arm being rotated into a high position (similar to the familiar arm position seen in holding up a LAMP). Words which seem to display the rotatory effect of elements in this group are ROLL ROUND TURN BEND MOULD CURL DOME MILL.

It will be noted that no examples have been given of words showing the effect of elements in the first group, the Vowel Line. This is because in any word the positions and movements associated with individual speech-sound elements (as illustrated in the Table) are melded or blended together; this is particularly the case for the vowel sounds, which affect the duration or physical extent of the associated action rather than the shape or character of it. Still, there are words such as UP TOP AT THAT which seem to have associated with them positions derived from the positions attributed to the vowel elements. This is of course a very sketchy outline of a system of correspondences which was elaborated in Allott (1973). The essential point is that no speech-sound by itself determines the action-patterning corresponding to a particular word. Each speech-sound is an articulatory and action subroutine which combines and interacts with the other speech-sounds in a word. The contribution made by an individual speech-sound in a word and thus in the associated action pattern is critically influenced by whether it is initial, medial or final, since the position or movement associated with the particular speech-sound has to be accommodated to and melded with the positions or movements associated with preceding or following speech-sounds.

Such is a possible foundation for the microstructure of the lexicon. To explain the system in all its complexity would take more space and time than are available on this occasion. All that it is intended at the moment to establish or make plausible is that some such scheme of correspondences is conceivable as the basis underlying the sound-structure of words in the lexicon and as providing a possible way in which the otherwise elusive link between the sound and the meaning of individual words can be found.

Can a scheme of this kind, for example, help us to understand the functioning of the words discussed earlier: CUP GIVE THE ME ? Using the correspondences shown in the Table, these words would be represented by the following:

The combination of the speech-sounds G I V would produce first an upward position of the bent arm corresponding with the sixth element in the Main Consonantal line - the hand held just at the level of the top of the head - which would be succeeded by and combined with a movement forward to the position associated with the vowel I, finishing with a vigorous forward movement corresponding to the Projective element V. In combination, the three speech-sounds, on the scheme of correspondences outlined, would have associated with them a total action very similar to that we think of as constituting the action of giving something to somebody.

The combination of the speech-sounds M and I would produce (similarly to the effect already described for LA) an inward turning movement of the hand and arm, pointing towards one's own chest.

The speech-sounds in THE (with a short vowel) are associated with an arm position appropriate to an element in the Main Consonantal group, with the hand and arm initially held in front of the face but moved forward, as the TH sound is melded with the position associated with the vowel. Note that the addition to THE of the speech-sound T and change of the vowel to A converts THE into THAT, adding a sharp forward movement indicative (deictic) in character. We see by this that THE is simply a reduced form of THAT, a weak demonstrative.

Finally, the speech-sounds combined in CUP have associated with them a movement forward from the bent position of the arm for C (hand in front of the mouth, up through the position associated with U, then moving forward sharply to end with P, leaving the hand and the arm extended forward with the hand in a cupped, concave position. The correspondences in action-patterning associated with the individual sounds combined in these words are simple and readily related to the meanings of the words.

This paper is already long enough and there is no possibility of going on to consider in the detail required how the proposals on the microstructure of the lexicon can be tested and verified in terms of the wider systematic organisation of the lexicon, It is a large task to identify in the lexicon of any language the extent of uniform relationships and regular patterning in the variation of sound and meaning over blocks of words.

The sceptic can only convince himself by trial and error, see if he himself can observe signs of a systematic relation between word-meanings and the sound-pattern of words. There is no room in this paper to investigate in depth the corroborative or circumstantial evidence for the validity of the proposed microstructure by studying the macrostructure of the lexicon, the remarkable groupings of words which, without having any generally recognised etymological relation, yet have a clear and progressive relation of meaning moving in parallel with broad similarities of sound-structure. Typical groupings are found in CLAP CLASP CLAMP CLIP CLING CLENCH CLUTCH CLUMP CLUSTER CRUNCH CRUSH CRUMBLE CRIMP CRAMP CRANK CROOKED CRUTCH CRAB GRAB GRASP GRIP GRAPPLE GROPE GRIND POINT POKE PIKE PIERCE PIN PRICK PROBE TOUCH TAB TAG TAP TACK TICK TAKE JAB JERK JOLT JOG JET JUT SPIT SPEW SPOUT SPURT SPATTER SPLATTER SPLUTTER SPLOSH SLOSH SLIP SLIDE SLITHER GLIDE SKIM SKIN SHIM SLIM SCRIMP SCRAPE SCRATCH STOP STEP STAND START STIR STRIDE STROLL STRUT STRAY STRAGGLE. There is an unending collection of similar words where variations in meaning seem to change in step with variations in sound-structure.

All one can say is that once the idea is accepted, or at least treated as a reasonable hypothesis, that the sound and meaning of words can be systematically related, the way is open for a quite different categorisation of the lexicon from the simple alphabetic or semantic ordering of the kind attempted by Roget, Trier or Weisgerber. Anyone wishing to pursue this might start by noting the surprisingly large number of words in any dictionary where, in attempting to convey the meaning of a word, the authors of the dictionary use words with related meanings which are, in sound-structure or in the most prominent speech-sounds, remarkably similar to the word they are seeking to define. For example, SLIDE in a dictionary is explained through GLIDE, STRIP by SLIP, SLIM by THIN and so on.

The time is perhaps overdue to return to the central thrust of the symposium, the bearing of what has been said on the origin and development of language. The return is quite direct. It assumes that the processes which operate in language as we use it today are processes which also operated in the earliest stages of the evolution of language. The importance, and so the vigour, of the attempt to communicate through sound a pattern of bodily action or the contour of something heard or seen would have been even greater then than at the present day. If action-patterns in modern man can be redirected into articulation-patterns, then in primitive man this would have provided a means for creating appropriate words, understandable words, to be used for marking out the features of his world. The development and the richness of the genuinely primitive lexicon, the first words of man, will have been dependent upon the progressive elaboration of man's capacity for skilled action, including skilled imitative action, coupled with the evolution of a sufficiently refined musculature of the throat and mouth to allow an effective transfer of patterning from bodily action to articulation. Not only must man's anatomy have evolved to allow modulation of the stream of breath to produce a wider range of distinct sounds but also there must have been the evolution of the musculature controlling the positioning and movement of the lips, the tongue, the vocal cords etc.

At the same time, evolution in the precision of arm and hand movements, with the survival premium for increasing skill, discrimination and accuracy, will have widened the range of his experience and of what he would wish, or need, to be able to communicate to his associates. It would seem no accident, if the link between the patterning of hand and arm movement on the one hand and the patterning of speech on the other is as close as has been proposed in this paper, that the elaboration of the neural basis of language should have taken place largely in the left cerebral hemisphere, in close association with elaboration of neural control for fine manipulation by the right hand and arm. Development of language and manual skills on the view presented in this paper would have proceeded together, reciprocally modifying and refining each other. We then could see the word in its real character as a unique pattern of neuromuscular organisation closely linked to overall muscular organisation and as a neural pattern readily integrated in the general neural structure of the individual, linking with all the neural patterning underlying the basic human functions of perception and action and playing a part in the overall development of human intelligence as the expression of the harmonious and effective integration of experience.