From: Bannister, D. (1970) Perspectives in Personal Construct Theory. London and New York: Academic Press, pp.223-253.

Men, the Man-Makers: George Kelly and the Psychology of Personal Constructs

John Shotter

“. ..historically the questions that men have asked turned out to be more important than the conclusions they reached.” George Kelly

Introduction

The aim of this essay is a very broad one and to do with the foundations of Psychology: It is to work through a number of the implications in George Kelly’s (1955) “The Psychology of Personal Constructs” and to discuss what might be the source of such constructs.

              Kelly’s model of man was, so one might say, of man with a model, an essentially reflexive idea which at first sight might seem to be an unproductive tautology. Tautology it may be, however it is certainly not unproductive; it is just those sort of logical structures, which in a sense contain themselves, that Chomsky (1957, 1965, 1968) is using to such powerful effect in his description of language. They allow, in the v. Humboldtian phrase, for “infinite use of finite means”. It is because of the “reflexive creativity” immanent in Kelly’s model of man that this essay refers to “Men, the man- makers.” This is one theme we wish to explore.

              Another related and no less important one we wish to explore is to do with the question “What must be the essential character of our knowledge, both of ourselves and nature, that is expressible in language (in symbols)? This is a question which, in one sense, Chomsky (1965) has been raising when he wonders whether “...the general form of a system of knowledge is fixed in advance as a disposition of the mind. ...” (It is tantamount to the question “Are there any clear limits to scientific knowledge -- especially to such knowledge of ourselves ?”).

              Our discussion falls into two main parts: In the first, we discuss the power of man’s self-images, and ponder upon the history of those currently represented in Psychology (and some which are not). We then take Kelly’s notions of construct and construct-system and see “what we can do with [end p.223] them” (Kelly’s criterion of acceptability for ideas). We find that they have consequences in the planning and regulation of an individual’s actions, and also in the co-ordination of our social interactions.

              In the second part, I suggest an extension of Kelly’s scheme to account for the process by which man acquires his ‘models’, taking up Kelly’s (1969) points on the priority of behavioural encounters. The ‘knowledge’ that we gain by such bodily activity (for that is what encounters are) is ordered only in bodily terms and cannot be reordered; it is in the sphere of social interaction that schemes of ordering-constructs and construct-systems -- are invented, and these may then be imposed upon such knowledge so that it can be reordered (or re-construed, in Kelly’s terms) in different ways. And thus, to our question on the limits of our explicitly expressible knowledge, we reply, “Yes there is a limit, it must be expressed in terms of differences, in terms of constructs, with all the inherent limitations that that entails.” All the other limitations on our knowledge are due to our bodily nature. Further, if ‘knowledge’ gained in our bodily encounters with the world and others cannot be expressed through a construct-system, because none exist, then that ‘knowledge’ remains inexpressible in anything other than spontaneous bodily activity -- it can only become both anticipatory and voluntary as a result of (in Kelly’s terms) a construal. So, while for us as individuals our ‘feelings’ are absolutes, ‘known’ in terms of the encounters that produce them, their systematic expression must be in relative terms, in terms of the differences that we can make.

              The essay is brought to a close by pointing out that, until now, a thorough-going behaviourism (like Kelly’s) has not been possible as previous approaches (e.g. Miller et al., 1960) have been confounded by mechanistic diversions. For the understanding of order, the focus of attention should not have been upon the orderly actions of the individual, but upon the sphere of social interaction from whence that order arises. Thus if Kelly has any occidental (there seem to be many oriental) spiritual brothers, they are Mead (1934), Vygotsky (1962), and Wittgenstein (1953).

Constructs Have Consequences -- In the Actions of Men

THE CRISIS IN MAN’S KNOWLEDGE OF HIMSELF

By now it is almost a platitude to say that there is some sort of crisis in western thought (and life). It is perhaps not quite such a platitude to say that it seems to stem from what is perceived as a discrepancy between what men know and feel about themselves, personally, and what their accepted public image (or plethora of images) allows as their nature. In short it is a division between hearts and minds. While, so it is said, there has been an [end p.224] explosion in knowledge, there seems to have been a contraction in our understanding of ourselves -- it is our self-image that has shattered.

              This is disastrous, for these are the images which we use to plan, regulate and co-ordinate what we think of as our rational activities in the public sphere. If these images are inadequate or conflicting, we run the grave risk of all our actions in that sphere being irrelevant to our needs or of being in conflict. Cassirer (1944), from whose Essay on Man I have taken this section heading, ended his statement of the problem thus:

“...our wealth of facts is not necessarily a wealth of thoughts. Unless we succeed in finding a clue of Ariadne to lead us out of this labyrinth, we can have no real insight into the general character of human culture; we shall remain lost in a mess of disintegrated data which seems to lack all conceptual unity.”

Is it possible, then, to construct for ourselves a self-image which does not, in his words “cripple our own form-shaping powers”, which will make us whole once again? Cassirer saw the key in the understanding of symbolism; we shall find the key to the understanding of symbolism (at least that aspect relevant to the sphere of Science, not that relevant in myth and magic) in the understanding of George Kelly’s notions of construct and construct- systems.

MAN AND HIS IMAGES

Kelly was one of the few psychologists to systematically point out the discrepancy between the “personal” and the “public” knowledge of men, between the way someone talked about himself and the way an external observer talked about him. He (Kelly, 1955) observed that while man-the- scientist could entertain “theories”, “ideas”, “thoughts”, “images”, “mental- pictures”, etc., of what man might be like (notions which serve to define what is being sought and to guide how to look for it), man-the-scientist’s- subject can only be, scientifically speaking, as the psychologist’s theories allow him to be. However, because so-called “mentalistic” terms are allowed no place in an objective science, the psychologist’s subject cannot be influenced by his “theories”, “ideas”, etc., they must be determined in some quite other way -- by, say “stimuli”. Thus the split between the two ways of understanding man is set and cemented in right at the foundations of modern Psychology . How has this come about and how was it justified?

              In the past, and currently too, psychologists thought that if they were [end p.225] going to be “scientific” (i.e., act as if they were going about their business as accredited scientists do), it was necessary to model their activities upon the ways of going on, invented myth has it, by Galileo and Newton for the study of the, then given , material world. These were the ways, the “paradigms” (to use Kuhn’s (1962) term), that Hobbes and Locke (which some look upon as the fathers of our modern approach) used for their study of man. In applying the “new method” of Galileo, and. starting from the premise that we too were material beings (what else could we be ?), Hobbes (1651) has this to say:

“. ..what is the heart, but a spring; and the nerves but so many strings; and the joints so many wheels, giving motion to the whole body.”

Thus inaugurating our present phase in Psychology -- the phase of “man is a machine actuated by its circumstances” -- as the last transformation in a fantastic sequence of such conceptual transformations which we can trace back at least to Plato.

              This sequence seems to have gone something like this: to begin with there seemed to be chaos and reality; only the real was intelligible (via ideal forms) ; only that which can be said explicitly may be true or false; only that which is arrived at by rational thought and is thus beyond dispute can be true; truth and falsity are matters of logic; the logic that deals with the real world is mechanics; thus mechanics must be applied to man for a true understanding of man, who is part of the real world. Each new phase was an abstraction from its predecessor, and its significance could only be fully understood within the terms of the whole tradition: it was this whole tradition which Kelly repudiated.

              If we had gone back to deeper roots, to the pre-P1atonic philosophers for instance, we might have found there seeds which, due to lack of nurture, have not yet grown and blossomed so; seeds which might have brought forth Kelly-flowers.

              What of Protagoras, who thought each man to be “the measure of all things, of all things that are that they are, and of things that are not that they are not,” thus for him there was no objective truth in virtue of which one man is necessarily right and the other wrong? But his writings are lost and only known through Plato.

              Kelly’s true Hellenic ancestor seems to be Heraclitus. That the world is full of “things”, he thought, was an idea arrived at as the result of “analysis” -- a human activity practised, he alleged, by such men as Protagoras, consisting of classifying the constituents of an aggregate according to a “table [end p.226] of opposites” -- this does not give us an account of how the world actually is: only how we think it to be. To attribute any absolute or fundamental significance to such an analysis would be a mistake (alternative analyses are to be had) ; it would be an inexcusable failure to understand what Heraclitus called, the logos-the way that words work to give only a one-sided determination of the ever changing flux, a determination arrived at as the result of one analysis among many.

              It must have been the dawning realization of the power of logic which made Heraclitus unintelligible to his successors, for a sequential expression of the results of different analyses yields what in logic are thought of as paradoxes. Heraclitus summarized his position, in what to us, but not necessarily to him, was an extended paradox:

“Things taken together are wholes and not wholes; being brought together is being parted; concord is dissonance; and out of all things, one; and out of one, all things. “

THE USE OF “CERTAIN KINDS OF LANGUAGE FORM” IN SCIENCE

At various points in this essay words will appear in single quote marks, this will be to indicate that ‘’as if” is intended. Thus men may have models (no single quotes) which they hold in their hands and the word “model” could only have been devised and learnt by reference to such situations. To say that “Man has a ‘model’ “ is to say that he is behaving just as if he had a model in his hands to which to refer in pre-enacting his possible courses of action prior to actually encountering the situation so modelled. It is as if he possessed such a model but in any obvious material sense does not; it is somewhere “inside” him. Clearly, in deciding upon such a usage for the word “model” behavioural criteria must be involved.

              The ‘’as if” language-game [to use Wittgenstein’s (1953) term] was well-known to Kelly (1964), for him it was a sub-game in the great “Language of Hypothesis” game. Illustrated below, in an extract from one of his writings on this topic, are many of the focal issues which he expressed in one form or another throughout all his writings. He is discussing whether a floor is really hard or whether it is really soft -- is it a matter of proof? Or:

“Suppose instead we employ the language of hypothesis. We say, in effect, ‘To be sure the floor may be regarded as hard, and we know something of what ensues when we cope with it in the light of such an assumption. Not bad! but now let us see what happens when we regard it as soft.’ Out of this further exploration may come, not so much confirmation that it really is hard, or soft – as Descartes would have reasoned – but a sequence of fresh experiences that invite [end of p.227] the formulation of new hypotheses. ..the object of the discourse (above) is only to suggest how a certain kind of form of language can enable us to extricate ourselves from the kind of realism to which our so-called objective language has bound us. ..one can see most clearly how man can be trapped by his indicative verbs and how, in turn, he has been led to believe that he must choose between mutually exclusive versions of reality. ..but the language of hypothesis enables his therapist (or the scientist, or the person himself) to say ‘But only suppose the floor is to be regarded as hard’”

For Kelly then, there is no division in Science into pure and applied, the one concerned with the proof of an idea and the other with its exploitation; Science’s primary concern is with possibility. In summary its characteristics are: Science functions in the sphere of instrumental activity; it does not provide a description of “reality”, but schemes for the ordering of experience ; scientific observation is not a passive process, but an active one, and what one observes is relative to what one does; there are no questions about which view is “true” in some ultimate sense, only the question of whether we can (and should!) use it to guide our actions in some way or other -- questions of logical justification can arise only within schemes not between them; the achievements of one science thus do not negate those of another (Kelly, 1969, p. 135); only in logic is choice mutually exclusive, of the A vs not-A kind; elsewhere, even in the conduct of science (Kuhn, 1962), it is what Kelly calls a “double-entity” choice -- choosing between one way of going on or another, choosing not to go on at all is only available to us in extreme circumstances.

              All these characteristics of science -- at least these-follow as a consequence of accepting Kelly’s model of man; we shall find more during the course of this essay.

              “Ways of going on,” then, can depend upon the invention of, for example, “a certain kind of language form,” which once in man’s possession can be put to special kinds of uses. Yet it is a construction of man, no less real because of its intangible nature than any of his more material constructions. To understand its nature (one way of understanding, Kelly would point out), is to discover both how he made it and how he uses it, and this, as we shall find later (p. 241), seems to be related to the problem of how we can come to do deliberately what we can already spontaneously (Vygotsky, 1962), and [end p.228] how what we learn is first reflected on our spontaneous activity and only after instruction by another person can be expressed in language. But to understand the details of this process a close observation of it is necessary. To think that there are no empirical issues at stake here just because the material world is not under study would be a great mistake.

              To turn now to Kelly’s model of man.

KELLY’S MODEL OF MAN: MAN WITH A ‘MODEL’

In the face of a Behaviourism which incorporated all the features of the Platonic-Hobbesian tradition soutlined above, Kelly’s heresy was to allow the psychologist’s subject all the capacities of his investigators. Thus all men could be regarded as “scientists” to the extent that they used “theories”, “ideas”, “thoughts”, etc., in the planning and regulating of their individual actions, and in the co-ordinating of their activities with others: Kelly’s model of man was one of a man who is as if he owned a model of his world as a personal possession, to use just as he saw fit-to use in rehearsing a situation prior to its actual occurrence. Such a model that contains itself, so to speak, is reflexive.

              The key to this idea is to realize that man is not meant to contain a model of himself as a material thing, but as a form-producing process. Just as a seed does not materially contain the plant, but produces it from the chaos of materials in the soil around it, organizing them into a plant which goes on to produce seeds for further plants, and so on, Kelly’s model must be understood in the same sense; but it is to do with man’s ability to impose upon the chaos around him forms which he finds intelligible. Thus when we use the term “invent” and say “Man invents the things in his environment”, we shall not mean that he conjures them up out of nothing, we shall mean that he imposes upon what was once for him chaos, constructs and construct- systems which enable him to decide what to do in the face of such chaos.

              The idea of “reflexivity” as a form-producing process is of relatively recent origin. As “self-replication” it occupies a central position in genetics (Beadle and Beadle, 1969) and as “recursive function theory” in the foundations of computing (Davis, 1958). It has turned up in Psychology in the computer simulation of problem solving (Newell et al., 1958) and is the essential property that makes Chomsky’s grammatical schemas generative.

              One way of viewing this essay is as a working through of similar such implications in Kelly’s position.

A MODEL FOR ‘MODELS’

Kelly’s “Theory of Personal Constructs” (1955) is to do with man’s psychological constructions -- in that way it is aptly named, but we shall find that, in one sense, constructs are not personal at all. [end p.229]

              A construct is invented when a man notices a difference, draws the attention of a fellow to it, and they agree upon a way of representing it. A construct is used when a man makes (expresses) a distinction using the agreed ways of representing it, such as: strong/weak, high/low, voiced/unvoiced (phonetic distinction), etc. ...To establish (or erect) a construct is more than simply making a discrimination; we shall find that a social psychological process is involved.

              A construct-system is invented when a whole set of related distinctions is involved. The best example of an explicit system that I know is the distinctive feature analysis imposed upon the sound system of a language (Jakobson et al., 1952). We shall discuss it on p. 244 of this essay.

              However, to repeat, constructs are imposed upon “something”, they are involved in a form-producing process not a knowledge acquisition one. Can this “something” have a nature quite apart from any analysis which we might impose upon it? (We are back, here, with the problem of “the-thing- in-itself” ; there is no avoiding it.) When we give an analysis of some object, what might be the nature of our ‘knowledge’ of that object upon which our construct-systems are imposed, and how might that ‘knowledge’ be acquired ? We shall deal with the first part of this problem here, the second part is left for a later section (p. 235).

              A working solution to the first part of the problem is deceptively simple : The “something” upon which constructs are imposed, the ‘model’, is like an actual material possession of the individual-as we have said before.

If we are going to give an analysis of an object, we take it in our hands and begin to describe it-this way and that, from one point of view then another, turning it over, looking inside it, feeling it, and so on. Our descrip- tion of it depends upon how we examine it. Can we do the same for objects which we know well, in our imagination ? Well, yes we can. ..to some extent (it is because of such differences which cannot be gone into here that I call this only a “working solution”). And we can see that constructs enter only into the process of description, and not into the memories of our once palpable contact with the object. For our memories are open to construal in a number of different ways, by the imposition of a number of different construct-systems upon them. We do not just remember our construals of a situation, we remember what for us was the situation.

              If this seems to upset received ideas about the nature of knowledge -- that true “knowledge” is what is put into books and stored upon library shelves -- well ...that is our heritage from Plato. In this issue there is a sharp distinction to be made between our ‘knowledge’ of a situation which we, [end p.230] as individuals, derive from our actions in it, and the knowledge of it transmitted to us by others, which they express within a certain scheme of established conventions, within a scheme invented by men. Thus there is a very clear distinction to be made between constructs and that upon which they are imposed. The latter is personal, absolutely so. We shall have to wait until later (p. 240) to see that there is a sense in which a construct-system, but not a construct, is personal also.

THE RELATION OF FACT TO THEORY

Psychologists seem to be in some confusion about the nature of facts; they do not seem to realize that their appreciation depends upon a psychological process which cannot itself be taken for granted. Not only do they seem to require Science to give us the facts and nothing but the facts, but they do not draw a clear distinction between facts of an everyday nature and “scientific facts”. Furthermore, they often fail to notice that the word “theory” refers indistinguishably to both a conceptual system (an organized body of knowledge somewhere inside us) and its formal expression (in ink-marks on paper), a requirement for scientific theories.

              Individuals who share in a culture seem to take “the world” (their world) -- in which they think, love, hate, play, do things and find one another’s actions intelligible -- as “given”. While everyone who possesses what some might call a culture’s Weltanschauung can agree about everyday facts, how can they agree about the possibilities which might be made to follow from them ? -- there are no existent circumstances to which to refer. However, if we could decide upon objects by which to represent them, we could refer to arrangements and rearrangements of these. It is in mediating judgements of this sort that scientific theories (their formal expression) seem to have their function.

              One of the important distinctions between a Weltanschauung and a scientific theory is that, while both reflect a conceptual system, the former is intuitively held, there being no way to reveal and characterize its nature, as alternative views with which to contrast it are (usually) unavailable, the latter, however, must be publicly expressible, and furthermore, if it is to be properly tested, of a form amenable to comparison with alternatives.

              Thus it is that no matter how substantial the circumstances might seem that an allegedly factual statement describes, there is in the appreciation of that statement a theoretical component, a psychological process is involved in its understanding. It is, to my mind, this process which Kelly’s theory illuminates.

              For Kelly, a person’s actions, and this includes what he says as well as what he does, can only be understood by another person to the extent that he possesses a construct-system similar to that upon which the person’s [end p.231] actions are predicated . Someone “can express himself only within the framework of his construct-system, words alone do not convey meaning” (Kelly, 1969, p. 83). And it is the job of whoever it is who must understand him to construct for himself a similar system of constructs. (Perhaps in Experimental Psychology the opposite often happens: during an “experiment” the subject comes to construct a system similar to the experimenter’s, operating only within his predetermined options-this is perhaps what “warm-up” or “practice” periods are for.) If the formulation of the construct-system is being undertaken within the sphere of science, then it must be made explicit and expressed by being represented materially to yield an account available for public examination and manipulation.

              In Physics the lack of a clear understanding of the relation between fact and theory is tolerable-any way of doing something against our environment may be useful to us. But in Psychology it is quite intolerable, it might lead us to accept, as a scientifically proven fact, that our nature is determined in quite unalterable ways, and we might easily forget that the ways in which we construe our own nature have been invented by us, and are relevant only to how we achieve our purposes and not to how we formulate those purposes in the first place. Science has nothing to do with the formulation of ends.

              To return to the relation between fact and theory, we can note that others have made the point, similar to Kelly’s, that “a fact” only has significance if an individual possesses a way of interpreting it, otherwise, it is literally nonsense. A “scientific fact” is not, Cassirer (1944) says:

“. ..given in any haphazard observation or in a mere accumulation of sense data. The facts of science always imply a theoretical, which means symbolic, element.”

For example, Galileo’s law of motion referred, not to actual bodies, but to idealized ones. And what Galileo did was not just to find his law by painstaking observation and measurement, but to invent the mode of idealization which determined the nature of the measurements he made.

              There is no way in which it makes sense to talk of a “pure fact”, in toto undistorted, with no construct-system imposed upon it. Kuhn (1962) puts it another way:

“Scientific fact and theory are not categorically separable, except perhaps within a single tradition of normal-scientific practice...[end p.232] Men whose research is based upon shared paradigms are committed to the same rules and standards for scientific practice. That commitment and the apparent consensus it produces are prerequisites for normal science.”

So, however one looks at it, what can be expressed in scientific terms, that is, explicitly written down or unambiguously encoded into some permanent physical system of expression, depends for its interpretation upon those encountering it possessing the appropriate ‘model’ and constructs: otherwise, it is just part of the unknown flux.

              Goethe expressed it all in a sentence : All that is fact is already theory.

DIFFICULTIES WITH OUR PRESENT IDEA OF WHAT SCIENTIFIC

PSYCHOLOGY SHOULD BE LIKE

A number of important implications for scientific Psychology follow from the discussion above: First it is not possible to base Psychology upon “facts”. It is not good enough to assert that Psychology is a science because psychologists approach their subject matter with a scientific attitude -- “a disposition to deal with facts rather than with what someone said about them” (Skinner, 1953). It is one of Psychology’s tasks, one which in my estimation Kelly has done much to elucidate, to account for how we come to agree upon what constitutes a fact, and how what is construed as a fact from one standpoint is not so from another .

              Next: If a science cannot exist apart from its means of expression yet within its existent means, there are serious limitations when it comes to giving expression to many well-attested modes of human feeling and conduct -as indeed there seems to be with the language of science, see what Wittgenstein (1953) said about Wittgenstein (1922) -- then psychologists should not feel inhibited as instituting new forms of reliable communication (if any can be invented).

              Perhaps they already exist and it is a matter of empirical investigation to discover them. This must have been the case with logic, for as Ryle (1949) says:

“Rules of correct reasoning were first extracted by Aristotle, yet men knew how to avoid and detect fallacies before they learned his lessons, just as men since Aristotle, and including Aristotle, ordinarily conduct their arguments without making any reference to his formulae. They do not plan their arguments before conducting [end p.233] them. Indeed if they had to plan what to think before thinking it they would never think at all; for all this planning would itself be unplanned. Efficient practices precedes the theory of it. ..”

              In Science, we have bound ourselves, as an article of faith, to aim (we always fall short, necessarily) at a particular language form made for reasoning -- a form which limits itself to names and patterns of relations between them. Surely, now, it is possible to allow other forms?

              Finally: It seems that once and for all Psychology must relinquish its Newtonian dreams (hasn’t Physics) of the one single unified edifice of knowledge available to morons in cookbook form. It is to commit our students to yet two hundred years more of a Psychology without people for Broadbent (1961) to say:

“Against the perspective of two thousand years the speed of our advance in studying behaviour seems more cheerful. At a rough guess, two hundred more years may bring the study of behaviour up to the level which physics reached in Newton’s time. The problem to be faced is less tractable than that of the organisation of the Solar System. ..! ”

For it ought not to be forgotten that Newton’s laws do not deal with particular bodies or masses, but with any point-masses whatsoever. Laws which are about any people whatsoever-hypothetical proto-human forms with features common to cave men, wild boys, aboriginals, modern western europeans, etc.,-- will exclude at the onset any account of what we think of as our humanity. For, as we know, babies born to us need not grow up to be what we understand as human; their humanity is transmitted to them after birth. It is inherited, but like houses and cities not like blue eyes, and it is of course the responsibility of a culture to make sure its young get their birthright: the inventions of their ancestors. The laws of proto-human forms phrased, if that is imaginable, in psychologically equivalent terms to point-masses, would ignore such inventions, thus denying us just those abilities by which we identify ourselves as human.

              These are just some of the implications that we can draw out of Kelly’s understanding of the relation between fact and theory. Later we shall take up a related issue: the way that “laws of nature” depend for their means of expression, their mode of existence, upon “rules of social conformity” – that is, “rules” are epistemologically prior to such “laws”, which cannot, thus, be turned round to provide an explanation of our behaviour. For explicit explanations we are locked within our own constructs.

              However, we must turn now from the discussion of what we can do with constructs once we possess them, to a consideration of how we might acquire [end p.234] them -- construct them -- and how we acquire the ‘knowledge’ upon which they can be imposed. We shall then be in a better position to understand their nature further.

Actions Have Consequences – on the ‘Knowledge’ and Constructs of Men

THE PRIORITY OF ENCOUNTERS

For Kelly, man was not something which in its natural state was static and upon which, action, motion and life had to be superimposed. It was taken as given that he was a self-active being in the world; also, he was a creature who made his own circumstances. What, then, might be the capacities of such a being? Why, he defines them for himself: “It is by his actions that man learns what his capabilities are. ..” (Kelly, 1969, p. 33). He utterly repudiated the “Darwinian assumption” that most psychologists seem to impose on men – although he was not sure that Darwin himself would have made it – that mankind had finally stopped evolving, and thus all that psychologists had to do was itemize his nature. It is a question of invention, not discovery .

“Out of. ..further exploration may come not so much confirmation. ..but a sequence of fresh experiences that invite the formulation of new hypotheses. ..”

That quotation appeared earlier, within the body of a more extended quotation from Kelly’s comments on the “Language of Hypothesis”. Here, as in many other places, he emphasized the role of action, behaviour, as the primary source of man’s new ‘knowledge’ both about the world and himself. But here again, I feel that Kelly did not sufficiently work through the powerful implications in his own position, even though the comments littered about his writings were obviously informed by an intuitive appreciation of them. We will explore the consequences of taking bodily activity as a logically prior datum. One might be forgiven, at this stage, for saying “So man does what he does because he does what he does, does he ? Tell me something new.” Admittedly it is an avenue of exploration which, prima facie, does not seem especially fruitful. However, it is not quite as circular as it may seem, it does have a basis, a starting point: We must accept that we are responsive to a situation in a way quite apart from any construction which we might put upon it -- we can in a rough and ready way respond to the actions of other beings even when they do not or cannot talk to us. And thus, prior to the possession of any construct-systems, we must assume that, innately, we respond in a differential manner to our circumstances (as must all living things), albeit in a counteractive rather than anticipatory manner.

              In this section of the essay, then, I would like to outline a mythology (Shotter, 1969) and a set of suggestive questions which will carry us from the [end p.235] datum of bodily activity to the theory of constructs, both personal and public.

              Showing the behavioural origins of constructs will amount to outlining a behaviourist theory both of behaviour and the genesis of what we think of as mental activity. Thus it will be substantially different from any of those within the Platonist/empiricist/rationalist/associationist/mechanistic traditions.

              I will first present a mythology – a new one to replace the old – and then the suggestions.

A NEW MYTHOLOGY OF MAN: MAN AS THE CREATORS AND POSSESSORS

AND USERS OF CONSTRUCT SYSTEMS

We begin with proto-beings who are not self-aware individuals at all (Mead, 1934). They live together, initially like animals: reacting to one another directly and spontaneously, but in ways modified by ‘knowledge’ initially derived from their encounters with the world and each other. Before they created language, they were aware of nothing of their own nature and planned nothing consciously or deliberately.

              But then they created amongst themselves a means for co-ordinating their communal activities: language. In doing so, not only did they make themselves aware of aspects of their otherwise spontaneous activities, they also provided themselves with a means for modifying these activities in a deliberate fashion. Once it was possible to request an act of another, and he of you, it was possible to request it of oneself (Luria, 1961). As ways of life (“forms of life”, Wittgenstein, 1953) became stabilized and it was possible to request organized sequences of actions of another, and he of you, it became possible to plan and to regulate one’s own activity (Vygotsky, 1962).

              While it was the internalized actions of the individual that made the material for thought (Piaget, 1952), it was the internalization of dialogue (Vygotsky, 1962) that provided schemes for the ordering and re-ordering of such material.

              Whereas the original creations of men were blind – there was no pre- established plan or guide – children now re-invent (Chomsky, 1968) them again, but with their mothers and friends as guides. Men can learn such things as language because men created such things-learning them is guided invention. But learning the things of men is different from learning the things of the world, the first are governed by rules of social conformity, the latter by laws of nature. Our means of expressing the latter depends upon the prior possession of the former (Wittgenstein, 1953).

THREE SUGGESTIONS TO DO WITH THE GENESIS OF “MIND”

The above is a story, pure fiction. Can we get away with saying “But then they created. ..language”, as if it raised no problems? When Locke (1690) [end p.236] was discussing language, thinking that men were endowed with it, he said “. ..nature, even in the naming of things, unawares suggested to men the originals and principles of all their knowledge. ..(Bk. 3. Ch. 1. sec. 5)” and that was the story he had to tell to get on with the analysis he wanted to make. It was a useful myth, as so too, I hope, is the one above in its own way .

              One consequence of it is that everything which we think of as “mental” activity, or a product of “mind”, is an invention of man-even “mind” itself is a cultural product. “Nothing which happens in the brain can be described except in terms supplied by the mind” (Mumford, 1967). And, Kelly would say, to find the seat of the mind do not look into the brain, look at those responsible for it, look at men, whole men with their fellows in the world.

              Now to frame some questions (we will take spontaneous activity as given):

‒            1. “What if ‘knowledge’ could only be derived from the self-instigated actions of the body?”

‒            2. “What if the unknown is ‘felt’ first via bodily encounters; and only later when it is so known can it be construed into parts, one way by one construct system, another way by another, where the invention of construct-systems arises out of co-ordinated activity within the sphere of social interaction?”

‒            3. “What if, within the sphere of social interaction, a construal (one among many) of spontaneous, context-bound, non-conscious, “need-dependent,” activity can be imposed so that an act (a unit) can come to be done deliberately, out-of-context, to alleviate no present need? It could then be interpolated into spontaneous activities thus modifying them, and action could then be emitted in anticipation, voluntarily.”

These are not to be taken as formal propositions to be proved, but only as suggestions to be discussed. Let us take each one in turn and see what we can get out of it.

“Thought derives from the abstraction of one’s own action upon things” (Elkind, 1969) – Does it?

Only the behaviour of organisms that in some sense can acquire ‘knowledge’ could possibly become anticipatory – initially the aspect of its activity influenced by its environment must be counteractive in character, and without being modified by ‘knowledge’, remain so. But even for such activity of this primitive kind to be possible what goes on outside the organism must affect its behaviour in a differential manner (we might want to say that its movements were mediated by its “feelings”, but that is not necessary at this stage). [end p.237] This necessitates a fundamental assumption that must be made about all organisms . they must contain “internal regulators” which function to ‘guide’ (in the absence of all other guides) their activity’- regulators which determine, regarding its current circumstances, whether at least the organism is to go on or to withdraw. It is to these very primitive regulators that we will appeal to guide the protohuman’s actions until he has acquired some ‘knowledge’ and some constructs for construing it, and thus become human.

              We have not yet, though, made contact with construct theory; simply acting and acquiring ‘knowledge’ --if one is going to be able to put alternative constructions upon it – is not enough. To show this I am going to use in a modified form an analysis of concrete bodily needs to be found in Dreyfus (1967).

              Imagine a wild-boy divorced from any human community. He might develop his own abilities and knowledge of the world in the following way: At first he would not know what was the nature of his world nor what his needs were; he would simply begin by being in some state of restlessness or discomfort. And this is not a determinate state that can always be alleviated in the same way. He would have to actively cast around for a way of easing it, inevitably, in the process encountering his environment. Once a way was discovered he would come to know his need in terms of what he had done to alleviate it, and to ‘know’ the structure of his world in terms of the actions within which he had encountered it. And next time he recognized himself to be in the same state of restlessness he could do what he did before, but this time in order to relieve it. Presumably, he could discover many needs to himself in this sort of way, and he would ‘know’ his world in terms of his needs and the actions that he must initiate to satisfy them. Furthermore, we might presume that his ‘inventions’ would become ‘polished’ with repetition, as he discovered that not all aspects of them were relevant to his “needs”.

              Both animals and wild-boys may ‘teach’ themselves in this way, and I feel that it does reflect Kelly’s intuition that “It is by his actions that man learns what his capabilities are. ...” However, it does cause some difficulty. It provides only one construction, one which is tied to the organism’s needs – alternative construals are impossible. So although we now have behaviour that is anticipatory, it is still not so in the sense Kelly meant. We can only solve this problem when we turn to suggestion number two.

              Before we do that, however, it is necessary to comment upon our title for this section [taken from Elkind (Piaget, 1968), who uses it to express Piaget’s key notion of “reflecting abstraction”.] : The “things” of the child’s world are not the “things” of Elkind’s or Piaget’s world. His is still the statement of an external observer. Does thought derive from the child’s actions upon Piaget’s “things” or from those upon his own “things”? While ‘thought’ may be derived from our actions, we seem to derive from them both our own nature and “things.” Elkind’s statement, elegantly phrased as it is, ultimately is unhelpful.

“. ..the development of thinking is... from the social to the individual (Vygotsky, 1962)”

As the title for this section, a discussion of suggestion number two, I have taken a statement from Vygotsky (p. 20) in which he was totally disagreeing with Piaget’s (1923) then interpretation of language development: That it was a transition from an egocentric to a socialized use of speech. While the child’s view of the world may be egocentric, and this expressed in his speech, his use of speech could be, Vygotsky showed, initially for nothing but social purposes; this, after all, was the situation in which it was learned.

              “Egocentric” speech emerges, Vygotsky thought, when the child starts to use his language to plan and guide his own actions. It is only incomprehensible to others because it omits to “mention” what is obvious to the speaker. Such a language form seems to disappear when the child’s speech becomes sharply differentiated between the two functions: overt communication and covert self-regulation and planning.

              Thus it was that Vygotsky invented a thesis, the full implications of which are still to be worked out: An organized pattern of social interaction (and there are many other examples other than language) initially between two people (a child and others), creates a means for that child, at a later date, to organize his own behaviour.

              The reason why this is so important is because it shows that there is so much more to learning to talk than just making acceptable patterns of noises with one’s mouth. It is not that first there is language , and once learned it can then be used by the child in co-ordinating his social relationships. For, as Wittgenstein (1953) put it:

“If language is to be a means of communication, there must be agreement not only in definitions but also (queer as this may sound) in judgements. ..” (no.242).

Associated with the process of getting a child to re-invent a language, as those around him and their ancestors originally invented it, are many other processes not obviously linguistic. To put it shortly here, although I will [end p.239] expand on this a little in the final phase of this essay, they are to do with getting him to structure, to partition and attribute significance to the world in the same way as his mother. And this process, if words are later going to serve any communicative function, begins before the child learns, as we think of it, to talk (Newson, 1969). In Wittgenstein’s terms, it seems as if “agreement in (at least some) judgements” between mother and child is a pre- requisite for the development of language.

              Thus a child inherits (like a house, not like eye colour) at least two possessions when he is inducted into a culture, separable only as an analytic convenience: A special way of ‘seeing’ the world, of attributing significance to it, and interwoven with it, a skilful way of co-ordinating his social interaction in its terms.

              It is not as if the wild-boy in the previous section would not develop his own special way of ‘seeing’ the world; he did. But what language enables our children to do, which the wild-boy could not, is to re-arrange the component parts of his ‘view’ into new patterns. And it is a ‘view’ that represents and incorporates knowledge gained from the many different stand-points of his fellows. Thus our child may now construct for himself many different construct-systems for alternative construals of a situation-something inaccessible to the wild-boy. Furthermore, our child may make a personal selection of constructs and construct-systems. And it is thus that we can once again give a sense to the word “personal” in personal constructs.

              But we are departing from some of Kelly’s (1969) very fundamental points here. He looked upon it as “the creative capacity of the living thing to represent the environment, not merely respond to it. ..to represent other forms of reality, while still retaining its own form” (p. 8). Thus the ability to “construe” situations characterized all living things. If the views above are correct, it is an ability only of man, because only men can ‘see’ the world as others see it. This, however, is a departure that makes no practical difference to Kelly, concerned as he was, only with men; but within the broad panoply of Psychology it is a point worth making.

“Our new activities are our old spontaneous ones deliberately reconstituted into new ones”

For the discussion of our final suggestion I have modified some comments of George Miller’s (Miller and Chomsky, 1963) in which he ponders the source of our new plans: “Presumably, our richest source of new plans are our old plans, transformed to meet new situations” (p. 486). While that may be our richest source, it cannot possibly be our only source. The trouble with Miller and Chomsky’s scheme is that it will not get us beyond the possibilities implied within our current symbolic schemes, and then only [end p.240] with our logical schemes at that . For in their scheme plans must be treated as “objects that can be formed and transformed according to definite rules.” Yet as Kelly (1969, p.31) says:

“There are first-time occurrences in the history of mankind. ..a psychology that pins its anticipations on the repetition of events it calls ‘stimuli’ ...will find its accurate predictions confined to the trivialities of man’s least imaginative moments and to the automatisms of persons given in despair.”

So how are we to get beyond these limitations in our symbolic systems and yet come to give an account of the human innovative process within them?

              Let us return for a moment to the idea of reflexivity: Chomsky (1965) has shown how a “finite means” can give rise to “infinite (in the sense of ‘uncountable’) creativity”, as long as the rules in the system have a “recursive” property. The point which Chomsky misses is that men not only “create” systematically according to their own rules -- they “create” the rule systems too! This type of invention (I would like to call it transcendental to contrast with systematic) proceeds not in terms of symbol manipulation, but by the invention of new activities, new “forms of life” within which others can share (and which later come to support a “language-game”). It is how man can modify his own actions without plans that is the problem; and this is the problem of free-will, volition and consciousness too.

              For the key to the puzzle, we can again turn to Vygotsky (1962, p. 90): “In order to subject a function to intellectual and volitional control, we must first possess it.” And it is instruction (of the child by the adult) that effects the transfer of spontaneous actions to the realm of the deliberate. Its effect is not to make the child do what he cannot, but to restructure what he can do already into new forms – and this first proceeds without language, and provides a frame of reference within which the particular language of the child’s culture can develop (Newson, 1969). However, once linguistic communication is underway it can function first, as the instrument (Vygotsky, 1962, p. 56ff) for the mother to cumulatively restructure the child’s actions, and later, for the person (that the child becomes) to restructure his own. As Goethe said: “Language makes people more than people make language.”

              This is obviously an area of investigation with undreamt of riches; the discussion above is little more than an assay which, in my estimation, indicates that full-scale mining operations are worthwhile. Its implication here is that there is a close involvement of construct theory with language and language development; that essentially what we think of as “free will” or “deliberate” action is to do, not with acting in accord with a construct-system, but with selecting or constructing that particular construct-system [end p.241] in the first place. Thus, although we disagree with Miller’s idea about the source of plans, the disagreement is only partial; we think that the ‘knowledge’ for plans comes from bodily activity and the structuring of it into component parts and the schemes for their organization and reorganization, from instruction.

CONCLUSIONS

Constructs, as we have already said, are binary distinctions. What I have attempted to do in the preceding sections is to construct at least two sets of generically (i.e., each new distinction arising out of the last) related construct systems for the analysis of behaviour and knowledge.

              Clearly, there is no monopoly on construct-systems, we must remember that constructs are just as much as all our other artifacts, constructs. And if we think that all human actions are guided by a definite and limited set of fundamental constructs which can be described, once and for all, we shall run again into all the difficulties from which Kelly tried to save us. Yet the choice of constructs is not at all an arbitrary one, and much of Kelly’s (1955) book is to do with such problems of choice at the personal level [as Kuhn’s (1962) is at the scientific level].

Current Issues

THE EXPRESSIBLE AND THE INEFFABLE: THE NATURE AND

DEVELOPMENT OF CONSTRUCT-SYSTEMS

“What must be the essential character of our knowledge both of ourselves and nature that it is expressible in language (in symbols)?” For Plato, the world of sense experience was a flux, but “true” knowledge was of “real” and unchanging things; it was infallible, it allowed one to do with an expectation of success what one could not do before. (It ought not to be forgotten that the Greeks were concerned, for the conduct of their practical affairs, to instruct their slaves appropriately-much as we are now concerned to programme our computers to relieve us, adequately, of some of our chores. ) Plato thought that true knowledge could be arrived at by viewing the flux against an unchanging set of limits or ideal forms. And this set the philosophic task for centuries after: To discover the nature of such forms. Thus it is that we still ask such questions as “What is perception ?”, “What is motivation?”, etc., expecting to be answered with a set of templates which will cover all possibilities.

              In Fig. 1, I have depicted the results of a Kelly-Heraclitian process of construct-system development. As we have already discussed, people en- countering the flux about them may sense it in a differential manner, and furthermore agree amongst themselves upon ways of indicating this. Taking [end p.242] system A, on the left, essentially what it depicts is that a group of people encountering the flux, f, have found in it a difference and have invented a distinctive indication, A_l, which allows them to show how to mark off f into two distinct regions f’ and f’‘. Noticing that there is a difference must come first, agreeing on its nature and representing it next, and realizing that the two distinct parts are still related last. After having made A_l they next, say, notice a difference within f’‘ and mark it with the term A_2. But again this involves the process of first simply sensing differences, then agreeing

 upon which are of significance and representing them symbolically, and only finally realizing that what has been made separate is still related, now much more complexly, to its opposites. The process of noticing differences and so on, in l other words, the process of developing construct-systems may, presumably, be continued indefinitely. But note this, once a distinction has been made, we may quite reasonably choose to ignore it, thus, although the diagram depicts, from top to bottom, systems of increasing differentiation, the lower are generically related to the higher and are logically included in them.

              The system so produced is best described as a system of compartments , [end p.243] where the compartments are distinguished from one another in terms of binary distinctions, i.e., constructs, and where each compartment may be identified by the relation that it bears to all the others in the system. This is where the repertory grid comes in: We might like to name the compartments, in the case of system A, I have named them, arbitrarily, a, b and c. Aspects of the flux in which we can sense Al but not A2 qualify to be called a’s, those without Al but with A2 qualify as b’s, while those with neither as c’s. (Not a very rich system as depicted here perhaps, but with eight binary distinctions (256 possible compartments) we can encompass the sound system of a language (about 40 distinct phonemes) with about 50% redundancy.)

              To hark back to Heraclitus on Protagoras: a set of constructs is obviously like a “table of opposites” ; however, a construct-system demands not just a list but properly inter-related constructs.

              While each compartment is, in isolation, uncharacterized, its nature is known in terms of its relations to all the other compartments in the system. [And we might, as Kelly (1961) did, imagine an algebra of such systems.] However, what it is important to realize is that constructs are not passively acquired, but result from action of some kind, and that gaining conscious possession of a construct involves the social psychological process discussed on p. 239. When it comes to construct-systems in science, this process is not just involved in the making of a construct, but is involved in the making of a whole system. The task of science (as distinct from technology) does not lie in just discovering new facts, but in discovering an orderly arrangement for many, many facts, one which depicts the nature of all their interconnections.

              Now, as has been mentioned before, what is observed depends upon what is done. In the diagram I have shown two construct systems A and B which have been imposed upon the same flux, f . To take a particular example, imagine the “flux” to be the unanalysed “proto-noise” of speech. The construct system A might be the distinctive feature analysis of Jakobson et al (1952), where speech sounds are distinguished in articulatory terms: vocalic/non-vocalic, consonantal/non-consonantal, compact/diffuse, etc. While the construct-system B might be a physicist’s analysis of the proto- noise in terms of: high/low frequency, high/low intensity, short/long duration, etc. The two sets of objects (a, b, c) and (x, y, z) so produced bear no relation to one another having been obtained by different means of analysis serving different purposes; there being no general or “neutral” analysis to be had as there is no general or superordinate communal activity [end p.244] from which it could arise. So while both phoneticians and physicists are talking about what we obviously feel are the same noises, they are both view- ing them through their own “paradigms” (to use Kuhn’s term), their own research training and practice, or their own scientific “form of life” with its associated “language-game(s)” to use Wittgenstein’s (1953) terms. Kelly talked of a construct-system as having a range of convenience; I hope the discussion above goes some way towards explicating that idea.

              Now, as what scientists observe depends upon what they do, the intelligibility of the account they give depends upon whether one is a participant in the appropriate scientific community, for their accounts are only of relevance (i.e., of use in exploiting the possibilities inherent in them) to those who possess the same means for affecting the world as those who originally produced them.

              This seems to suggest that explicit construct-systems follow from a particular orderly activity (see suggestion 2, p. 237), and need not be at all implicated in that activity. This raises a most important issue. A child does not first learn the distinctive features (constructs) of his language in order to decide what is the right way to talk: he must first ‘talk’, try to talk, in order to find out what the distinctive features of his language are. [And he seems to go through the sort of development depicted in the diagram, i.e., his early constructs are different from his final ones, but generically related to them. The process of differentiation seems to stop when he has made enough distinctions for all practical purposes (Brown, 1965, Ch. 6).] Thus an analysis, in terms of some final set of constructs, such as Jakobson et al.’s, cannot be involved in the process of learning to talk, it can only be constructed and imposed as a consequence of knowing how to talk. Talking is learned by trying to talk to an adult, by bodily activity being shaped under instruction. [Here we are very close to Skinner (1957) as I shall acknowledge in a moment, p. 249.]

              It is because of this that we cannot make machines, which lack bodies and thus non-linguistic means of communication and so on, learn to talk as we do, no matter what this author (Shotter, 1968) once thought.

              Unfortunately, we cannot spend time working out the implications of this diagram here, though it obviously has many to do with systems of knowledge, concepts, universals, etc. But it is clear, I hope, that the answer to our initial question is: “The essential character of the knowledge we encode in symbols is that it is expressed in terms of differences, i.e., it is relative, and based in ‘knowledge’ derived from our direct encounters with the “world”.

OBJECTIONS TO THE IDEA THAT EVERYTHING ESSENTIAL TO

HUMAN INTELLIGENCE CAN BE FORMALIZED

It might seem that Chomsky’s (1957, 1965, 1968) brilliant analyses of linguistic [end p.245] rule-systems has, along with our linguistic competence, brought the whole nature of our competencies as such within the realm of logic and computa- tion. At the moment Chomsky (1968) is “assigning to the mind, as an innate property, the general theory of language that we have called ‘universal grammar’.”

“. ..very deep inborn principles, probably of a highly restrictive nature, that determine how knowledge of language emerges in the individual, given the scattered and degenerate data available to him (Chomsky and Hampshire, 1968) ...the general features of language structure reflect, not so much the course of one’s experience, but rather the general character of one’s innate capacity to acquire knowledge -- in the traditional sense, one’s innate ideas and innate principles” (Chomsky, 1965).

And Chomsky seems to think that in some way these very deep inborn principles can be characterized. As indeed he must if his ideas about a “universal grammar”, and language acquisition, and so on, are not to be vacuous. This is a powerful claim and if correct would have startling consequences, if not in all of Psychology , at least for construct theory : If Chomsky is right, and our linguistic (and possibly other) competencies are so structured because of “innate mental dispositions,” then the consequences are exciting. For it would suggest that the form of our knowledge is limited in discoverable and well-defined ways (Chomsky, 1965, Chomsky and Hampshire, 1968), i.e., that a basic set of constructs could be found from which all others could be derived.

              However, if such people as Kelly and Wittgenstein are right, languages are not built upon such constructs, constructs are a consequence of the process. Languages appear to share deep common characteristics simply because all peoples of the world share the same basic bodily form, live in the same world, discover their capacities to themselves in the same way, and so on. In other words, the fact that our linguistic structures seem to be limited in well-defined ways is no more remarkable than that we locomote on our feet rather than on our knees and elbows.

              The activities that we carry out in the social sphere do not just depend upon the “dispositions” of the “mind”, but upon the whole person. Chomsky’s constructs are constructs; they can only be imposed upon a language by someone who already knows that language. Grammatical transformations are, in his own words “structure dependent, in the sense that they only apply to a string of words by virtue of the organization of those words into phases” [end p.246] (Chomsky, 1968). That is, they can only be applied if it is already known how to construe sentences into phrases.

              Obviously language is innate in the sense that it has come out of man’s own resources, but we inherit its structures not like blue eyes but like all the other constructs of man.

MENTALISM, MECHANISM, BEHAVIOURISM, SOCIAL INTERACTION

“. ..words are connected with the primitive, the natural, expressions of the sensation and used in their place. A child has hurt himself and he cries; and then the adults talk to him and teach him exclamations and, later, sentences. They teach the child new pain behaviour. “So you are saying that the word ‘pain’ really means crying?”- On the contrary: the verbal expression of pain replaces crying and does not describe it” (Wittgenstein, 1953, no.244).

For Behaviourists to attempt to banish mentalistic terms from their vocabulary-terms that they called “mentalistic” because they were supposed to refer to private inner states -- was quite unnecessary.

              When as an adult we refer to ourselves by use of such terms, we do it against a personal background of thought and feeling which we must recognize as familiar in some way. Thus there is for us, we feel certain, a particular private inner state associated with the use of such words, and it is to that which we think they refer. Brit that is not how we were taught them; that is not how the words were used by those who taught them to us. They observed our behaviour and used such terms to refer to us, not to themselves. It was a case of “adultmorphism” for Behaviourists to think that the epistemologically prior use of mental terms was for self-reference. Further- more, it was yet another example of the “double-viewed” approach, to use such terms in setting the goals of a Behaviourist psychology-an account of “mental” activity in terms of S-R laws. They did not realize that even in their very contention that such laws were “objective” or “public” they were appealing to implicit “rules of social conformity” involving just the process such “laws” were meant to illuminate: perceptual judgements, agreements, their symbolic representation and so on.

              Kelly’s Psychology is not mentalistic, quite the opposite. For him we do ‘talk’ to one another with our behaviour.

“Two people, say a mother and a newborn child, may not have a full intellectual meeting of minds the first time they try to enter into a discourse with each other in the maternity ward. But by sharing their encounter with events-including the events produced by their own behaviour-some mothers and daughters do develop a fair understanding, each of what the other is talking about” (Kelly, 1969, p. 28).

[end p.247]

And it is no exaggeration to say that in fact Kelly’s is a thoroughgoing behaviourism [like George Mead’s (1934)], so thoroughgoing that one cannot use it to find out within it anything about man’s very fundamental natures any scientific accounts that we might achieve we will still have invented! But yet it deals with just those activities of man which are thought of as intellectual, conscious and deliberate.

              In relation to this, turning now to the problem of mechanism, it is noteworthy that it is just the “higher” mental functions of man that have proved easiest to simulate upon a computer. Something, such as the extensive recognition of friends faces, which even what we think of as the most moronic among us can do, is very difficult to simulate-if not impossible for it to be done in the same sense as we do. In Wittgenstein’s terms, the language-games of our “higher” mental functions, e.g., mathematics and logic, arise out of extremely circumscribed and orderly forms of life: those that revolve around the arrangement and re-arrangement of patterns of marks on paper. That such orderly patterns of social interaction should give rise to activities which can be mechanized is not, then, surprising. A machine to do these tasks could be an expression of those very same inventive impulses that originated the social interaction in the first place. Thus it is not at all remarkable and we should expect to be able to invent a rational description (a mathematical model) for all our rational activities, as Chomsky has done for those aspects of our language usage that Bernstein (1964) would call “elaborated”.

              Thus we may only be able to simulate on a computer, in any genuine sense, those of our activities which we ourselves have both invented and, most importantly, closed to any further inventive development.

              Comments such as these, I think, should make us suspect that there is more to being intelligent than having a “mind” (and that computers are more like electronic “minds” than “brains”).

              The simulator’s dreams of designing “a species of superior intelligence to replace ourselves as lords of the Earth” (Sutherland, 1968), seem to be nothing more than thinly disguised “death-wishes” to be able to escape from the responsibilities of being able to act consciously and deliberately by choice.

“No wonder so many of us would like to become scientists and be content to win prizes without having to take the awful responsibility for people (Kelly, 1955).”

The Frankensteinian-myth, of making creatures more powerful than our selves but which still remain within our control, is obviously very powerful. And scientists’ activity no matter what public relations story they put out, is still clearly tinged, unavoidably, by mythico-magical thinking.

              For example, computer programmes must be written in ‘words’, in symbols [end p.248] -- whether these are patterns of punch holes on paper tape, patterns of magnetization in core-stores, or whatever. Is it not “word-magic” to hope that one day, if the pattern is long enough, or complex enough, we shall hit upon the right combination to make the machine “intelligent”? Also, if “principles of reinforcement” (Skinner, 1957) were all that there was to learning language, we would have to accept, as Locke said, that the origin of language really was a shaping process by Nature: “. ..nature, even in the naming of things, unawares. ...” If we call that a myth, a story we tell ourselves to make the world and ourselves intelligible, then the notion of “innate ideas” is no less a myth. And in the absence of any alternatives, such myths guide us in our corporate affairs until we can invent more empirically well-founded alternatives. There are, I think, reasonable grounds for pre- ferring Kelly’s myth.

              Finally, in this section, I would like to give some indication of how one can proceed in a behaviouristic fashion while investigating the most complicated of topics: the instruction of a young child by its mother.

I am fortunate enough to be involved in a project at Nottingham investigating pre-verbal communication between mothers and their 10-20 months old infants. I feel that we are beginning to uncover the processes at work by which they create a mutual construct-system between them before language proper develops. We are simply watching {and video-taping) how mothers go about trying to get their young children to do form-board tasks, i.e., putting different shaped pieces in their appropriate place, otherwise the situation is entirely unstructured by us.

“Mothers shape the child’s whole activity by selecting and emphasizing happenings which are task-relevant (as the mother sees it) at the expense of those which are not. In this sense the mother structures the child’s spontaneous activities in a way which he himself would not do if left to his own devices. Out of all this mutual activity, mother and child develop a shared frame of reference. Each knows the happenings to which the other is responsive, and they are thus in a position to communicate their intentions to one another. It may well turn out that this form of rapport is essential to the development of linguistic communication as a later stage” (Newson, 1969).

In other words, the mother is responsible for some of the very fundamental constructs available to the child. She, obviously, is not responsible for his activity and what he discovers to himself as result, but she is responsible for getting him to select among his actions.

              Clearly here, some version of a reinforcement theory is needed {and here we acknowledge Skinner’s ideas of “shaping” and “reinforcing”), but quite why a child is responsive to some of his mother’s actions and not [end p.249] others, why her “Ohoo ...o there’s a good boy” seems to relax him and terminate his ongoing activity, we only just now are working out ways to investigate.

              All we can say at this stage is that patterns of interaction are, in some degree, idiosyncratic. The mother must watch the child, and the child the mother, there are no pre-established rules, only continual improvisation. It is the creative exchange, par excellence, of people making people.

Conclusions

The essence of a recursive scheme is that it can give rise to an indefinite number of different forms. This is the nature of Kelly’s model of man; he may take on many forms. In no sense can Kelly’s Psychology boil down to a Galilean or Newtonian formula, quite the opposite: it gives us the beginnings of an account of how the scientific process functions, how rational thought can grow out of thoughtless and irrational activities; it suggests that rather than seeking after the one single “true” system we should proceed with a proliferation of construct systems, each, if they are of any relevance to man’s needs at all, with their own special range of convenience. As we discovered on p. 244 a particular analysis is achieved by a particular means for a particular purpose, and the analysis is only intelligible and of use to those who under- stand and possess the means via which it was originally achieved. Thus if a construct-system is to be of any use at all it must bear a relation to the methods that people have available for taking some sort of action in the world. But even if it can meet these conditions, whether it should then be used (“should” always implies agreements amongst people) is another matter. Once we possess an accredited construct-system we can use it like we use a map, letting it dictate most of our actions, proceeding not in terms of a differential sensing of our environment, but by letting it determine a sequence of ready- made units of activity relevant to the achievement of certain purposes. Thus it is in the sphere of instrumental activity that construct-systems have their importance; however, it is in the sphere of non-instrumental activity (I shall for convenience call it “play”), that construct-systems are invented. Play is the sphere of activity where we are not under compelling external circumstances, they are such that we can “selectively non-attend”.

              To some the mere mention of the irrational, or non-rational, is to court disaster; here I have spent much time discussing its value. This is not because I wish to advocate the abandonment of rationality, far from it. I hope that if we can understand its roots more clearly we can transform it from a fragile plant, always fading away, to a robust one, a hardy perennial.

              What Kelly shows us is the nature of our actions when they are regulated by rational schemes, what he cannot show us is how to choose between alternative schemes; this sort of intelligent and responsible choice cannot [end p.250] itself be encompassed within any rational scheme, our individual choice is still important.

              The consequences of Kelly’s Psychology in the long run are quite unforeseeable, and not without its dangers because of that. We should not, I think, as psychologists, as a specialist group with specialist methods, make any sort of prescriptions about the possibilities we might invent. We must appreciate that for a person to ‘see’ a possibility as relevant to himself, it must have a basis in his own preferred construct-system. Presenting people with new options between which to choose but with no basis upon which to choose, can only lead to what might be called “the existential horrors”.

              In summary then, for the three areas of concern, Psychology, Science and People, we can say: In Psychology we have distinguished between the ‘knowledge’ which we derive from our actions and the schemes of organization, derived from social interaction, which we can impose on it. Thus that which the child learns about his physical world is different from that which he learns of human inventions; he has to re-invent (Chomsky, 1968) the latter but discover (Skinner, 1953, 1957) the former. (Thus it is not so much that Chomsky’s and Skinner’s views are mutually exclusive; they have not admitted that their construct-systems have a limited range of convenience.) Psychology must find a way of illuminating both these areas; it cannot, just because it arbitrarily chooses to ape the natural sciences, ignore our inventive activities.

              For Science, the implications are that it cannot provide a one single “true” account of “reality” with pure science to prove the view, and applied science to exploit it. The essence of Science is that it should invent construct-systems which, ultimately, make empiricism unnecessary (whereas Technology exploits the empirical method). Science is to do with the creation of networks of relations which can be arranged and rearranged according to determinate rules thus allowing us to derive possibilities from actualities. The different descriptions of phenomena, the possible derived from the actual, are ‘maps’ to use in our investigations; they tell how, even though we may do different things, “how to go on in the same way”. Thus the question is not of which view must be logically true, but of whether we can use it, not just in the material sphere but in the personal sphere also. But a description is only of such use if it relates in some way to the methods that people have available for taking some sort of action in the world, that is, relates to their ‘knowledge’ derived from bodily activity.

              And finally for People, as we have already said on p. 225 their most important possessions are their construct-systems of themselves; of themselves in their culture; and of their culture in the world at large. Such construct- systems tend to be self-fulfilling as they determine peoples’ rational and deliberate ways of going on. To meet new circumstances (which we as a [end p.251] consequence of our one actions create !) we must continually transform and re-create ourselves. By not realizing the nature of this process and thus not having proper provision for it in our accepted public images, we seem to have allowed our self-image to fragment, and in the process, ourselves.

              My purpose in this essay has been to try to show that it might be possible to combine Kelly’s Psychology with insights from the writings of Wittgenstein, Mead, Vygotsky and Cassirer to produce a new self image of man, for man, which will integrate his scattered fragments.

References

Bernstein, B. (1964). Elaborated and Restricted Codes, (Gumperz, J. J. and Hymes, D., eds.). Am. Anthro. Sp. Pub. 66, 2, 55-69.

Beadle, G. and Beadle, M. (1969). “The Language of Life”, Panther Books, London.

Broadbent, D. E. (1961). “Behaviour”, Eyre and Spottiswoode, London.

Brown, R. (1965). “Social Psychology”, The Free Press, New York.

Cassirer, E. ‘1944). “Essay on Man”, Yale University Press, New Haven and London.

Chomsky, N. (1957). “Syntactic Structures”, Mouton, The Hague.

Chomsky, N. (1965). “Aspects of the Theory of Syntax”, M.I. T .Press, Cambridge.

Chomsky, N. (1966). “Cartesian Linguistics”, Harper and Row, New York.

Chomsky, N. (1968). “Language and Mind”, Harcourt, Brace and World Inc., New York.

Chomsky, N. and Hampshire, S. (1968). A Universal Grammar, The Listener 79,687-691.

Davis, M. (1958). “Computability and Unsolvability”, McGraw-Hill, New York.

Dreyfus, H. L. (1967). Why computers must have bodies in order to be intelligent, Rev. Metaphysics 21, 13-32.

Hobbes, T. (1651). “Leviathan”.

Jakobson, R., Fant, G. and Halle, M. (1952). Preliminaries to Speech Analysis, M.I.T. Ascoust. Lab. Rept., 13, (2nd Ed. 1963, M.I.T. Press, Cambridge).

Kelly, G. A. (1955). “The Psychology of Personal Constructs”, Vols. 1 and 2, Norton, New York.

Kelly, G. A. (1962). The abstraction of human processes, Proc. XIV International Congress of Applied Psychol. 2, Gerhard, Neilson, Copenhagen.

Kelly, G. A. (1964). The Language ofHypothesis: Man’s Psychological Instrument. J. indiv. Psychol. 20, 137-152.

Kelly, G. A. (1969). A Mathematical Approach to Psychology, 1961, read to Moscow Psychological Society, April 10, 1961 and Ontological Acceleration, In “Clinical Psychology and Personality: The Selected Papers of George Kelly (Maher, B., ed.), Wiley, New York.

Kuhn, T. S. (1962). “The Structure of Scientific Revolutions”, University of Chicago Press, Chicago.

Locke, J. (1690). An Essay Concerning Human Understanding.

Luria, A. R. (1961). “The Role of Speech in the Regulation of Normal and Abnormal Behaviour, Pergamon Press, London.

Mead, G. (1934). “Mind, Self and Society”, University of Chicago Press, Chicago.

Miller, G. A. and Chomsky, N. (1963). Finitary Models of Language Users, In “Handbook of Mathematical Psychology”, Vol. 2 (Luce, Bush and Galanter, eds.), Wiley, New York.

Miller, G. A., Galanter, E. and Pribram, K. H. (1960). “Plans and the Structure of Behaviour”, Holt, Rinehart and Winston, Inc. New York.

Mumford, L. (1967). “The Myth of the Machine”, Secker and Warburg, London.

Newell, A., Shaw, J. C. and Simon, H. A. (1958). Elements of a theory of human problem solving, Psychal. Rev. 65, 151-166.

Newson, J. (1969). Comments on the nature of pre-verbal communication between mothers and their children in the age range 10-20 months, Mimeo, Department of Psychology, Univ. of Nottingham.

Piaget, J. (1923). “The Language and Thought of the Child”, Routledge and Kegan Paul, London.

Piaget, J. (1952). “The Origins of Intelligence in the Child”, Routledge and Kegan Paul, London.

Piaget, J. (1968). “Six Psychological Studies”, University of London Press, London.

Ryle, G. (1949). “The Concept of Mind”, Hutchinson, London.

Saussure, F. de (1960). “Course in General Linguistics”, Peter Owen, London.

Shotter, J. (1968). A note on a machine that ‘learns’ rules, Brit. J. Psychal. 59, 173-177.

Shotter, J. (1969). What can programmed minds do ? In “Aspects of Educational Technology III”, (Mann, A. P. and Brunstrom, C. K., eds) Pitman, London.

Skinner, B. F. (1953). “Science and Human Behavior”, Macmillan, New York. Skinner, B. F. (1957). “Verbal Behavior”, Appleton-Century-Crofts, New York.

Sutherland, N. S. (1968). Machines like men, Science J. 4, 10, 44--48.

Turing, A. M. (1936). On computable numbers with an application to the Entscheidungs problem, Hrac. Land. Math. Sac. ser. 2, 42, 230-265.

Vygotsky, L. S. (1962). “Thought and Language”, M.I.T. Press, Cambridge.

Wittgenstein, L. (1922). “Tractatus-Logico-Philosophicus”, Routledge and Kegan Paul, London.

Wittgenstein, L. (1953). “Philosophical Investigations,” Basil Blackwell, Oxford.

Notes: