A Study of Children's Thinking

Margaret Donaldson

• Publisher: Routledge
• Year: 2013
• Print ISBN: 9780415263993
• eBook ISBN: 9781136420047
• Edition: 1
• Topic: Medicine
• Subtopic: Health Care Delivery

Tavistock Press was established as a co-operative venture between the Tavistock Institute and Routledge & Kegan Paul (RKP) in the 1950s to produce a series of major contributions across the social sciences.
This volume is part of a 2001 reissue of a selection of those important works which have since gone out of print, or are difficult to locate. Published by Routledge, 112 volumes in total are being brought together under the name The International Behavioural and Social Sciences Library: Classics from the Tavistock Press.
Reproduced here in facsimile, this volume was originally published in 1963 and is available individually. The collection is also available in a number of themed mini-sets of between 5 and 13 volumes, or as a complete collection.

CHAPTER 1

The Prediction of Ability

WE ARE ALL constantly making predictions. We predict that there will soon be snow; or that we shall not reach home before we run out of petrol; or that the government will be defeated at the next election. Also, however, we make predictions of a slightly different kind. We say, for instance: ‘Listen to that noise! There must be a mouse gnawing at the wood.’ This last example is different from the others in that we are here not really foretelling the future, except in so far as we are predicting what we should find if we were to look to see; and it would be possible to argue that this should not be called prediction at all. But it is important to realize that many of the predictions of science are of this latter kind. The laws of science often do not tell us about sequences of events following one another in time, so that knowledge of earlier ones enables us to predict later ones, but rather about interrelated happenings, so that knowledge of one enables us to know about others that are occurring at the same moment.

An analogy may help to make this point clearer. In recent years people have begun to study seriously what is now called ‘extra-sensory perception’ – awareness, that is, of things which have not been directly perceived by the senses – and it is claimed that there are two cases to be distinguished. One of these is ‘precognition’ or knowledge of events which have not yet happened. This is the counterpart of what many people think of as the only type of scientific prediction. But there is also ‘clairvoyance’ which is claimed to consist in a knowledge of events or states of affairs that are contemporary with the having of the knowledge but inaccessible to the normal organs of sense-perception. And this has its scientific counterpart as well.

It may seem a little fanciful to speak of science as an instrument of extra-sensory perception – and yet this, in a way, is what it is. It enables us to know about things and events which we have not perceived directly. And it has its ‘clairvoyant’ as well as its ‘precognitive’ side. In the former case, what is predicted is that, if you look to see, you will observe that some specified event is occurring or has occurred or that some specified relationships hold. It is then only the looking to see which is in the future.

In the scientific study of human behaviour we are concerned, as in all science, with prediction. Here, as elsewhere, attention has tended to concentrate in the past on precognitive prediction. It may be, however, that in this study the clairvoyant type of prediction will prove to be of particular importance. Perhaps some of the greatest practical consequences will rest in the end on increases in our ability to make inferences about a present state of affairs from rather limited evidence.

Within the class of precognitive predictions, there is a distinction to be made that deserves attention. This is the distinction which Karl Popper (1957) has drawn between what he calls the ‘prophetic’ and the ‘technological’ or ‘conditional’ predictions. The prophetic predictions are the foretellings of things which it is beyond our power to prevent or control, such as the coming of a typhoon. The technological predictions, by contrast, tell us by what means a given event can be brought about: they tell us what we must do if we want to achieve something. Popper's own example of a conditional or technological prediction is the statement that if we want to build a shelter to withstand the force of a typhoon we must build it in a certain way.

When Popper makes this distinction he is discussing and criticizing notions of history as a process of inevitable unfolding about which only prophetic predictions can be made. However, not only the history of mankind but also the individual history of each one of us has been held by some people to be, in certain respects at least, an inevitable unfolding. And there is nothing of which this has been claimed more forcibly than the development of intelligence.

The great early advocate of this view was Francis Galton. In 1869 he published his book Hereditary Genius, the main argument of which was that intelligence is largely independent of environmental variation and that in each one of us its development is governed by hereditary constitution.

But, of course, if the whole of our intellectual development were inevitably determined from the start, much of education as we know it would be superfluous. If, then, a prophetic view of intelligence is to be reconciled with the retention of schools, a distinction has to be drawn between what can be prophesied and what cannot; and this has commonly been done by means of the distinction between intelligence and attainment. Galton himself did not make much of this distinction, and indeed he came near at times to expressing the opinion that education was superfluous, at least for the most able; but those who have come after him and who have agreed with him in his conception of intelligence have for the most part seen attainment as subject to a limited measure of conditional prediction, the limitation being imposed by intelligence. Intelligence is then regarded as the ‘innate potential’, that which makes attainment possible. Education has only to make it actual – and can, indeed, do no more.

This view is much less widely held than it used to be and many educationists and psychologists seriously mistrust it; but it still underlies much educational thinking in Britain and in the United States. It is a curious fact that the country which is at present most influenced by a conception of history as an inevitable unfolding – namely, the Soviet Union – is much less influenced than Western countries by a similar conception of the development of the individual.

We might ask, then: what is the evidence one way or the other? And this question has been asked repeatedly. It may be more illuminating, however, to begin by asking not what evidence we possess but what evidence we would be justified in accepting as conclusive proof of inevitability; for the answer to this question may perhaps determine how efforts to get evidence ought to be directed.

For this purpose, let us suppose first of all that arguments about how to test intelligence were settled, so that no difficulties of measurement would be complicating the issue. What sort of evidence would we then require?

It turns out that this question is hard to answer. It would not be enough for our purpose to have evidence that one set of environmental variations had not affected the issue – not enough, for instance, to know that a pair of identical twins, reared in different environments from infancy and tested in adulthood, had proved in the end to be of very similar intelligence. It would not even be enough to know that a very considerable number of different sets of variations had not caused intelligence to vary.¹ For the claim we are considering is that nothing in the way of environmental variation (given, of course, an environment sufficiently favourable to sustain life) can cause intelligence to vary – and this is a very big claim. J. B. S. Haldane (1955) has remaaked that if we make a claim of this kind we are behaving rather like the physicists of a century ago who called certain gases ‘permanent’, meaning by this that they could not be liquefied. It turned out in the event that these gases were no more ‘permanent’ than any others: it was merely that the physicists who so described them did not know how to liquefy them. When they said the gases could not be liquefied they were describing their own incapacity.

Environmental variations that might conceivably affect intelligence are of so many different kinds that the task of testing their effects, even by very limited sampling, would present enormous problems and has certainly never been attempted. But even if it were done – and this is the point which most needs to be stressed – no evidence that development was unaffected by fortuitous environmental variation could ever establish the impossibility of affecting it by deliberate action in the light of real understanding. The history of the physicists and the gases is again relevant. The discovery of impermanence did not come in the end through a random sampling of all the innumerable possible conditions that might affect gases: it came as all discoveries come (though sometimes chance helps them) through careful research and the gradual growth of knowledge. It is in the same way that we shall discover, if we ever do discover, how to affect the development of intelligence. But meantime, there is no justification at all, and it is hard to see how there ever could be justification, for a claim that its development is inevitably predetermined and entirely beyond our control.¹

It is worth noting that this argument amounts to a rejection of any absolute and final distinction between things which can be conditionally predicted and those which can be prophesied, given that the criterion of a prophecy is our inability to affect the issue. This is because we can never be sure that our inability is final, and consequently anything which at one time admits only of prophetic prediction may come, with advance in our understanding, into some possibility of control. Popper would presumably accept this, because it is part of his own argument that the progress of our knowledge is itself something we can never predict. Thus it cannot be maintained that there is anything which we can never come to know – except for this one limit: that we cannot know what we shall know.

If the above arguments are sound, we must cease to claim that intelligence is innately determined – if by this we mean that environment cannot affect the issue.² In this case, though, what becomes of the distinction between intelligence and attainment? Clearly, the old basis for distinguishing the two has gone. On the view that intelligence was ‘innate potential’, its relation to attainment was that of ‘rendering possible’. But if we are not prepared to assert that the whole development of intelligence is innately determined, can we still hold to the view that intelligence is in some sense ‘potential’, or must we abandon that conception also? Can we give up trying to distinguish between what is ‘innate’ and what is ‘acquired’, yet still keep the distinction between ‘intelligence’ and ‘attainment’ and the notion that the one ‘makes possible’ the other?

There are two circumstances which would lead many people to give a negative answer to this question. In the first place, there is the old habitual association of the words ‘innate’ and ‘potential’. These two have been used together so regularly and for so long now that their inseparability is liable to be taken for granted. When words come to be linked in this way, it grows easy after a while to slip into the unreflecting assumption that the one notion implies the other, so that the possibility of retaining one and rejecting the other is not even considered.

But there is a second reason why the idea of intelligence as potential is in danger of rejection. There is a current tendency to have misgivings about the whole concept of potential ability and its value for any scientific study. This appears to arise from the idea that whatever is potential is quite unobservable, an idea which is seriously, and curiously, mistaken. We are constantly observing potential. When we say of someone, ‘He could be a strong swimmer, if he knew how’, we are making a judgement of potential that is based on direct observation of physical characteristics such as well-developed muscles.

It is of more than incidental interest to notice that the distinction between potential and realized ability is more likely to be overlooked by an English speaker because of the fact that in his language the one word ‘could’ is used in both senses: ‘He could be a strong swimmer, if he could swim’, we might say, though we would tend to avoid this awkward juxtaposition wherever possible. In the French language, however, the distinction is very neatly provided for by the words pouvoir and savoir. The English ‘he cannot swim’ is quite ambiguous. But the French clearly distinguish il ne peut pas nager (because of lack of potential, a radical incapacity) from il ne sait pas nager (because he has never learned, never realized his perfectly adequate potential for the achievement of this skill).

Now the question of whether the strength of a man's muscles is innately determined raises just the same difficulties as does the similar question about the development of his intelligence, and these we have already considered. What must now be observed is that our inability to prove the complete innateness of his muscle strength in no way prevents us from regarding that strength as potential. And we can regard it in this way without necessarily being in a position to say exactly what influences have been exerted on this development by the food he has eaten or the ways in which he has exercised his body. Our ignorance on these matters does not alter the fact that, in relation to the specific attainment of swimming, his muscles represent a potential of unrealized ability.

In a similar way, the concept of intelligence as potential ability is not necessarily bound to the concept of intelligence as innate, and we can perfectly well give up the latter without having to deny the notion that intelligence makes attainment possible.

There is, however, a difficulty still to be considered. In so far as intelligence is not an inevitable unfolding, it may very justly be claimed that it is itself a form of attainment. And if intelligence is allowed to be attainment, what becomes then of the distinction between the two?

This difficulty is not so great as it at first seems. The fact that man is an animal does not mean that we cannot distinguish him from other animals or that we may not sometimes be justified in using the word ‘animal’ in a way which contrasts with, instead of including, ‘man’. Similarly, then, there may be certain sorts of attainment which, because of some distinguishing features, can reasonably and usefully be given a special name: the name ‘intelligence’. And it would be in keeping with traditional usage if this name were reserved for attainments which can be shown to be of special value for the clairvoyant and precognitive prediction of other attainments. But Ferguson (1954) expresses an important truth when he says: ‘The concept of intelligence … is no longer a useful scientific concept except as subsuming some defined set of clearly distinguishable abilities.’

It has been argued in this chapter so far that we cannot prove that the development of intelligence is a process of inevitable unfolding. At the same time, it has been pointed out that this is by no means equivalent to an assertion that heredity does not affect the issue or affects it only a little. A sensible suggestion by Hebb (1949), which has been fairly widely adopted and has done a great deal to clear up confusion, is that we assume the existence of a certain genetic complement which is relevant to the development of intellectual power. This we call Intelligence A. We acknowledge, however, from the beginning, that we cannot directly test Intelligence A – that is, innate capacity in the strict sense. What we can hope to measure with some success is Intelligence B – the individual's effective developed intelligence.

Now the practical implications of giving up all claim to be able to test Intelligence A are worthy of consideration. So long as we speak or think as if we aim to find out about innate capacity we shall be tempted to try to use our test results for long-term prediction – to forecast, for instance, when a child is eleven or younger how he will perform in an examination that is five or six years away; or to forecast when a student enters a university how he will do in his final examinations. This is because innate capacity is something which lends itself to prophetic prediction, and prophetic predictions tend to be (though, as Popper points out, they need not be) long-term ones. Obviously enough, if we think we are dealing with an inevitable unfolding, we shall feel justified in making predictions about the fairly remote future. But if, on the other hand, we conceive ourselves to be dealing with a potential ability which develops and increases in the measure in which it is realized, we shall not be so likely to proceed in this way, for we shall understand that the child's abilities five years from now may depend not only on his present state but on what happens to him – and on what he does – in the interval. Consequently, we shall realize that where there is any proposal to select a child for one kind of education rather than another our prediction will affect its own fulfilment, for what the child's abilities will be five years from now may depend in some measure on what education we decide he is to have.

This sort of circumstance is one which all students of human behaviour have to be prepared to encounter, as Karl Popper points out. In the social sciences, he says, a prediction ‘may in an extreme case even cause the happening it predicts: the happening might not have occurred at all if it had not been predicted’.

How, then, would it be reasonable to make use of tests of intelligence if it is allowed that the development of ability may not be wholly predetermined, and unaffected by our deci...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Table of Contents
• Preface
• 1. The Prediction of Ability
• 2. The Notion of General Intelligence and the Choice of Test Items
• 3. The Study of the Processes of Thought
• 4. Matching Problems
• 5. Three-term Series Problems
• 6. Series Extrapolation Problems
• 7. Related Series Problems
• 8. Formal Deductive Reasoning
• 9. Conclusions
• Appendix I – A Further Study of Overlap Error in Three-term Series Problems
• Appendix II – Statement of Problems
• References
• Index