give to the latter a very marked emotional tone, and hence lead to affective inhibition where it is concerned. Doubtless this accounts for the fact that some children who do well in an intelligence test may be making progress in school at a slower rate than their success in the test would lead one to expect. These considerations reveal a merit of the intelligence test which makes it one of the children's safeguards. The writer numbers among his own cases examples both of specific and of general inefficiency in school work which investigation showed clearly to be due to causes of an affective character. Yet the intelligence test penetrated these inhibitions and revealed the true grade of mental capacity to be superior to that manifested in school.

In addition to the inherent nature of the tests themselves, it is necessary to consider for a moment the conditions of their administration. The individual tests are given to one child at a time, and, in general, no one else is present but the examiner. Evidently, such conditions may in certain circumstances, and with children of certain temperaments, be conducive to the setting up of inhibitions. The removal or prevention of inhibitions arising in this way depends for its success almost entirely on experience in the general technique of administering the tests; and the skilled tester (who, it need hardly be said, should be a sufficiently trained psychologist) will quickly recognise symptoms indicating the presence of inhibitions and will know how to deal with the latter. It may safely be said that, with a tester who understands his work, any serious effect due to affective inhibitions arising from the conditions of administration of the tests can be practically eliminated.

When we turn to consider group tests, however, the case is somewhat different. It is true that, among the large number of items which go to make up a group test, there may perhaps be a few which possess marked affective tone for certain children. There seems no reason to suppose that the effect of this, if it exist at all, will be more than slight; and, in any case, the very multitude and variety of the items is likely to ensure that any affective consequences will average out among the different children so as to leave unaffected the reliability of the norms of performance as a basis of comparison.

More serious, however, in the case of group tests, are the possible effects of the conditions of administration. In the majority of group scales the separate tests are given out at intervals, each, after brief verbal instructions from the examiner, being carried out under a strict time limit. The possible effect on certain types of child (e.g. the 'nervous' or 'anxious,' or even 'cautious' child) of this state of affairs can easily

be imagined, and the conditions of the test necessarily render it impossible for the examiner to eliminate disturbing factors by dealing with the children individually. It does not seem possible to get rid altogether of the influence of such factors in a group test; yet, on the other hand, the results of the tests show that the mass effects produced by the disturbing elements are negligible, though, of course, injustice may be done to individual children. In the case of individuals, the effects are probably reduced to a minimum in those scales in which the tests are not timed separately but only as a whole, the child being left to work quietly through the questions by himself, without interruptions at intervals by the examiner1.

It will be clear, then, that an examination of the tests themselves reveals but little in their nature to warrant the conclusion that affective factors may frequently militate to a serious extent against a child's performance. But evidence of a more definite, and therefore more decisive, nature is found to be available when we turn to a scrutiny of the results of the tests. In considering these results we should expect to find the influence, if any, not only of specific inabilities with regard to particular tests arising from specific inhibitions, but also of general inefficiency due to affective factors of a more far-reaching type.

It was pointed out at the beginning of this paper that, when disturbing elements are eliminated, individuals will differ in their respective grades of mental capacity, intelligence tests being directed to the discovery of the index, constant for a given individual, which expresses his particular grade of intelligence. A child's degree of intellectual maturity is expressed as a 'mental age.' As intelligence develops with increasing age, so will the mental age increase. But the child's intellectual rank can only be expressed by comparing him with the 'average' child, and the result of this comparison is termed the 'mental ratio' (i.e. the ratio of the child's mental age to his actual age) or 'intelligence quotient' (I.Q.). There now arises the question whether the I.Q. can be taken as that constant index which denotes the child's grade of intelligence.

Let us suppose that the influence of affective inhibitions frequently and markedly interferes with children's performances in the tests. How could this be detected in the results of the tests? Evidently by retesting the same children after various intervals of time, some of the intervals being of considerable duration. We should then expect to find

1 Cf. for example, Dr Godfrey Thomson's Northumberland Mental Tests (Harrap), and the present writer's Simplex Group Intelligence Scale (Harrap).

that the results of retests of the same child differed much from one another; for we could hardly suppose that the effects of the inhibitions were so nicely proportioned quantitatively as always to produce the same proportional (not absolute) effect at any age (thus leaving the apparent I.Q. constant so far as they were concerned), especially in view of the fact that at later ages a child is tested by questions wholly or partly different from those employed at earlier ages. On the contrary we should expect the interference of the affective factors to be quantitatively unequal, and more or less arbitrary, and hence to find irregular variations in the I.Q.

What is found in practice? The very opposite is the case. There is a large and growing amount of data which shows that the I.Q. of a particular child, as measured by the tests, remains practically constant within the limits of experimental error. Retests have been conducted under varying conditions, by different examiners, and after varying intervals of time (some of several years' duration), and all go to reinforce the conclusion that the tests give a nearly constant I.Q. for each child1. There are no signs (except in markedly pathological cases) of the irregular variations which would inevitably arise from affective interference by inhibiting factors of the kind we are considering. Retesting therefore shows, not only that the I.Q. is at least a close approximation to the constant index aimed at, but also that affective inhibitions have at most but a negligible influence on the children's performances both in specific tests and in the scale as a whole.

We are thus bound to conclude that there is nothing in the nature of the tests themselves or in the results of their application which could warrant us in believing that affective inhibitions seriously interfere with the measurement of intelligence. At the same time we must end as we began by pointing out that finally conclusive evidence can only be obtained by retesting children after psychotherapeutic treatment. It is to be hoped that this important experiment may be carried out in the near future. Allied experiments of equal interest might consist in the application of mental tests during hypnosis, and also in cases of dissociated personality. In dissociations of the co-conscious type the comparison of the results of mental measurements made on the different personalities would be particularly interesting.

1 The only general exception to this statement occurs with mental defectives (especially the lower grades) where a noticeable drop in I.Q. seems to occur with increasing age.

39

SUGGESTION.

By J. CYRIL FLOWER.

THE Confusion which exists in psychological theories of Suggestion is due, no doubt, to a number of causes, chief among which is the fact that there is no common agreement as to the principle upon which definition is to be made. One method is to lump together certain kinds of reaction which are obviously similar, and to refer them without more ado to the capacious pigeon-hole labelled 'Suggestion.' This method is the carrying over of the rough and ready classifications of common sense into the field of psychology. What we mean in ordinary speech by suggestion is actually a group of mental operations and reactions which appear, as effects, to resemble one another. Thus if A in perplexity asks the advice of B, his question probably is, "What can you suggest?" And B is regarded as having made a suggestion, good, bad or indifferent, either (1) if he offers definite ideas, or (2) if by gesture, silence or other means, he insinuates something, vaguely or definitely. So a book or a play is said to be 'suggestive' either because it communicates fresh ideas and stimulates the thought and imagination of the public, or because it hints in veiled language at things which are more attractive in the disguise of innuendo than in the everyday garb of direct expression. The common element in these phenomena which are thus classed together as suggestion is the influence by one person or group of persons upon another person or group of persons which is exercised without recourse to physical force. Obviously if everything which comes under this broad heading is included under one term, the term will be so general as to have little or no value. A word that means too much may serve in common speech, where it can always be more nicely defined by its context, but it can only lead to hopeless confusion in science. If, therefore, the term is to survive in psychology, it must be defined by something which penetrates deeper than mere external resemblance. Accordingly there is a second method which proceeds by psychological analysis, and which aims at defining suggestion by reference to the psychological mechanism or mechanisms involved. But although psychologists are commonly agreed on the necessity of this exacter method, they are by no means agreed that its application leads to one precise and adequate definition

of Suggestion. Nor is the reason for this far to seek. Psychology is not a science of phenomena which can be isolated and studied as a selfcontained group of interacting elements. Psychology has to deal with mind and its manifestations, and mind cannot be temporarily detached from the world while it submits to analysis and experiment but has to be studied in relation to the world upon which it acts, and which acts upon it. Thus psychological analysis depends for its results upon the point of view from which it is made. The academic psychology which largely held the field till recently was primarily interested in mental processes so far as they could be described and formulated by the methods of introspection and observation, the assumption being that what appears to be the character of a mental process for an intelligent observer, is its character. This whole assumption, however, was challenged by the application of the idea of evolution to mind, and profounder results were at once obtained by the comparative method, which substituted for the mere personal approach, the biological approach to the problems of psychology. A further contribution of vast importance has come from the study of mental disease and irregularity. The situation, therefore, is that psychology can hardly be looked upon as an independent science, but rather as a handmaid to the sciences. 'Pure' psychology is an abstraction; real psychology is as various as are the fields of experience in which mind is involved. Every psychologist must, that is to say, deal with mind in the concrete, in relation to some definite situation or set of conditions, and the point of view from which he approaches the subject will necessarily influence all his conclusions. Thus the philosopher who studies man as a 'rational animal,' the biologist who studies him as the most recent phase of an evolutionary process, and the mental specialist who studies him as a complex of forces which may easily get out of order and have to be set right again, will inevitably analyse the psyche and draw inferences concerning it from points of view so different and with interests so diverse, that it is not surprising that the resulting 'psychology' is not always coherent. The whole topic of suggestion is hedged around with precisely these difficulties. Rationalistic psychology does not much relish the facts of suggestion, and accordingly it dismisses such facts as it cannot otherwise dispose of to the realm of 'the abnormal'-one of those vague terms which covers a multitude of ignorances. Biological psychology welcomes suggestion, because it likes to discover how very irrational man is, and how nearly related therefore he is to the ape, the dog, the tiger and the jackal, and accordingly it gives the widest possible scope to the operation