Intelligence and Its Measurement

L. L. Thurstone
Carnegie Institute of Technology

What is meant by intelligence?

Intelligence as judged in every-day life contains at least three psychologically differentiable components : a) the capacity to inhibit an instinctive adjustment, b) the capacity to redefine the inhibited instinctive adjustment in the light of imaginally experienced trial

( 202) and error, c) the volitional capacity to realize the modified instinctive adjustment into overt behavior to the advantage of the individual as a social animal.

The inhibition of an instinctive adjustment involves the substitution of an overtly passive deliberative attitude against environmental, social, and instinctive pressure. The degree of intelligence is partly measurable by the relatively early inhibition of the instinctive adjustment at a stage when it is not yet defined.[1] Conceptual thinking is rendered possible by the inhibition of the instinctive adjustment at a very early stage of its formation. A concept is an unfinished act inhibited before it has become personally concrete and it therefore has more extension and less intention than the idea which is an act inhibited after it has become motorially more defined. The capacity to redefine the inhibited instinctive adjustment in the light of imaginally experienced trial and error is essentially the analytical element of intelligent work. In routinized intelligent work the mental trial and error experience may be rationally controlled, but in unroutinized intelligent work the mental trial and error experience has its causal antecedents in loose affective and conative analogies. These subconscious analogies of intelligent work constitute alternatives of less resistance at a preconscious stage of the unfinished, undefined and inhibited instinctive adjustment. Analytical capacity may be thought of as flexibility of affective and conative associations which precede the cognitive phase in the definition of the stimulus into the response.

These three phases of intelligent behavior may be labelled for convenience a) inhibitive capacity; b) analytical capacity, and c) perseverance. One may possess the inhibitive capacity without keen powers of analysis. This leads occasionally to an appearance of intelligence by silence. Such a mind is often capable of recognizing and adopting the successful analyses of other minds without being able itself to produce the analyses.[2] The inhibitory capacity for the early, abstract, impersonal stage of instinctive adjustment does not necessarily coexist with inhibitory capacity for the later more defined and personally significant stage of the instinctive adjustment.

( 203) Hence a man may possess inhibitory capacity for abstract impersonal thinking and lack self-control for those instinctive adjustments which have reached uninhibited a personally and socially significant and overtly defined stage. A man may possess considerable inhibitory and analytical capacity and still be socially quite ineffective because of insufficient volitional drive to realize his intelligence into consistent action. Such a mind is not infrequently the subject of ridicule from those who would justify their own mental worth in spite of modest analytical powers.

Analytical capacity concerns the preconscious and early conscious stage in the definition of instinctive adjustment. It implies the capacity to inhibit instinctive pressure at that stage so as to allow flexibility of definition of the response. This has been picturesquely called by Knowlson free interplay with the subconscious.[3] It is only natural to expect that such inhibitory capacity should appear when the instinctive drives are relatively weak. Hence superior analytical ability is not infrequently associated with inferior instinctive energy in its nutritional, sexual, gregarious, and self-assertative forms. When such an individual discovers his inferior instinct equipment he may center his self-satisfaction on his analytical powers as a compensation. Such individuals need the assistance of other volitionally stronger minds in order to make their labors socially effective.

When the analytical capacity is associated in the same mind with a fair degree of instinctive and volitional energy we have the bright person in practical life. There is a fundamental distinction between bright and profound intelligence. The mind which possesses a fair amount of inhibitive, analytical, and volitional capacity will in general aim for effectiveness in the immediate present. Such a mind is bright. When, however, a mind is more generously endowed in all three of these respects or when it is unbalanced by reduced volitional capacity, it deals with the earlier and more abstract relations in its adjustments. Such a mind is profound and its analyses concern temporarily remote ends. I am quite sure that our intelligence tests in which the candidate races against time for a few minutes do not measure adequately the more inhibited and deliberative profound type of intelligence. It does fair justice to the more frequent well balanced bright mind.

( 204)

I should summarize my definition of intelligence as follows : Intelligence is the capacity to inhibit instinctive behavior in fin unfinished stage of its formation and to modify it at that stage by means of an imaginal stimulus which is relatively remote from that which is immediately and perceptually present. The imaginal stimulus in intelligent behavior is of preconscious origin. It becomes conscious as an alternative to be controlled, accepted or rejected rationally, but its origin is a preconscious, irrational, uncontrolled association of affective and conative similars. This definition is consistent with that of Baldwin in his discussion of meaning in which he says : "It is in the passage from the bare recognition of each item presented as being just what it is, to its treatment as being in some sense not what it is, but what it may become or be used as, that psychic meanings as such arise." Bergson says that "intelligence, considered in what seems to be its original feature, is the faculty of manufacturing artificial objects, especially tools to make tools, and of indefinitely varying the manufacture."

There are two features in intelligent behavior that I have tried to emphasize, namely, that it implies inhibition and that it consists largely in rendering conscious an unfinished act. The sooner the act becomes conscious, the more crude and unfinished it is when it becomes conscious, the greater will be the range of possible overt behavior into which it may become particularized, and the higher will be the intelligence of that particular moment. I have previously described this point of view with reference to the momentary psychosis.[4] The element of inhibition in intelligent behavior may be partly justified if thought of in relation to its various ramifications such as the inhibitory functions of the cortex, the recession of the stimulus in intelligent behavior,[5] the frequently drawn contrast between intelligent behavior and instinct or habit-determined behavior, the modification of instinctive protopathic behavior into intelligent epicritic behavior,[6] and the analytical intelligence of the inhibited introvert as contrasted with the analytical impatience of the extrovert.

By what means can intelligence be best measured by group tests? This question can be given at least three different interpretations.

( 205) We may set out to measure pure intelligence as a trait which is only partly responsible for social success. In this case we shall probably have no criterion by which to judge the test. We may set out to measure brightness using as our criterion the estimates of avowedly competent judges of brightness and dullness with the necessary admixture of volitional and emotional ingredients. We may set out to measure ability to do a specified practical task for which we assume that intelligence is an indirect partial requisite. This latter type of mental measurement is the customary one. Our criterion is the relative success of the candidate in school, college, salesmanship, clerical work, or what not. Any test procedure which has diagnostic value with reference to a specified criterion is acceptable. It is much more difficult to measure intelligence as such and I am not sure that it has ever been done. For administrative purposes, we do not need to measure pure intelligence. In fact our diagnoses are probably more effective if we test other significant factors in our tests, although it would be scientifically interesting to be able to tease out the intellectual, volitional, affective, social, heredity, and experiential elements.

It would be well if we could draw a sharp line of distinction between service and research in mental test work. These two purposes are often confused. The results of research may be summarized for direct application in some mental test service. Some research data may be obtained from service records but the reverse is rarely successful. Some mental testers are justifying their use of a test in test service simply because they have norms for it. This should never be done. No test should ever be used for any kind of service unless it is known what it is that the test diagnoses. This necessitates a criterion for every test. If the test correlates well with a criterion such as chronological age, it can be used to determine age. But a test may be good for one criterion and poor for another criterion. We should never talk about a "good" test without telling what it is that the test is good for—namely, the criterion. Another relatively common mental test illusion is that if a new test has a high correlation with the Army Alpha, and if the Army Alpha is good for some specified criterion, the new test is forthwith assumed to be good for that criterion too. This does not follow. The only safe way is to take the trouble to determine the diagnostic value

( 206) of the new test without inferring it through another test as an intermediary.[7]

It has always seemed to me rather peculiar that we measure children's intelligence in terms of their chronological age. I should like to suggest that possibly the underlying assumptions would be less troublesome if we stated the intelligence of an eight-year old in terms of his percentile rank with reference to other eight-year olds, or in terms of the standard deviation of test scores for eight-year-old children. From such a direct rating one could of course readily ascertain that chronological age the mean intelligence test score of which any given child attains.

What are the most crucial next steps in mental test research? Now that we have accumulated enough norms and correlation coefficients to make one dizzy it is about time that we begin to formulate some mental test principles. Our literature abounds with instructions for giving particular tests and with statistical data of administrative interest. But we have so far no science, no principles, no psychology in the mental test literature. We have done well with our empirical cut-and-dry methods in mental test work. We can do better if we rationalize our findings in this field. Unfortunately there are relatively few mental testers who are interested in deriving psychological fundamentals from mental tests.

I should like to see another line of mental test work opened up, namely, the diagnosis of the volitional and emotional characteristics which determine our character traits. Intelligence is only one of the elements in mentality and it has been overworked because of being accessible to measurement. We should investigate the possibility of diagnosing character traits by some new kinds of mental test, self-analysis forms, and other procedures. Who knows but that individual differences in the various characteristics of our reflexes may be diagnostic of character traits? Why is it that executives are in general taller and heavier than other people with similar habits of life? Why is it that we guess right more often than wrong about the character traits of a stranger by merely looking at his face? These questions should be rationalized as science. What are the possible schemes by which we may successfully classify character traits? What types of mentality are produced by the combination of intelligence and different emotional and volitional traits?'

( 207) By what physiological and psychological technique can we diagnose these types? What are the mental processes that distinguish special ability? Can these be taught to others? These are worthy problems for mental test psychologists.

It is high time that we quit justifying ourselves as psychologists by simply standardizing mental tests. If we attack the individual diagnosis of character traits as energetically as we have been giving group tests the results will be of far reaching psychological, educational and social significance.


  1. By "early" I do not refer to childhood or the education of youth. I mean by "early" the incipient response as it is being consciously particularized in every moment of normal waking life.
  2. When such a mind is by superior social traits adapted for administrative work it will lean on competent assistants for analyses. This type of mind is common in industry and in a deplorably exaggerated form in politics. It was described even by Machiavelli in “The Prince."
  3. "Originality," by T. S. Knowlson.
  4. "The Anticipatory Aspect of Consciousness," Journal of Philosophy, Psychology and Scientific Method Oct. 9, 1919
  5. Holt : "The Freudian Wish and Its Place in Ethics."
  6. Rivers : "Instinct and the Unconscious."
  7. 'Yule : "Introduction to the Theory of Statistics," p. 250.

Valid HTML 4.01 Strict Valid CSS2