A. The Definition.
In order to direct attention to the quantitative description of intellectual deficiency which is here proposed, let us state the percentage definition in its most general form. Individuals whose mental development tests in the lowest X per cent. of the population are PRESUMABLY INTELLECTUALLY DEFICIENT, unless their deficiency is caused by removable handicaps. Above these is a group of Y per cent. within which the diagnosis of intellectual deficiency is uncertain on the basis of our present tests. The size of the presumably deficient X group is to be determined by the number of intellectually weak which society is at present justified in indefinitely isolating. The doubtfully deficient Y group should include all those who are so intellectually deficient as to be expected to need assistance indefinitely. The feeble-minded, or MENTALLY DEFICIENT, are those who require social care indefinitely because of deficiency in mental development. They include the X group, that portion of the doubtful Y group which is found to require isolation, guardianship or social assistance, and any others not detected by the tests but requiring prolonged social care on account of their failure to develop mentally. Under the principle which we stated at the close of the last section the combination of Y ability and persistent serious delinquency brings the case within the group presumed to be feeble-minded.
Besides the greater definiteness and significance of such a definition of intellectual deficiency, it affords the simplest practical criterion for determining the borderline of passable intellects with a scale of mental tests. A detailed comparison of the percentage plan with other forms of quantitative definition will be found in Part Two. We may note here, however, that it guards against a number of the absurdities of current descriptions of the borderline with measuring scales. It is a criterion which may be consistently applied to the borderline of both the immature and the mature. It may be adapted with comparative ease to any system of tests. It aids in comparing the frequency of intellectual deficiency among different groups, for example, among different types of delinquents, regardless of whether the investigators have used the same series of tests, provided only that each series has been standardized for similar random groups.
Any form of quantitative definition, on the other hand, involves certain assumptions which must be defended before it can claim to be of advantage for practical purposes.
B. The Assumptions of a Quantitative Definition.
(a) Deficiency is a difference in degree not in kind.
Fortunately the tendency to describe the feeble-minded person as if he were a different species from the normal has been definitely attacked by two noteworthy researches, that of Norsworthy ([159]) and that of Pearson and Jaederholm ([164]) ([167]). In these two investigations mentally deficient children either in special classes or in institutions have been compared with groups of normal children from the same localities on the basis of objective tests. The results are uniformly supported by numerous other studies of deficient and normal groups with the Binet and other tests. The conclusion is, therefore, thoroughly established that there is no break in the continuity of mental ability. It grades off gradually from average ability, and continually fewer and fewer individuals are to be found at each lower degree of ability. The borderline of deficiency will, therefore, not be a mental condition which clearly separates different kinds of ability, but a limiting degree of capacity to be decided upon by social policy in attempting to care for those who most need social guardianship. Since ability changes gradually in degree it is necessary to indicate a doubtful border region of degrees of ability on which expert judgment must supplement the test diagnosis. Below the doubtful region the diagnosis is clearly supported by objective test criteria, so that the only question to raise is whether the condition is caused by removable handicaps. The percentage definition thus strictly conforms to the best objective studies of mental deficiency in treating deficiency as a difference in degree.
It should, perhaps, be said that this view is in direct conflict with the opinion that mental deficiency is accounted for as a Mendelian simple unit character. The opposing view has been advocated by Davenport (95, p. 310) and others in the publications of the Eugenics Record Office, and accepted by Goddard (112, p. 556). It has been so fully answered by Pearson ([164]) and Heron of the Galton Laboratory ([127]) and by Thorndike ([198]) that there is no occasion to take up the question in detail. We seem to be reaching an understanding so far as our present problem is concerned. If the explanation of the inheritance of mental ability is through Mendelian characters, nevertheless intellectual ability is the result of such a complex combination of units that it may best be thought of in connection with the unimodal distribution of ability adopted in this study. No random measurement of mental ability has ever shown any other form of distribution.
The attempt has also been made by Schmidt ([179]) to find qualitative differences between normal and feeble-minded children by means of tests, and by Louise and George Ordahl ([162]) to find qualitative differences between levels of intelligence among feeble-minded children. While these studies are very suggestive in pointing out the tests which most clearly indicate differences between individuals, they seem to me to fall far short of showing that the qualitative distinctions are anything more than larger quantitative distinctions. It is not clear that the authors intended them to mean anything more than this, so these studies do not seem to conflict seriously with our assumption that intellectual ability grades off gradually and uninterruptedly from medium ability to that of the lowest idiot.