Social Attitudes
Mental Deficiency and Crime
E. H. Sutherland
Professor of Sociology, University of Chicago
Table of Contents | Previous
APPROXIMATELY 350 American studies of the intelligence of delinquents are accessible; these studies contain the results of psychometric tests of about 175,000 offenders. The present survey is an attempt to appraise this method of studying delinquents and to summarize the results which have accrued from its use.[1]
The first conclusion derived from this survey is that the proportion of delinquents diagnosed feebleminded has been steadily decreasing. In 1913 Goddard published in his Feeblemindedness a list of studies of the intelligence of criminals, in which the study at one extreme diagnosed 89 per cent of the group as feebleminded, the study at the other extreme diagnosed 28 per cent feebleminded, and the median or middle study, which will here be used as a simple index of the group, had 70. In 1915 an Ontario Commission on the Feebleminded presented a list of American studies of the intelligence of delinquents, in which the median had a percentage of 62 feebleminded. In 1918 Williams in The Intelligence o f the Delinquent Boy published a similar list, with a median of 34. In 1919 a Massachusetts Special Commission on Defectives published a list with a median of 28. In six years the median in these studies dropped from 70 to 28. The authors of the particular studies and the compilers of these lists asserted that these con-
( 358) -clusions were very conservative. The lists, however, were not fairly representative of the studies which were being made but tended to select those in which the percentages of tested deficiency were high. Table I is a compilation of all accessible reports of psychometric tests of delinquents from 1910 to 1928, with the exception of a small number of studies in which the delinquents tested were clearly not a representative sample of the institutional groups from which they were taken. This table shows clearly a decrease in the upper extreme and in the
| Years | Number of Studies |
Percentage Feebleminded in Median Study |
Percentages Feebleminded: Range |
| 1910-14 | 50 | 51 | 4 - 96 |
| 1915-19 | 142 | 28 | 1- 82 |
| 1920-24 | 104 | 21 | 1 - 69 |
| 1925-28 | 46 | 20 | 2 - 58 |
median. If the median study be taken to represent the period, the proportion of delinquents diagnosed feebleminded decreased from about 50 per cent in 1910-14 to 20 per cent in 1925-28.
In 1912 the Training School Bulletin contained a report of a questionnaire investigation of the extent of feeblemindedness among delinquents in reformatory institutions. The replies were based on the general observation of inmates by superintendents and other officials, not on mental tests. Replies were received from 34 institutions in which 13,188 delinquents were incarcerated; the median in this group of estimates was 14.5 per cent, and the range 0-41. If the trend in the psychometric studies of delinquents continues for another decade it will be not far from the guesses made a decade and a half ago on the basis of personal acquaintance.
If this entire group of delinquents be broken up into the various institutions in which they are incarcerated the same trend is evident in each type of institution. Table II is a classification of the psychometric studies of delinquents by types of in-
( 359) -stitutions. Longer time intervals are taken for some of these groups in order to secure a sufficiently large number of studies in each group to justify comparisons. This trend may be
illustrated in the Indiana Boys' School, in which routine tests have been reported over a series of years; from 1914 to 1917 the proportion of delinquents diagnosed feebleminded was 59 per cent; from 1922 to 1927 it was 10 per cent.
| Institution and Years |
Number of Studies |
Percentage Feebleminded in Median Study |
Percentages Feebleminded: Range |
| Juvenile institutions | |||
| 1910-14 | 24 | 45 | 4-96 |
| 1915-19 | 65 | 32 | 2-78 |
| 1920-28 | 56 | 17 | 2-65 |
| Reformatories | |||
| 1910-14 | 15 | 44 | 14 - 89 |
| 1915-19 | 28 | 25 | 6 - 66 |
| 1920-28 | 39 | 26 | 3 - 60 |
| State prisons | |||
| 1910-19 | 30 | 23 | 1 - 75 |
| 1920-28 | 34 | 19 | 1 - 60 |
| Jails, workhouses | |||
| 1910-19 | 12 | 31 | 3 - 84 |
| 1920-28 | 9 | 9 | 1 - 69 |
| Other institutions | |||
| 1910-19 | 26 | 39 | 8 - 83 |
| 1920-28 | 11 | 24 | 8 - 37 |
This downward trend in the proportions of delinquents reported feebleminded may be interpreted in two ways: one is that intelligent people are relatively more likely to commit crime now than they were a generation ago, the other is that the methods of measurement of intelligence have changed. The invalidity of the first interpretation cannot be demonstrated,
(360) but the second seems much more plausible in view of the wellknown changes in the methods of measurement of intelligence. About 1915 much criticism of mental testing methods developed. Many of the testers pointed out the lack of standardization in these tests. In 1917 Gilliland reported that if 100 delinquents in the Columbus workhouse were given mental tests the percentage found feebleminded might vary from i9 per cent to 50 per cent, depending upon which one of twelve standards was used. In the same year Miss Fernald reported that if 100 delinquent women, inmates of Bedford, were graded according to nine different methods then in general use the proportions diagnosed feebleminded would range between 34 per cent and 100 per cent, depending upon the particular standard selected. In those early days of mental testing the influence of Goddard was very great; he had asserted that the more expert the mental tester the larger the proportion of delinquents he would find to be feebleminded. Many of the testers attempted to demonstrate their superiority in that manner. During these years Healy, Bronner, Miner, Wallin and others were questioning the reliability of the high percentages of feebleminded among the delinquents; Pintner and Paterson and others had suggested methods of standardizing the tests. The criticisms, discussion, and comparison of results did result in lowering the standards of normal intelligence so that many who had previously been classified as feebleminded came to be included among the normal. These changes appear to be the principal factor in the decrease of the proportion of delinquents diagnosed feebleminded.
A second conclusion from this survey of psychometric studies of delinquents is that the proportions of delinquents feebleminded have varied widely even in the studies made since 1919. In this group of more recent studies at one extreme stand two studies in which only 1 per cent of the delinquents were found to be feebleminded, at the other extreme one study in which 69 per cent were found to be feebleminded. One fourth of these studies report less than 10 per cent of the delinquents feeble- minded, one fourth report from 10 to 21 per cent feebleminded, one fourth 22 to 36, and one fourth more than 36 per cent.This scattering may be interpreted as the result of difference either in the intelligence of the groups tested or in the methods
(361) of diagnosing intelligence. Both interpretations are evidently partially correct. Even if testing procedures were absolutely identical differences in the results would be expected because the groups tested vary widely in their composition. On the other hand, wide variations in results appear even when groups which are nearly identical in their composition are tested by different persons. In 1920 the inmates of the penal and reformatory institutions of Wisconsin were tested by representatives of the National Committee for Mental Hygiene, with the percentages of feebleminded ranging from 8 per cent in one institution to 16 per cent in the institution at the other extreme.[2] Since 1923 the state psychologist of Wisconsin has been reporting each year about 30 per cent of the inmates in those same institutions to be feebleminded. This increase of loo per cent or more in the proportions feebleminded is surely not due to a real increase in feeblemindedness but to the fact that a different person has been giving the tests. Again, routine examinations are given to the inmates of the state prisons of Minnesota and of Illinois; 40 per cent of the inmates of the state prison of Minnesota are found feebleminded, about 7 per cent of the inmates of the state prison of Illinois.
Methods of testing and methods of grading the tests vary widely. Some testers use an I.Q. Of 75 as the upper limit of feeblemindedness, some use 7o, and Dearborn and Burt maintain that the dividing line, so far as it can be located by tests, should be closer to 50 than to 70. Some take into account nothing except the tests, while others include a wide variety of other facts regarded as pertinent, especially race, nationality, formal education, economic status, and previous behavior. Some diagnose as feebleminded only those with low intelligence scores who are not found, in a complete physical and psychiatric examination, to be psychotic, epileptic, neurotic, psychopathic personalities, deteriorated by drugs or alcohol, or in other ways mentally or nervously abnormal, while others distribute the entire sample on a unilinear curve of intelligence. Some include only vocabulary ability in their tests, others include a wide variety of other abilities. Some administer tests when the delinquents first enter the institution, others after the delinquents have had considerable time in which to adjust to the institu-
(362) -tional life. Some have a much better technique than others of putting the delinquent at ease and inducing him to perform to the maximum of his efficiency.
Mental tests of delinquents are certainly not standardized to the extent of eliminating important personal variations. It is probable that if ten persons of national reputation could test an identical group of delinquents, being free to use their own methods and criteria, the results would be strikingly different. We might expect Dr. Adler to report that about 7 per cent of the group were feebleminded, Dr. Doll 1o per cent, Dr. Healy 13 per cent, Professor Terman 23 per cent, Dr. Kuhlman 35 per cent, Professor Root 40 per cent, and Dr. Hickson 70 per cent. To be sure, if they all used the same tests and followed prescribed directions no such variation would be found, but they do not use the same tests and criteria. Consequently a report regarding the proportion of a delinquent group feebleminded is of primary significance in locating the mental tester upon a scale of mental testing methods. In this sense the psychometric tests of delinquents throw more light upon the intelligence of the mental testers than upon the intelligence of the delinquents.
A third conclusion from this survey of psychometric tests of delinquents is that feeblemindedness has not been demonstrated to be a generally important cause of delinquency. The studies from which this conclusion is derived are of two types: first, comparison of the mental scores of delinquent and of the general population; second, comparison of behavior of the feebleminded and normal minded.The early mental testers were very certain that they had discovered the most important cause of crime. Goddard and others called the feebleminded "potential criminals" as though other people were not potential criminals. Moreover, Goddard stated that a diagnosis of feeblemindedness "fully explains any delinquency of which the child has been guilty." In accordance with that proposition many of the early testers, after finding 50 per cent of the delinquents tested feebleminded, concluded that feeblemindedness was the cause of 5o per cent of all delinquency and crime. The first difficulty about that conclusion was the lack of a satisfactory measure of the intelligence of the general population with which the delinquents were compared. At first it was believed that about one-half of 1 per
(363) cent of the total population was feebleminded, later that it was 1 per cent or even 2 per cent. Some later studies have raised the proportion still higher; 4.24 per cent of the school children tested in one California county were found feebleminded, 5.0 per cent in Minnesota,' and 8.6 per cent in Rhode Island,' while the Army Tests have frequently been interpreted as evidence that 24 per cent of the general population is feebleminded. The distribution of the general population in respect to intelligence has not been satisfactorily determined but the trend appears to be toward higher proportions feebleminded. Some studies have been made of school children and of delinquents in the same state by the same persons, but the relative intelligence is not consistent even in these. The state psychologist of Minnesota found 5 per cent of the public school children examined in the state feebleminded, and 30 per cent of the inmates of all penal reformatory institutions in the state feebleminded. On the other hand, the National Committee for Mental Hygiene found 8.6 per cent of the school children of Rhode Island and 8.4 per cent of the inmates of the penal and reformatory institutions of the state feebleminded.' Army Tests have been given to adult delinquents by Adler,' Lincoln,' Doll," Murchison,' Stone," Weber," and others; these tests have been given in the federal penitentiary at Leavenworth, state prisons of Kansas," Illinois, Indiana, New Jersey, and Maryland, state reformatories of Illinois, Indiana, Ohio, and Nebraska, and municipal prisons of St. Louis." These tests, given by differ-
( 364) -ent persons in different institutions, are strikingly consistent in the conclusion that adult delinquents score about the same as the draft army of the same race and nationality in the same state. Serious questions have been raised regarding the validity of these tests and the validity of using the draft army as a sample of the general population, but the consistency in results is a fact that cannot be overlooked.
A second reason for discounting the early conclusion regarding the importance of feeblemindedness as a cause of delinquency is that the delinquents who have been tested are always a selected portion of the entire delinquent population and are probably selected partly because of their feeblemindedness. Delinquents who are smarter than policemen are less likely to be caught than those who are less intelligent. Mental tests of police in Los Angeles, Palo Alto, Detroit, and Cleveland indicate that, if this proposition is correct, a very large portion of the general population could commit delinquencies and not be caught. Also, the feebleminded person is more likely to be convicted if arrested, more likely to be committed to an institution if convicted, and less likely to be paroled early if committed to an institution than is the non-feebleminded delinquent. The Voluntary Defenders Committee in New York City dealt with 574 defendants charged with felony, of whom 12 per cent were diagnosed mentally defective; of the 395 convicted 16 per cent were mentally defective; of those convicted 65 per cent were committed to institutions, but of the feebleminded who were convicted 82 per cent were committed to institutions; in general, 45 per cent of all persons charged with felonies were committed to institutions, while 75 per cent of feebleminded persons charged with felonies were committed to institutions after conviction.[14] In Joliet in 1921 the mentally defective constituted 28.6 per cent of the prison population, but only 15.6 per cent of those paroled. These comparisons indicate that since the tests are generally given to delinquents in institutions, they are given to a selected group of delinquents.A third objection to the conclusion regarding the importance of feeblemindedness as a cause of delinquency is the report by several investigators that the group with the highest rate of
(365) delinquency is not the definitely deficient group but the dull group. Wallin, especially, has stressed this. Some others have reported that the superior group ranks next to the feebleminded group, so that the delinquency rate drops from the deficient group to the normal and then rises from the normal to the very superior. The evidence on this point is not sufficient to justify detailed comparisons but at any rate raises a doubt regarding the validity of the conclusion that there is a straight-line relationship between intelligence and delinquency.
In view of the upward trend in the proportion of the general population reported feebleminded, the downward trend in the proportion of delinquents reported feebleminded, the striking similarity in some studies in the distribution of intelligence of the delinquents who are tested and the portion of the general population tested, especially by Army Tests, and the fact that delinquents who are tested are likely to include an undue proportion of feebleminded because of the greater probability of arrest, conviction, and commitment to institutions, and the smaller probability of early parole, the claim of the early testers regarding the general importance of feeblemindedness as a cause of crime does not seem to be substantiated.The second type of studies from which a conclusion regarding the importance of feeblemindedness as a cause of delinquency has been derived, consists of attempts to determine delinquency records of groups of feebleminded persons. Goddard in his Feeblemindedness makes a report regarding the delinquency records of 1,987 feebleminded persons and 4,111 normal members of the same families as follows: alcoholic 5.4 per cent of the feebleminded, o.6 per cent of the normal; sex offense 8.3 per cent of the feebleminded, 0.5 per cent of the normal; crime 1.2 per cent of the feebleminded, 0.0 per cent of the normal; three offenses 15.1 per cent of the feebleminded, 1.i per cent of the normal. This difference is very striking, but it needs to be discounted because of three things: first, the much better opportunity to study the life records and thus discover the delinquencies of the feebleminded, a large proportion of whom were in institutions; second, the fact that more than half of the members of these families were placed in a third class as "undetermined" with reference to intelligence, which might, if distributed, change significantly the relative delinquency rates
(366) of the other two classes; third, the fact that the diagnosis of the mentality of persons living two or three generations ago is highly unreliable.
Miss Hansford made a study in 1918 of mental defectives in a county in Indiana.[15] She found 527 feebleminded persons in this county or in state custodial institutions on commitment from this county; only 1.1 per cent of these feebleminded persons were in institutions for delinquents. This finding of a very small proportion of the feebleminded with this record of delinquency is not conclusive because of the lack of a comparable record for the general population, and especially because of the well-known fact that feebleminded persons, even if known to be delinquent, are likely to be committed to custodial institutions rather than to institutions for delinquents. The author, however, makes this general statement on the basis of her knowledge of the situation: "Those persons who are charged with the most serious crimes as well as those who are arrested most frequently on such charges as intoxication and disturbing the peace, are not the persons listed as feebleminded nor are they relatives of mental defectives.”
In 1922 a study was made of 117 consecutive cases known to the New York Charity Organization Society;[16] a diagnosis of mental defect had been made in all of these cases and all had been closed in 1916-17. At the end of five years 17 were dead and 28 in custodial institutions; 3 were in prison, 1;4 in the community with a record of promiscuous sex relations, 5 in the community with a record of vagabondage, a total of 18 per cent of the entire group with a record of delinquency; 1;7 per cent were in the community and classified as a fairly social type, 20 per cent in the community and classified as economically marginal but not delinquent so far as known. This study, also, is not conclusive, primarily because of the lack of a control group; if the careers of feebleminded clients of the Society could be compared with the careers of all clients of that Society the result would be more significant.
In 1919 Dr. W. E. Fernald made a study of the after-careers
(367) of feebleminded inmates discharged from Waverley between 1890 and 1914; of 1,537 discharged, 537 were discharged to other institutions, 175 to other states, and 279 could not be located; the number studied was, therefore, 647, of whom 16 per cent had a record of delinquency or of arrest on charge of delinquency.[17] The question that may properly be raised regarding this study is whether feebleminded persons committed to Waverley were not a selected group of feebleminded, namely, those who had shown unusual tendencies toward delinquency, and whether, therefore, the after-careers of this particular group are representative of the behavior of all feebleminded persons.
Several studies of the after-careers of children in ungraded classes in the public schools have been made. In 1915 Miss Farrell studied the records of 350 boys and girls who had been in ungraded classes in New York City during the preceding eight years, and found 1 per cent of them in penal institutions.[18] In 1916 the Superintendent of Schools of Detroit reported that of 100 children who had been in special classes in Detroit 27 per cent had come into contact with correctional agencies by the school.[19] Of 203 children in special classes in Cincinnati 22 per cent had come into contact with correctional agencies by the time they had reached a minimum age of eighteen.[20] On the other hand, Professor Wallin found for 2,774 consecutive children sent to the school clinic in St. Louis between 1914 and 1921, suspected of mental deficiency, that the average mental age of those who had a record of school delinquencies was higher than of those who had no such record; 24 per cent of the delinquents were diagnosed feebleminded, 35 per cent of the non-delinquents.[21] Also Witty and Nelson report that
(368) though the distribution of intelligence is about the same for the two sexes, the number of boys in ungraded classes is about twice as large as the number of girls. The conclusion is that children sent to ungraded rooms are selected not only on the basis of deficiency but also of behavior, and the after-career of children from these rooms is, therefore, not a suitable measure of the relation between delinquency and feeblemindedness.
In general, though these studies of the behavior records of feebleminded persons show some very large percentages of delinquency among the feebleminded, they all need to be discounted because of the selected nature of the group studied and because of the lack of a suitable group for purposes of comparison. These studies, therefore, tend to substantiate the conclusion that the great importance of feeblemindedness as a cause of delinquency has not been demonstrated.
No one doubts that wide differences between delinquents exist in respect to intelligence, and that these differences are indicated to some extent by mental tests. Even though the tests are not standardized, it may be possible by using the same testing procedure within an institution to find differences between classes of delinquents, such as first offenders and recidivists, burglars, and sex offenders. The next five conclusions are concerned with such relative differences between delinquents tested by the same procedures; they are concerned with a comparison of one group of delinquents and other groups of delinquents rather than a comparison of delinquents and non-delinquents.
A fourth conclusion is that within the institution the feebleminded delinquents behave about as well as the non-feebleminded. The evidence on this point is not extensive but is consistent. In Whittier State School, California, a small negative correlation was found between the intelligence and the goodness of behavior (r=-.093 +.057) and this was consistent for each of three groups, white, Negro, and Mexican-Indian.[22] Clark reports on the basis of this and other evidence: "Boys of superior intelligence were proportionately more frequent and chronic offenders [within the institution] than boys of any other intelligence group." The reports of Auburn State Prison,
(369) New York, since 1921: have included statistics of the mental classification of all entrants to the institution and of all prisoners punished for infraction of prison rules. These statistics show that 52 per cent of those who entered the institution from 1921 to 1926 were diagnosed feebleminded, and that 51 per cent of those who were punished for institutional delinquencies were feebleminded, that is, that the feebleminded were punished with almost exactly the same frequency as the total population.[23] The total number of persons punished during this period was 23 per cent of the total number entering the prison, and the number of feebleminded persons punished was 22 per cent of the number of feebleminded persons entering the prison. This indicates that feebleminded prisoners behave within the institution about as well as the average of the population. The borderline group had the highest ratio of punishments, the normal the lowest.
A fifth conclusion is that these studies do not in general indicate that the amount of intelligence exercises a decisive influence in determining goodness of behavior on parole. A diagnosis of inferior intelligence was made in 1921 for 15.6 per cent of the prisoners paroled from Joliet; 1:5.5 per cent of those who violated their parole were of inferior intelligence. Thus the prisoners of inferior intelligence behave on parole about the same as the average of all prisoners. Burgess found in Illinois no consistent relation between amount of intelligence and success on parole; of those paroled from Pontiac the groups which had rates of violation of parole above the average of the entire group were, in order, those of "superior intelligence," "very inferior intelligence," "low average intelligence"; in Menard the groups above the average in violation of parole were, in order, "very superior," "high average," and "superior"; in
(370) Joliet "average" and "low average." In each case the difference was small. Burgess concluded that "parole violation is no more frequent-if as frequent-among those of inferior than among those of higher intelligence." [24] Borden reports that among delinquent boys in New Jersey "the lower the intelligence the more likely a boy is to succeed" on parole (r = -.164).[25] On the other hand Warner found that 25 per cent of those who violated parole from the Massachusetts Reformatory were feebleminded, as compared with 19 per cent of those who succeeded on parole.[26] Elmira Reformatory reported in 1919 that 35 per cent of those who were returned for violating parole were feebleminded, as compared with 17 per cent of the entire population of the institution in that year. Clark found a correlation of +.19 between intelligence and success on parole from Whittier State School, but this correlation varied from -.51 for those who worked in transportation while on parole to +.74 for those who worked in agriculture.[27] Healy found in his study of outcomes of cases dealt with in Chicago that 10 per cent of those who succeeded were feebleminded, 16 per cent of those who failed, and that feebleminded girls had a larger proportion of successes than feebleminded boys.[28] Preston School of Industry in California reported in 1918 that 37 per cent of all delinquents failed on parole, but 44 per cent of the defective delinquents failed on parole. Taking this conflicting evidence altogether it appears to justify the conclusion that feebleminded delinquents on parole in general behave about the same as all delinquents on parole.
A sixth conclusion is that no generally significant relationship between intelligence and recidivism has been demonstrated. A study of 1,288 inmates of 34 county jails and penitentiaries in New York in 1925 showed that 5.8 per cent of those arrested
(371) once were mentally deficient, 9.9 per cent of those arrested twice, 9.6 per cent of those arrested three times, and 6.7 per cent of those arrested four or more times.[29] In Rhode Island a higher proportion of defectives was found among recidivists than among first offenders in the state prison, the industrial school, and the county jail, but a lower proportion in the house of correction.[30] Miss Fernald found no significant association between intelligence and recidivism among Bedford women; the mental age of first offenders was 75.64, of recidivists 75.68; the coefficient of correlation between mental age and number of previous convictions was -.054 +.052; the mental age of those with from 9 to 20 previous convictions was ß.5.I, of those with no previous convictions 75.6.[31] Slawson found in his study of juvenile delinquents in New York that the association between intelligence and recidivism was very small and the probable error very large.[32] Slawson and Fernald made corrections for differences in age in their studies, without producing significant associations thereby. Murchison found recidivists more intelligent than first offenders, but on analyzing this concluded that of those guilty of fraud and sex offenses first offenders are much more intelligent than recidivists, of those convicted of statutory offenses (illegal sale of drugs or liquor, possession of burglar's tools, etc.) less intelligent, and of those convicted of other crimes practically identical.[33] In general this evidence is conflicting but tends to justify the generalization that the relationship between intelligence and recidivism is very slight.
A seventh conclusion is that these studies show no consistent relation between age of delinquents and intelligence. Miss Lowe in a study of unmarried mothers in Minnesota found an average I.Q. of 63.6 for those over 35 years of age, an average of 92 for those between the ages of 15 and 19, and the other age groups were distributed in regular order between these extremes so that the lower the age of the delinquents the
(372) greater her intelligence.[34] Erickson reports a similar distribution in the penal and reformatory institutions of Wisconsin.[35] On the other hand, Miss Fernald found a steadily increasing median mental capacity as chronological age at first sex offense increased, with one slight exception; she found the mean mental capacity of those convicted first before 21 to be 75.70, of those convicted first after 21 to be 75.65, and a general coefficient of correlation between mental capacity and age at first conviction of -.07 +.052. She concludes, "It is impossible even to state the direction of the association, i.e., whether the more intelligent women tend to be convicted at an earlier or at a later age than do the less intelligent." [36]
A final conclusion is that the type of crime is affected somewhat by intelligence. Those convicted of fraud are generally a more intelligent group, those convicted of sex offenses generally a less intelligent group. The evidence is not conclusive, however, regarding many other particular offenses. Those convicted of homicide have the highest intelligence rating of any group in Illinois, the lowest in the Western Penitentiary of Pennsylvania, and near the middle in the institutions studied by Murchison. Glueck, Murchison, and others have attempted to combine these specific charges into more general categories and reach the conclusion that those who commit acquisitive or predatory crimes are the highest in intelligence, those who commit sex crimes the lowest, and those who commit crimes of personal violence (other than sex crimes) are intermediate but not far above those who commit sex crimes. The statistics of Connecticut State Prison for 1925 to 1928 agree with this conclusion so far as the superiority of those convicted of acquisitive crimes is concerned but reverse the rank of the other two groups; 28.7 per cent of those convicted of acquisitive crimes, 44.4 per cent of those convicted of sex crimes, and 49. 1 per cent of those convicted of crimes of personal violence were reported to have a mental age of eleven or less. The evidence is fairly conclusive, however, regarding the general relative su-
(373) -periority of those convicted of crimes of acquisition, and the relative inferiority of those convicted of sex crimes.
This survey of studies of delinquents by psychometric methods results in two general conclusions: First, the mental testing procedures as actually used fall far short of providing objective, standardized knowledge. Results are decidedly influenced by personal elements in the mental testers. The proportions of delinquents reported feebleminded have been decreasing but are tending to reach a greater degree of uniformity, but even in recent years much scattering in results appears. The tests, thus far, do not furnish sufficient evidence to justify a conclusion that the studies which report less than 10 per cent of the delinquents feebleminded are more reliable than those which report more than 36 per cent of the delinquents feebleminded.Second, the early belief that feeblemindedness was in general and by itself a very important cause of crime is not definitely substantiated. Rather, the conclusion is that it seems to be a relatively unimportant factor, in view of the decreasing difference between the reports of the distribution of the intelligence of the delinquents and of the general population, and in view of the relative similarity of behavior of delinquents within institutions and on parole, and in view of the relative similarity of first offenders and of recidivists, and of the age at which delinquents commit their offenses. It appears to be of some significance in determining the general type of crimes and of a few specific types of crimes.
The most significant conclusion from this evidence is that the relation of feeblemindedness to delinquency cannot be determined by dealing with it in isolation from other factors. We find feebleminded persons well behaved in some situations, delinquent in others; even in one objective situation some feebleminded persons become delinquent, other feebleminded persons do not become delinquent. The significance of feeblemindedness apparently can be determined only when studied in relation to a great many other personal and situational factors. For instance, the evidence in regard to the relation between intelligence and success on parole is conflicting. In one study it is suggested that the reason for the larger proportion of successes among feebleminded may be because the feebleminded are more frequently sent to rural districts and employed in agri-
(374) -culture. In another study it is suggested that the reason for the greater frequency of success of feebleminded on parole may be that the parole officers devote more attention to the feebleminded than to the normal. Apparently the relation between intelligence and success on parole can be determined only by studying these two factors in relation to the whole personal and situational complex. Similarly in regard to each other aspect of the relation between intelligence and delinquency it appears to be methodologically desirable to study these factors in relation to other factors rather than in isolation.
SELECTED BIBLIOGRAPHY
Adler, Herman M., and Worthington, Myrtle R., "The Scope of the Problem of Delinquency and Crime as Related to Mental Deficiency," Journal of Psycho-asthenics, 1925, Vol. XXX, pp. 47-57.
Anderson, V. V., "A Study of the Careers of 321 Feebleminded Persons Who Have Been in Special Classes and Are Now Out in the Community," Journal of Psycho-asthenics, Vol. XXVII, pp. 138-49.
Anderson, V. V. (Director), Report of Mental Hygiene Survey of Delinquency and Dependency Problems in St. Louis, St. Louis Department of Public Welfare, 1922.
Borden, H. G., "Factors for Predicting Parole Success," Journal of Criminal Law and Criminology, 1928, Vol. XIX, pp. 328-38.
Bruce, A. A., Burgess, E. W., and Harno, A. M., The Indeterminate Sentence Law and Parole System in Illinois, Illinois Department of Public Welfare, 1928.
Clark, W. W., "Success Records of Delinquent Boys in Relation to Intelligence," Journal of Delinquency, 1920, Vol- V, pp. 174-82.
Doll, E. A., "The Comparative Intelligence of Prisoners," Journal of Criminal Law and Criminology, 1920, Vol. XI, pp. 191-97.
Erickson, M. H., "A Study of the Relationship Between Intelligence and Crime," Journal of Criminal Law and Criminology, 1929, Vol. XIX, pp. 592-635.
Farrell, Elizabeth E., "A Preliminary Report on the Careers of 350 Children Who Have Left Ungraded Classes," Journal of Psychoasthenics, 1915, Vol. XX, pp. 20-26.
Fernald, Mabel R., et al., A Study of Women Delinquents in New York State, New York, 1920.
Fernald, W. E., "After-care of the Patients Discharged from Waverley for a Period of Twenty-five Years," Ungraded, 1919, Vol. V, pp. 25-31.Glueck, Sheldon, Five Hundred Criminal Careers, New York, 1930.
Goddard, H. H., Feeblemindedness, Its Causes and Consequences, New York, 1914.
Hamilton, G., "Community Aspects of Feeblemindedness," Mental Hygiene, 1923, Vol. VII, pp. 312-18.
( 375)
Hansford, Hazel I., "A Social Study of the Mental Defectives in County H, Indiana, in 1918," Indiana. University Studies, 1923, Vol. X, No. 59.
Healy, W., and Bronner, Augusta F., Delinquents and Criminals, Their Making and Unmaking, New York, 1926.
Kelley, T. L., "Mental Aspects of Delinquency," University of Texas Bulletin, No. 1713, March, 1917.
Kuhlman, F., "A State Census of Mental Defectives," Minnesota State Board of Control, Quarterly, November, 1926, Vol. XXIII, pp. 23-35
Lincoln, E. A., "The Intelligence of Military Offenders," Journal of Delinquency, 1920, Vol. V, pp. 31-40.
Malzberg, B., "Relation of Mental Defect to Delinquency," Journal of Criminal Law and Criminology, 1919, Vol. X, pp. 218-22.
Miner, J. B., Deficiency and Delinquency, Baltimore, 1918.
Murchison, Carl, Criminal Intelligence, Worcester, 1926.
New York, Report of Committee on Psychopathic Delinquents in New York, Report of Commissioner of Prisons, 1925.
Raphael, T., and Others, "Socio-psychiatric Delinquency Studies from the Psychopathic Clinic of the Recorder's Court of Detroit," Mental Hygiene, 1924, Vol. VIII, pp. 433-65
Root, W. T., Psychological and Educational Survey of 1,916 Prisoners in the Western Penitentiary of Pennsylvania, Western Penitentiary, 1927.
Slawson, J., The Delinquent Boy, Boston, 1926.
Spaulding, E. R., An Experimental Study of Psychopathic Delinquent Women, New York, 1923.
Stone, C. P., "A Comparative Study of the Intelligence of 399 Inmates of the Indiana Reformatory," Journal of Criminal Law and Criminology, 1921, Vol. XII, pp. 238-56.
Wallin, J. E. W., "The Diagnostic Findings from Seven Years of Examining in the Same School Clinic," Journal o f Delinquency, 1923, Vol. VIII, pp. 169-95.
Warner, S. B., "Factors Determining Parole from the Massachusetts Reformatory," Journal of Criminal Law and Criminology, 1924, Vol, XIV, pp. 172-207.
Weber, C. O., and Guilford, J. P., "Character Trends versus Mental Deficiency," Journal of Criminal Law and Criminology, 1926, Vol. XVI, pp. 610-12.
Weidensall, C. Jean, The Mentality of the Criminal Woman, Baltimore, 1916.
Williams, J. H., The Intelligence of the Delinquent Boy, Whittier State School, Whittier, California, 1916.
