Mental Retardation

The problem of definition

The nature of intelligence

The nature–nurture issue

The defect and difference orientation

Motivational and emotional factors

BIBLIOGRAPHY

Mental retardation is a problem of serious social concern. In view of the large number of persons considered to be mentally retarded, such concern is certainly justified. Using the conventional criterion of 3 per cent of the population, the U.S. President’s Panel on Mental Retardation (1963) estimated that almost 5.5 million children and adults in the United States are mentally retarded. The criterion for mental retardation established in the “Manual on Terminology and Classification in Mental Retardation” (Heber 1959) and adopted by the American Association on Mental Deficiency as well as the Biometrics Branch, National Institute of Mental Health, is that all those at least one standard deviation below the population mean intelligence quotient (IQ) are considered retarded. If one accepts this criterion, and many do not, there are almost 30 million mental retardates in the United States. If the more conservative estimmate is employed, mental retardation is twice as prevalent as blindness, polio, cerebral palsy, and rheumatic heart conditions combined (Doll 1962).

The typical textbook pictures the distribution of intelligence as normal or Gaussian in nature, with approximately the lower 3 per cent of the distribution encompassing the mentally retarded. A common class of persons is thus constructed, a class defined by intelligence-test scores below 70. This schema has misled many laymen and students and has subtly influenced the approach of experienced workers in the area. For if one fails to appreciate the arbitrary nature of the cutoff point of 70, it is but a short step to the formulation that all those falling below this point compose a homogeneous class of “subnormals.” Since the conceptual distance between “subnormal” and “abnormal,” with its age-old connotation of disease and defect, is minimal, the final step is to regard retardates as a homogeneous group of defective organisms, immutably different from those persons possessing a higher IQ.

The view that mental retardates represent a homogeneous group is seen in numerous research studies where comparisons between retardates and normals are made on the basis of IQ classification alone. The view that mental retardates, as a group, are “different” is most vividly encountered in comparative studies where mental retardates are conceptualized as occupying a position on the phylogenetic scale somewhere between monkeys and children of average intellect. It is of some interest to note that people deficient in respect to intelligence-test performance are usually not called “mental deficients” but rather are commonly referred to as “mental defectives.”

The defect orientation to mental retardation originally emphasized the notion of moral defect and stemmed anywhere from the belief that retardates were possessed by a variety of devils to the empirical evidence of their exhibiting an inordinately high incidence of socially unacceptable behaviors, such as crime and illegitimacy. More recently, the notion of defect has referred to defects in either physical or cognitive structures. This defect approach has a certain unquestionably valid component. There is a sizable group of retardates who suffer from a variety of known physical defects. Mental retardation may be due to such factors as a dominant gene (as in epiloia), a single recessive gene (as in gargoylism, phenylketonuria, amaurotic idiocy), infections (such as congenital syphilis, encephalitis, rubella in the mother), chromosomal defects (as in mongolism), toxic conditions (such as radiation in utero, lead poisoning, and Rh incompatibility), and cerebral trauma. For a complete listing of the many types of mental retardation the reader is referred to the “Manual on Terminology and Classification in Mental Retardation” (Heber 1959).

The diverse etiologies noted above have one factor in common: in every instance examination reveals an abnormal physiological process, that is, there are specific or related defects in physiological functioning. Such persons are abnormal in the orthodox sense, since they suffer from a known disease defect. However, in addition to this group, which forms a minority of all retardates, there is the group labeled “familial,” or more recently “undifferentiated,” which comprises approximately 75 per cent of all retardates. This group presents the greatest mystery and has been the object of the most heated disputes in the area of mental retardation. The diagnosis of familial retardation is made when an examination does not reveal the physiological manifestations noted above and when retardation exists among parents, siblings, or other relatives. As will be seen in a later section, several theoreticians have extended the defect notion to this type of retardate. On the basis of differences in performance between retardates and normals on some experimental task rather than on physiological evidence, they have advanced the view that all retardates suffer from some specifiable defect over and above their general intellectual retardation. However, these theoreticians differ as to the specific nature of the defect. The experimental paradigm employed to demonstrate such defects involves equating groups of normals and retardates on Mental Age (MA), thus roughly controlling for general intellectual level, and demonstrating differences in performance between the two groups on some experimental task.

This more general defect approach thus lends support to the conceptualization of the mentally retarded as a homogeneous group of physiologically defective persons. Some order can be brought to the area of mental retardation if a distinction is maintained between physiologically defective retardates, with known etiologies, and familial retardates, whose etiology is unknown.

For the most part, work with physically defective retardates involves investigation into the exact nature of the underlying physiological processes, with prevention or amelioration of the physical and intellectual symptoms as the goals. Jervis (1959) has suggested that such “pathological” mental deficiency is primarily in the domain of medical sciences, whereas familial retardation represents a problem to be solved by behavioral scientists, including educators and behavioral geneticists. Diagnostic and incidence studies of these two types of retardates have disclosed two striking differences. The retardate having an extremely low IQ (below 40) is almost invariably of the defective type. (This does not mean that one cannot find defective retardates at every level of retardation. In fact, brain-damaged individuals may be found at every point along the IQ continuum.) Familial retardates, on the other hand, are almost invariably mildly retarded, usually with IQs above 50. The defect position emphasizes the innate, if not immutable, difference between retardates and normals.

The problem of definition

The decision of whether a person is considered retarded is often based not upon his intellectual characteristics but upon legal and occupational factors as well as his general level of social adjustment. The matter has been put most succinctly by Maher who stated:

What constitutes mentally retarded behavior depends to a large extent upon the society which happens to be making the judgment. An individual who does not create a problem for others in his social environment and who manages to become self-supporting is usually not defined as mentally retarded no matter what his test IQ may be. Mental retardation is primarily a socially defined phenomenon, and it is in large part meaningless to speak of mental retardation without this criterion in mind. (1963, p. 238)

This emphasis on social factors in defining mental retardation may lead to more confusion than clarity as indicated by the discrepancies found among various incidence and survey studies. The data of Table 1 would indicate that the incidence fluctuates not only across age categories but also according to the locality. If mental retardation is defined strictly in terms of IQ, and assuming a certain constancy of IQ score, we would expect no difference in the incidence of mental retardation at different ages. The standardization data of the Wechsler-Bellevue Scale confirm this expectation.

The incidence figures reported in Table 1 are understandable if one realizes that they reflect diagnoses based on some combination of IQ and the success of the individual in meeting social demands. For example, the extremely low incidence under 5 years of age may reflect the minimal social demands made on young children. The highest incidence obtained at the 10-14 age level occurs when the child is faced with school and more demanding intellectual tasks. It is probably in this age range that the relationship between IQ scores and meeting societal expectancies (i.e., successful school performance) is greatest. Stated somewhat differently, it is probably in this age range where the use of either the IQ or the child’s success in meeting social demands would result in his being classified as mentally retarded. A test-score orientation to mental retardation results in the view that approximately 3 per cent of the population is mentally retarded. The social competence viewpoint, however, results in a much smaller incidence. Data obtained through surveying representative samples of large populations or the entire population of certain limited regions in England and Scandinavia indicate that about 1 per cent or less of adults are classified as mentally retarded (e.g., Fremming 1947).

Armed with the information that a person’s social adequacy has much to do with whether or not he is considered retarded, we begin to get some inkling of the arbitrariness involved in such a classification.

The nature of intelligence

Whether mental retardation is defined by an intelligence-test score or by the person’s social competence, which many claim reflects his intelligence, the essential aspect of mental retardation is lower intelligence than that displayed by the modal member of an appropriate reference group. There is little agreement, however, when the question precipitated by this statement is raised, namely, “What is the nature of intelligence?” We cannot avoid this question by invoking the very unsatisfying cliche that “intelligence is what an intelligence test measures,” since it is perfectly apparent that the test constructor must have some definition of intelligence in mind, either explicitly or implicitly, before he can select test items. However, some consensus can probably be found for the view that intelligence is a hypothetical construct which has as its ultimate referent the cognitive processes of the individual. Given this, we are still faced with the unresolved issue of whether intelligence represents some single cognitive process which permeates every intelligence test or nontest behavior or whether it represents a great variety of relatively discrete cognitive processes which can be sampled and then summated to yield some indication of the amount of intelligence a person possesses [seeIntelligence And Intelligence Testing].

In either case, the more important questions involve an understanding of exactly how such cognitive processes develop over the life span and exactly how innate and environmental factors interact to influence such development. Approached in this way, the problem of defining intelligence becomes one with the problem of the nature of cognition and its development.

Cognitive versus psychometric approach

It follows that if we are to understand the nature of intelligence, we must consult those workers intent on investigating the nature and development of cognitive processes (e.g., thought, memory, concept formation, and reasoning) rather than focus on the work of test constructors and psychometricians. There has been little cross-fertilization between these two groups, which have approached the investigation of intellective functioning quite differently.

The former group of investigators utilizes a variety of techniques and, through extremely detailed analyses, attempts to tease out the intricacies of man’s cognitive functioning. These theorists have tried to evolve a theory of human cognition and its development. If intelligence tests had been developed by this group, psychology might have avoided the perplexing state of affairs encountered in trying to define intelligence. Tests devised by such a group would, by necessity, be indicators of the formal features of the cognitive structure at various times in the life cycle. Recently, workers within this framework (e.g., Laurendeau & Pinard 1963) have taken an interest in the problem of assessment. Although the task of providing an acceptable theory of the development of cognition is far from finished, Laurendeau and Pinard have been able to take the first step toward the construction of an intelligence test based on the formal features of cognition that were isolated by Piaget [seeDevelopmental Psychology, article on A Theory Of Development].

From a historical point of view, the practical demands of society for a test which would measure intellectual functioning meant that intelligence became the province of the second group, namely, the testers and psychometricians. Furthermore, for a variety of reasons, American thinking was not receptive to the approach taken by the cognitive theorists. The practical and empirical nature of the work of the testers can be seen in the efforts of Alfred Binet, whose intent was not to investigate the nature of intelligence but rather to discover those test items which would discriminate between successful and unsuccessful school performance. As has been pointed out, Binet viewed his empirically selected tests as a social screening device rather than as a theoretical interpretation of the nature of intelligence. The success of Binet’s tests in predicting school achievement led workers to the belief that such tests were inextricably bound to man’s intelligence or cognitive functioning.

For the psychometricians, it then became clear that the nature of intelligence could be understood by examining the nature of the tests that were employed to measure it. By discovering the correlations obtained between subtests within a given battery or across different tests, it was felt that the structure of intellect would be revealed. Despite the statistical rigor involved, no very satisfactory theory of intelligence has come out of the correlational or factor-analytic methods. There is, in fact, little agreement among workers even with regard to the one constant theoretical issue throughout this body of work, namely, the global versus the specific nature of intelligent behavior.

It should be emphasized that the weakness of current theories of intelligence has led to a conceptual impasse in the area of mental retardation. If there is no satisfactory theory of intelligence, then the essential aspect of mental retardation must escape us and we must be content with superficial statistical and social approaches to this complex problem. We do not necessarily have to await a completed theory of intelligence, however, to cut through much of the complexity, disputation, and confusion encountered in the area of mental retardation. Some clarification appears possible through the simple process of reorienting or restructuring our approach to intellectual retardation. A rather sizable step forward is taken if our commitment to a simple test approach is abandoned in favor of a concern with cognitive processes.

The process-content distinction

The plea is not that we abandon tests, for every cognitive theorist must eventually employ tests, as defined in the broadest sense. The plea is that workers in the field turn their attention from the superficial content of tests (i.e., the right or wrong answer) and come to grips with the problem of the cognitive structures and processes that give rise to content. It is this distinction between structure and content that has for too long escaped most workers in the area of mental retardation. Conventional tests are viewed by process-oriented cognitive theorists as too analytic and artificial in character and as measuring an end product and not a process (Laurendeau & Pinard 1963, p. 481).

In general, however, Piaget’s approach, with its developmental and normative emphasis, has had very little appeal to workers in the area of mental retardation, since such workers are committed to the study of individual differences. In this context, the test-constructing efforts of Laurendeau and Pinard appear very promising, since these followers of Piaget have formulated the states of cognitive development in terms of the nature of the cognitive operations achieved, thus emphasizing the nature of the cognitive structure and its accompanying processes. In their work we thus see a bridge beween a truly cognitive approach to intelligence and the need in the area of mental retardation for an instrument with which to make individual comparisons.

The focus on the content of test behaviors has been carried over to many nontest behaviors, and often an insufficient distinction is made between intelligence and intelligent behavior. It is the author’s view that intelligence must refer to the formal characteristics of the cognitive structure and the processes that accompany it, whereas intelligent behavior should refer to the content of behavior in respect to the appropriateness (often defined in a relatively arbitrary manner) with which an organism carries out an act. (See Maher 1963, for another discussion of this distinction between intelligence and intelligent behavior.) For example, in a sheltered workshop the author recently encountered a retardate working with a surprisingly complex piece of machinery. His ability to use this machinery defied current knowledge concerning the capabilities of the retarded. The director of the workshop explained that the retardate had been taught to operate the machine through a shaping process not unlike that employed by B. F. Skinner in training pigeons to play Ping-pong. It was also learned that the retardate could handle the machine quite adequately provided its position was not changed. To emphasize this point, the machine was rotated on its axis approximately 90° the retardate then became somewhat agitated and was no longer able to operate the equipment. (Piaget has also commented that remarkable intellectual feats performed by children on some task or other cannot be repeated following relatively minor alterations in the task stimuli.) In terms of the finished product, the retardate was behaving just as intelli-gently as an operator with a normal IQ. However, this accomplishment does not indicate that the retardate has become normal in intelligence. It is obvious that he is using a much more primitive cognitive process to achieve his intelligent behavior than is the normal person. This example should also make it clear that process analyses demand that the investigator make a more careful analysis of the content than that provided by a superficial “product” or criterion of correctness.

Social competence and mental retardation

The content approach is expressed also in the social competence definition of mental retardation. What are the intellectual demands of social competence? We do not know, and very little effort is made to discover what they might be. In the area of mental retardation, social competence usually means the ability to maintain oneself without too frequent contact with state schools, state hospitals, welfare agencies, and police officers. Though social competence defined in this way reflects certain cognitive abilities, it may also reflect a variety of factors reminiscent of nonintellectual aspects of intelligence-test performance. We refer here to factors such as luck, social values, attitudes toward other people, and emotional needs that are relatively in-dependent of intellectual level. Thus, present intelligence tests may predict social competence better than an ideal intelligence test because of the overlap of nonintellectual variables which influence both intelligence-test scores and social competence.

Social competence does not inevitably reflect normal intellectual functioning any more than its absence in the emotionally unstable, the criminal, or the social misfit reflects intellectual subnormality. Social competence is much too heterogeneous a phenomenon and reflects too many nonintellectual factors to be of great value in understanding mental retardation. The basic problem is that the concept of social competence is so value laden, and its definition so vague, that it has little heuristic utility. Windle (1962) has pointed out that the social competence definition of mental retardation is applicable only to institutionalized populations, whereas quite different definitional criteria must be employed with noninstitutionalized retardates. The only clear and acceptable operational definition of social competence would appear to be related to whether the individual has managed to function outside an institutional setting. Even Heber (1962), who has made the strongest case for employing social competence, has admitted that objective measures of adaptive behavior are presently unavailable. He has also stated that the present ambiguity of the social competence construct is such that in practice intelligence-test performance must remain “the most important and heavily weighted of the criteria used.”

There is a further problem with the social competence construct related to a fallacy which has permeated much of our thinking concerning the retarded. We have somehow come to believe that it is impossible for anyone who is “truly” retarded to meet the complex demands of our society. The bulk of retardates who have MAs in the 9-12 range (remembering that an MA of 16 is the upper limit for an individual of average IQ), have the intellectual wherewithal to meet the minimal demands of our society. This becomes immediately apparent if one raises the question of how much intellectual ability is required to arise in the morning, dress oneself, catch a bus or walk to a single location, perform some undemanding sort of labor, and re-turn home. Indeed in the 1920s and 1930s it was discovered that there were no less than 118 occupations in our society suitable for individuals having MAs from 5 to 12. As late as 1956 it was noted that 54 per cent of jobs require no schooling beyond the elementary level (Whitney 1956).

Another major aspect of social competence is the ability of the individual to abide by the values of the society, that is, obey laws, and so on. While the incidence of crime among the retarded is higher than among the nonretarded, this increase of incidence is not very great, especially if one controls for social class. Here again it is an error to view obedience to the law as somehow beyond the ability of the retarded. One simply has to apply the concept of the stages of moral development as investigated by Jean Piaget (1932): fairly young children are capable of a morality based on absolutism, that is, the rules inhere in the very fabric of existence and are not to be broken under any circumstances. Individuals who never achieve a higher stage of moral development are certainly not developmentally adequate, but neither are they likely to break many laws.

In order to make social competence a useful indicator of cognitive functioning, we must thus abandon some simplistic notion of social competence in favor of a variety of continua theoretically based upon the cognitive demands of the social requirements involved. Such indexes could then be considered independent indicators of intellectual functioning. Empirical efforts of this sort may be seen in the Vineland Social Maturity Scale and the Worcester Scale of Social Attainment. A more theoretical effort may be found in the work of Phillips and Zigler (1964) where both intelligence test scores and conventional social competence indexes are combined into an index of developmental or maturational level.

The nature–nurture issue

Attention is now turned to the role of cognitive capacity in mental retardation. Maher (1963) believes that the concept of capacity has considerable heuristic value for workers in the area of mental retardation. Intellectual capacity means something akin to Hebb’s (1949) intelligence A, that is, an innate potential for the development of intellectual functions. Those who have argued that the intellectual capacity notion is a relatively useless one (e.g., Ferguson 1956) appear to be invariably committed to an environmentalistic or learning orientation.

Maher’s position is that the capacity concept has value as related to “the differences between individuals in rate of acquisition of responses under similar learning conditions. Such a concept necessarily implies the existence of structural differences between individuals and is incompatible with a psychology of the empty organism ...” (1963, p. 250). It is in this last sentence that we see the theoretical value of the capacity concept, since it forces us to conceptualize individuals as biological organisms innately differing in respect to the potential manifestation of a multitude of traits. Thus the concept of capacity is intimately related to the biological concept of the genotype.

There has been an interesting effort to make intelligence, and thus mental retardation, a matter of acquired skills and transfer phenomena in the classical learning theory sense (see Ferguson 1956). Although Ferguson appears to abhor a biological concept of intelligence, he nevertheless falls back upon it in dealing with those aspects of early learning which do not reflect transfer effects. In addition, his treatment of transfer as a uniformly manifested phenomenon overlooks differences in ability to transfer from one task to another which may very well be a reflection of biological capacity.

Given such an orientation, we can derive the optimistic view that complete control over learning experiences would do away with individual differences and, thus, mental retardation, at least of the nondefective variety. Such a view, though appealing, flies in the face of what has been observed. Herculean efforts of teaching and training have not resulted in marked change in the intellectual level of most retardates.

The environmentalist, while acknowledging the importance of biological capacity, treats human behavior as the outgrowth of an infinite number of experiences. It is of interest to note that one environmentally oriented theorist (McCandless 1964) has argued that although heredity and environment interact in the production of intelligent behavior, we need only concern ourselves with environment, since “we can do something about environment.” This approach implies that the manipulations of environments are expected to have constant results and, furthermore, ignores the obvious possibility that children with particular capacities will need specific environmental events in order to maximize their cognitive development. The one group that has seriously considered the nature of the interaction between genotype and experiences in producing certain behaviors (phenotypes) has been the behavior geneticists. Employing infrahuman subjects, these investigators have presented evidence that the effects of particular experiences and the behaviors to which they give rise depend upon the biological nature of the organism (e.g., Hirsch 1963).

The attempt to determine the proportion of variance attributable to heredity or to environment is full of difficulties (H. Jones 1946). Despite the shortcomings of the nature-nurture work on intelligence, it is still possible to derive certain conclusions. (For more complete reviews of this work the reader is referred to H. Jones 1946; McCandless 1964.) Studies of parent-child and of sibling resemblances in intelligence, a variety of twin studies, and studies on children in foster homes have made it clear that inherited intellectual endowment is a much more important factor in intelligence than those who are environmentally oriented would have us believe.

At the same time one must not forget the importance of environmental factors to manifest intelligence. The role of environment is evident even in extreme cases where a known gene defect is the cause of mental retardation. In the case of genetically determined phenylketonuria, subnormal intelligence occurs only in an environment which provides phenylalanine in the diet of the affected individual. A specific change in the environment (i.e., withholding phenylalanine from the diet) will prevent the occurrence of subnormal intelligence.

An issue in the nature-nurture controversy of special pertinence to mental retardation concerns the degree to which the environment may produce individual differences in intelligence in contrast to affecting the absolute achievement level of man. It is one thing to assert that the environment may play a role in determining the range of individual differences found among men. It is another thing to assert that environmental events can cause the individual with a normal intellectual endowment to become retarded or, for that matter, shift the entire range of intelligence in such a way that no individual would display that degree of intellectual impairment that we now label retarded.

Different environments (e.g., rural versus urban, racial-cultural, social class) are associated with differences in intelligence. To what extent such differences reflect environmental as opposed to inherited factors remains an open issue. The position of the majority seems to be nurture oriented, and the argument advanced is that it is the social class or cultural environment which produces retardation. To state the matter more simply, the hereditarian asserts that one is in a lower socio-economic class because one is less intelligent, whereas the environmentalist asserts that one is less intelligent because one is in the lower socio-economic class. More specifically, mental retardation is sometimes seen as a major consequence of social deprivation. Such a view assumes that children are capable of “normal” intellectual functioning if we but expose them to enough “cultural enrichment.”

Environmental factors and IQ changes

A matter of considerable import in testing the above hypotheses is the magnitude of change that could be effected as a result of changes in the environment. Many investigators have been relatively pessimistic in their conclusions. McClearn (1962) has pointed out that the magnitude of the difference in IQs attributable to environmental factors, though statistically significant, has been so minute as to be practically negligible.

In support of the environmentalistic point of view, however, instances can be found where rather marked improvements in IQ have been reported following some type of environmental manipulation. The reader is referred to the review by McCandless (1964), whose statement is perhaps one of the strongest in favor of the environmentalistic position.

Other studies have indicated that when a geographic area is subjected to social improvement, such as better schools and improved communication, there is a tendency for the IQs of all the inhabitants to improve. Wheeler’s study of Tennessee mountain children (1942) is of considerable interest. Testing over three thousand subjects in 1930, he found that IQs progressively declined from a mean of 95 at age 6 to a mean of 74 at age 16. Testing a new sample ten years later, he found a mean increase in IQ of approximately 10 points at every age level. However, the steady decline with age, from a mean of 103 at age 6 to a mean of 80 at age 16, was again discovered, despite the general increase.

There has been a certain inconsistency in studies that have attempted to relate IQ changes to environmental factors. In some instances, significant correlations have been found between various subjective ratings of the “goodness” of the environment and increase in IQ (e.g., Thorpe 1946). But in other instances no environmental correlates could be found to account for changes in the IQ (e.g., H. Jones 1946). Jones has given some especially striking case histories of children who have manifested marked changes in IQ without the apparent involvement of environmental factors.

A continuing problem has been the failure to designate just what constitutes a good environment for optimal intellectual development. Little has been added to the implicit view that the American middle-class home represents some sort of standard. A related matter, of course, is the problem of defining cultural or social deprivation. The social deprivation concept has been loosely applied to certain events in early childhood which are characterized as antecedent to certain social behaviors. There is little agreement about either the early events or the resultant behaviors. Major dimensions of childhood deprivations that have been suggested are social isolation, cruelty and neglect, institutional upbringing, adverse childrearing practices, and separation experiences across a wide range of severity. Even factors such as these need much further definition and clarification.

The view that, given a fairly standard environment, it is extremely difficult to improve the quality of cognitive functioning is consistent with the bulk of findings resulting from efforts to improve children’s performance on Piaget-type tasks. Of course, familial retardates do not come from what we consider standard environments. Even with these children there is considerable evidence that no great intellectual improvement is produced through environmental manipulation, and this holds true for a variety of techniques. The reader is referred to E. E. Doll’s excellent history of mental retardation (1962) for evidence on this point. Binet, with his concept of “mental orthopedics,” and Jean M. L. Itard, with his great faith in the possibility of improving the quality of intellect, were responsible for the philosophy underlying the early work with retardates in this country. After several years of employing a variety of techniques, many of which are today being rediscovered, it became apparent that this optimism was unwar-ranted. In the early days, training schools in this country were just what the name implies. They became custodial institutions only when it became apparent that many retardates could not be trained to a level that would make them self-sustaining in the society at large. A reaction appears to have set in at this time, and the view that we could do nothing for retardates except provide them with a comfortable domicile became dominant. There is much for contemporary workers to learn from this marked swing in attitude toward the retarded. It suggests that undue optimism is dangerous, since it breeds undue pessimism.

The conclusion that may be reached concerning the relevance of the heredity and environment controversy for mental retardation has been well stated by Penrose, who after a lifetime of work with the retarded wrote:

The most important work carried out in the field of training defectives is unspectacular. It is not highly technical but requires unlimited patience, good will and common sense. The reward is to be expected not so much in scholastic improvement of the patient as in his personal adjustment to social life. Occupations are found for patients of all grades so that they can take part as fully and usefully as possible in human affairs. This process, which has been termed socialization, contributes greatly to the happiness not only of the patients themselves but also of those who are responsible for their care. ([1949] 1963, p. 282)

It is perhaps within this area of socialization that we can do a great deal to enhance the everyday effectiveness of the retarded. Personality and character traits were discovered to be more influenced by environment than was intellectual level (e.g., Leahy 1935). Such findings bolster the argument that there are many modifiable factors which are important in the determination of social adjustment. It is not rare to encounter individuals with the same intellectual make-up demonstrating quite disparate social adjustments. Perhaps the question is not how to improve the cognitive functioning of familial retardates but rather how to maximize the adjustment of such individuals, whatever their intellectual capacity may be. That considerable change in performance can result from the manipulation of nonintellective (i.e., motivational) factors will be made clear in subsequent passages.

A two-group conception

Hirsch has asserted that we will make little headway in understanding individual differences in intelligence and many other traits unless we incorporate into our thinking the fact that to a large degree such differences reflect the inherent biological properties of man. As Hirsch has noted, we can no longer make the “gratuitous uniformity assumption that all genetic combinations are equally plastic and respond in like fashion to environmental influences ...,” and he added that “without an appreciation of the genotypic structure of populations, the behavioral sciences have no basis for distinguishing individual differences that are attributable to differences whatsoever where there is a common history” (1963, p. 1442).

Work in population genetics appears capable of bringing considerable order to the area of mental retardation. We need simply to accept the generally recognized fact that the gene pool of any population is such that there will always be variations in the behavioral or phenotypic expression of virtually every measurable trait or characteristic of man. From the polygenie model advanced by geneticists, we would deduce that the distribution of intelligence would be a symmetric bell-shaped curve, which is characteristic of such a large number of distributions that we have come to refer to it as the normal curve. This theoretical distribution is a fairly good approximation of what is actually encountered in the observed distribution of intelligence. In the polygenic model of intelligence (see Hirsch 1963; Penrose 1949), the genetic foundation of intelligence is not viewed as dependent upon a single gene. Rather, intelligence is viewed as the result of a number of discrete genetic units. (This is not to assert, however, that singlegene effects are never to be encountered in mental retardation. As noted earlier, certain relatively rare types of mental retardation are the products of such simple genetic effects.)

A variety of specific polygenic models have been advanced that generate theoretical distributions of intelligence that are congruent with observed distributions (e.g., Burt & Howard 1956). Again caution is in order. An environmentalistic model positing five environmental factors acting additively would also generate an approximation to a normal curve. However, such a model appears much less capable of encompassing the raw data encountered in investigations of intelligence. An aspect of polygenic models of special interest for the area of mental retardation is that they generate IQ distributions ranging approximately from 50 to 150. Since an IQ of approximately 50 appears to be the lower limit for familial retardates, it has been concluded (e.g., Burt & Howard 1956; Penrose 1949) that the etiology of this form of retardation reflects the same factors that determine “normal” intelligence. Approached in this way, the familial retardate can be seen as normal, where “normal” is defined as representing an integral part of the distribution of intelligence that we would expect from the normal manifestations of the genetic pool in our population. Within such a framework, it is possible to refer to the familial retardate as less intelligent but it would make no sense to say that he is abnormal. He is just as integral a part of the normal distribution as are the 3 per cent of the population that we view as superior or that more numerous group of individuals that we consider to be average.

The two-group conception of mental retardation calls attention to the fact that the second group of retardates, those who have known physiological defects, represents a distribution of intelligence with a mean which is considerably lower than that of the familial retardates. Such children, for the most part, fall outside the range of normal intelligence, that is, they have an IQ below 50, although there are certain exceptions; brain-damaged children with IQs as high as 150 have been found. Thus the empirical distribution of intelligence may best be represented by two curves. Considerable clarity could be brought to the area of mental retardation if we were to do away with the practice of conceptualizing the intelligence distribution as a single continuous normal curve. The more appropriate representation is to depict the intelligence of the bulk of the population, including the familial retarded, as a normal distribution having a mean IQ of 100 with lower and upper limits of approximately 50 and 150. Superimposed on this curve would be a second nearly normal distribution having a mean IQ of approximately 35 and a range from 0 to 70. The first curve would represent the polygenie distribution of intelligence; the second would represent all those individuals whose intellectual functioning reflected factors other than the normal polygenic expression (i.e., those retardates for whom there is an identifiably physiological defect).

This two-group approach to the problem of mental retardation has been supported by Penrose (1949) among others. The very nature of the empirical distribution of IQs below the mean, especially in the 0-50 range (see Penrose 1949) seems to demand such an approach. This distribution is exactly what we would expect if we combined the two distributions discussed above, as is the general practice. This two-group approach is of particular significance to the issue of mental retardation, since it calls for a reappraisal of the entire concept of normality. Hirsch has pointed out that such a concept, as presently employed, is of little value:

Implicit in our use of “normal” is reference to some region of a distribution arbitrarily designated as not extreme—for example, the median 50 percent, 95 percent, or 99 percent. We choose such a region for every trait. Among n mathematically independent traits– for example, traits dependent on n different chromosomes—the probability that a randomly selected individual will be normal for all n traits is the value for the size of that region raised to the nth power. Where “normal” is the median 50 percent and n = 10, on the average only one individual out of 1024 will be normal (for ten traits). (1963, p. 1437)

Thus, if we consider the whole person with his many variable physiological and psychological systems, it would be extremely rare to find an individual we would consider normal. Indeed, if we were to find him, his very normality would be considered abnormal in the sense that he represented a rare event. In the area of mental retardation the concept of abnormal should be confined to those cases with known physical defects wherever these cases may be found in the distribution of intelligence.

A two-group approach makes the problem of the etiology of the familial retarded just as assailable as the problem of etiology in pathological retardation. In respect to the etiology of familial retardates, McClearn has stated that “these individuals undoubtedly represent the lower tail of the distribution generated by assortment of the polygenes underlying ‘normal’ intelligence, and should no more be considered abnormal than those whose intelligences are an equal distance above the mean” (1962, p. 186).

Once we adopt the position that the familial mental retardate is not defective or pathological but is essentially a normal individual of low intelligence, then the problem of familial retardation becomes part of the general problem of developmental psychology. In terms of cognitive development, the familial retardate would then be viewed as progressing from one intellectual stage to the next in the same sequence as is encountered in other children. He would, of course, progress from stage to stage at a slower rate than other children, and the final stage that he achieves would be lower than that achieved by the more intelligent members of the population. In terms of cognitive functioning alone, the familial retardate with a chrono-logical age (CA) of 10 and an MA of 7 would be conceptualized as being cognitively similar, that is, at the same developmental level as a child with a CA of 7 and an IQ of 100. (The reader must remember that the MA, which is invariably based on the IQ, can be considered only a very rough indicator of the cognitive or developmental level; however, to date, it represents the most adequate measure available.)

To say that two such hypothetical children are cognitively similar is not to assert that they will necessarily behave exactly the same on the intellective and nonintellective tasks with which society confronts them. If nothing else, the retardate is three years older, and if the performance involves lifting weights, or a task that he has encountered much more frequently than the 7-year-old normal child, we would expect the retardate to be superior. Furthermore, performance on even cognitive tasks reflects the wide variety of factors that are the product of the past history of the child rather than his cognitive ability alone. To the extent that these two children have different histories, have experienced different environments, and have developed different values and motives, we would expect differences in performance.

It is no great mystery that a group of children with IQs of 70 and a group with IQs of 100 matched on chronological age differ on a variety of tasks. These children are at different developmental levels, and such differences are exactly what a developmentalist would expect. The mystery is the repeated demonstration that even when groups are matched on MA, the retardate does less well, or at least behaves differently, than the MA-matched “normal” child. Two distinctly different explanations for this phenomenon have been advanced. One view is that these differences reflect a variety of experiential or motivational differences. The second position is that the familial retardate is really not a normal individual developing at a slower rate but is rather an inherently different or abnormal type of organism who, at every level of development, is suffering from some defect in his physiological or cognitive structure. These hypothe-sized defects are then viewed as producing differences in behavior even in those instances where the MA is equated. In the next section we shall consider the defect orientation, and the motivational position will be discussed further in the final section.

The defect and difference orientation

This section deals with those theoretical and empirical efforts that have advanced the view that all retardates, including those conventionally diag-nosed as familial, suffer from some specifiable defect. These efforts are in opposition to the view that the familial retardate suffers from nothing more than a slower and more limited rate of cognitive development. The evidence typically offered by the difference, or defect, theorist is that even when groups of normals and retardates are matched on MA, which grossly controls for differences in the rate of development, the two groups behave differently. This difference in behavior is advanced as proof of the existence of some physiological or cognitive defect which itself is responsible for the slower rate of development. Where the hypothe-sized defect is an explicitly physiological one, it would appear to be a simple matter to obtain direct validation for the defect’s existence. Such evidence would come from biochemical and physiological analyses as well as from pathological studies of familial retardates. A number of such studies have, of course, been carried out. Although there is an occasional report of some physical anomaly, the bulk of the evidence has indicated that the familial retardate does not suffer from any gross physiological defects. Indeed, if such evidence were readily available, the defect theorist would give up his reliance on the more ambiguous data provided by studies examining molar behavior. The failure to find direct evidence for the existence of a physio-logical defect in the familial retarded has not deterred—and probably should not deter—theorists from postulating such defects.

In spite of the negative physiological evidence, such workers as Spitz (1963) maintain that all retardates, including familials, are physically defective and that our failure to discover defects in the familial retarded is due to the relatively primitive nature of our contemporary diagnostic techniques. It is perfectly legitimate for these workers to assert that, although presently not observable, the physical defect that causes familial retardates to behave differently from normals of the same MA will some day be seen. These theorists operate very much like the physicists of a not-too-distant era who asserted that the electron existed, even though it was not directly observable. Analogously, defect theorists in the area of mental retardation validate the existence of a defect by first asserting that its existence should manifest itself in particular phenomena, that is, in particular behaviors of the retarded. They then devise experiments in which, if the predicted behavior is observed, the existence of the hypothesized defect is confirmed. This approach is legitimate and has become increasingly popular.

The majority of theories in the area of mental retardation are basically defect theories. It should be noted that these theories differ among themselves. One difference involves the theoretician’s effort to relate the postulated defect to some specific physiological structure. The theoretical language of some defect positions is explicitly physiological, that of others is nonphysiological, while that of others has remained extremely vague. Such differences are related to the specific nature of the defect postulated. Particular defects that have been attributed to the retarded include the relative impermeability of the boundaries between regions in the cognitive structure (Kounin 1941; Lewin 1926-1933); primary and secondary rigidity caused by subcortical and cortical malformations, respectively (Goldstein 1943); inadequate neural satiation related to brain modifiability or cortical conductivity (Spitz 1963); malfunctioning disinhibitory mechanisms (Siegel & Foshee 1960); improper development of the verbal system resulting in a dissociation between verbal and motor systems (Luria 1963; O’Connor & Hermelin 1959); and the relative brevity in the persistence of the stimulus trace (Ellis 1963).

Luria and verbal mediation theory

Some of the more influential of the defect positions will be examined here, turning first to the position of the Russian investigator, A. R. Luria, whose work has influenced investigators in England and the United States. In respect to Russian efforts it should be noted that, given the political philosophy of the U.S.S.R., workers in the area of mental retardation have no alternative but to accept a defect position. As in the United States, the Russians divide the retarded into three groups, although they use the older terms “idiot,” “imbecile,” and “debile.” However, the generic term for mental retardation is “oligophrenia.” The practice, followed in this article, of distinguishing between the group of approximately 25 per cent of retardates having known organic impairments and that larger group having unknown etiologies is simply not permitted by Soviet investigators.

Although this section of the article is directed at illuminating differences of opinion concerning this larger group, there is a general consensus in the United States that this type of retardate, which we conventionally classify as familial, is the product of complex genetic determinants and cultural deprivation. In contrast, as the subcommittee of the President’s Panel which recently visited the U.S.S.R. has noted (see Mental Retardation in the Soviet Union 1964), Soviet philosophy does not accept the view that mental retardation is deter-mined by genetic factors, nor is cultural causation considered a possible explanation. Thus workers in this area attribute all grades of mental retardation to central-nervous-system damage, suggesting that it occurs initially during the intrauterine period or during early childhood and then results in a disturbance of the child’s subsequent mental development.

It is clear, then, that in the Soviet Union the diagnosis of mental retardation necessarily involves the specification of a defect in some neuro-physiological system, and it is noteworthy that professionals, including researchers and teachers, working with the retarded are called “defectologists.” Knowledgeable visitors (see “Mental Retardation in the Soviet Union” 1964) have pointed out that, given such an approach, diagnosticians will go to great lengths to “discover” some slight indication of possible organicity. However, in observing the pupils in the Soviet schools for the debile, it was apparent that those in attendance were primarily retardates who would be diagnosed as familial in the United States. With rare exceptions, these were the children of unskilled workers and in some instances were actually the children of the graduates of such schools.

Consistent with the over-all Russian philosophy, general intelligence tests have been banned since 1936 by the Communist party because such tests are considered to be methods which discriminate against the peasants and the working class in favor of the culturally advantaged. Diagnosis in mental retardation is made by neurologists and psychophysiologists, who rely heavily on gross pathological signs in the case of the severely retarded and minor physical defects, minute examinations of electroencephalograph (EEC) patterns, and certain qualitative (nonstandardized) tests of perception, conditioning, and concept formation (with special emphasis on the identification of specific types of language disorders) in the case of the more mildly retarded.

Luria’s efforts and the difficulties they pose for non-Russian workers can only be understood in terms of such an orientation toward mental retardation. In his work on verbal mediation, Luria has demonstrated that the behavior of retardates resembles that of chronologically younger normal children in that the verbal instructions do not result in the smooth regulation of motor behavior. His findings clearly indicate that on all his tasks requiring verbal mediation, the retarded subjects have considerable difficulty. In light of these behavioral data, Luria has inferred that the major defect in the retarded child involves both an under-development or a general “inertness” of the verbal system and a dissociation of this system from the motor or action system. The general effect of this dissociation, vaguely conceptualized as a disturbance in normal cortical activity, is that a verbal response cannot serve as an adequate regulator of voluntary behavior.

Unfortunately, it is impossible to utilize Luria’s data to throw any light on the issue of whether the cognitive processes of retardates, typically diagnosed as familial, differ from normal children of the same MA. As noted earlier, a Russian defectologist would not accept this as a legitimate question. Since there is no concern with the IQ, there is no way to determine the MAs of the retardates and normals compared in Luria’s work. Furthermore, the etiological question of whether his retarded subjects are of the physiologically impaired or the familial-cultural type remains unanswered. However, in light of Luria’s discussion of “profound atrophic changes... expressed in the under-development of the complex neuron structures of the first and third strata of the cortex’ and his classification of these retardates as imbeciles rather than debiles, it would appear that these subjects probably suffer from gross physiological impairment.

It must be concluded, then, that these data have extremely limited relevance to the issue of whether those retardates whom we conventionally classify as familial suffer from some physiological defect. We must therefore look to English and American workers for more adequate tests of the basic proposition that all retardates, including the familial, differ from normals in the degree to which they employ verbal cues in regulating voluntary behavior. For example, O’Connor and Hermelin (1959) have found no significant difference between normals and retardates in the number of trials required to learn a size discrimination. however the finding that retardates required significantly fewer trials to learn a reversal was interpreted as supporting Luria’s position. O’Connor and Hermelin reasoned that on the original learning task the normal child employs both motor and verbal mediational responses in his learning, while the retarded child relies primarily on the motor response. When the reversal is introduced, the normal child must unlearn both the original motor and verbal responses. The retardate, having to un-learn only the motor response, would thus be expected to learn the reversal problem more easily. The findings of O’Connor and Hermelin are trouble-some in light of their inconsistency with earlier studies (e.g., Stevenson & Zigler 1957) in which mental retardates were not found to be superior on a discrimination reversal task. In an effort to resolve this discrepancy, Balla and Zigler (1964) ran a reversal-learning study involving several different reversal tasks and different types of retardates at different MA levels. This study provided no support for the Luria position.

Milgram and Furth (1963) compared retarded and normal children of the same MA on a series of concept tasks assumed to vary in the degree to which language might facilitate performance. Their findings were consistent with expectations derived from Luria’s position. However, in an experiment comparing retardates and normals of the same MA on their ability to employ verbal mediators, Rieber (1964) obtained findings that were inconsistent with those that would be derived from Luria’s theory. It thus appears that the evidence to date which has been mustered to support Luria’s position remains equivocal.

Spitz and cortical satiation theory. Another major defect position is that of Herman Spitz (1963), who has extended the Köhler-Wallach cortical satiation theory to the area of mental retardation. Spitz has argued that all retardates suffer from inadequate neural satiation which is related to brain modifiability or cortical conductivity, and has tested this position by comparing normals and retardates of the same CA. Again it should be noted that no direct physiological evidence has been presented to indicate that familial retardates suffer from inadequate neural or cortical functioning. Furthermore, there is direct physiological evidence (Lashley et al. 1951) which calls into question the validity of the entire Kohler-Wallach position [seeGestalt Theory].

As in the case of the earlier gestalt workers, Spitz has primarily employed perceptual tasks to test his position. His extensive program of research has now been summarized (Spitz 1963), and any complete review would be beyond the scope of this paper. Spitz’s most convincing evidence has been obtained on those perceptual tasks (e.g., figural aftereffects and Necker cube reversals) that are thought to be sensitive to hypothesized cortical satiation effects.

The heuristic value of Spitz’s position can be seen in his recent efforts to extend his postulates beyond the visual perception area and to employ them to generate specific predictions concerning the phenomena of learning, transposition, generalization, and problem solving. Spitz has noted a number of studies in these various areas which lend credence to his basic position. He has also been quite explicit in noting the limitations of his view. He has pointed out that, contrary to his theory, cortical satiation as measured by his perceptual indexes does not “in general correlate with IQ, but rather only differentiates the average performance of two distinct groups.” The extensive overlap between normals and retardates on his tests of satiation led him to conclude that “the satiation variable must be only a very small one in the total complex of intelligent behavior.”

Spitz has also been appropriately concerned with the fact that the test-retest reliability of the scores of his retardates is not impressive. Furthermore, he has noted that the lack of any correlation of individual scores across certain of his satiation tasks is troublesome for his position. Across modalities and even in the same modality, correlations have been moderate or nonexistent. In addition to these concerns, Spitz has been sensitive to the issue of how accurately the subject’s response, often a verbal report, reflects the perceptual response being investigated. (See Spivack 1963 for a discussion of this problem in respect to research with the retarded.)

Adding to these difficulties is the fact that several investigators have now discovered that responses to cognitive and perceptual tasks are influenced by a variety of motivational factors (e.g., Zigler & deLabry 1962; Zigler & Unell 1962). In addition certain aspects of Spitz’s work have come in for criticism on the grounds that his findings are inconsistent with those of other investigators. Spivack (1963) has voiced this concern in a review of research on perceptual processes in the retarded, noting that certain of Spitz’s findings “are in marked contrast to the findings of others.”

Of more importance to the central question of this section is the conclusion that Spitz’s data throw little light on the issue of whether familial retardates are inherently different from normals of the same MA. Taking a stand reminiscent of the Russian position, Spitz has argued that the distinction between familial and organic retardates is misleading. In Spitz’s view, all retardates suffer from brain damage in the broader sense, and he has argued (see Garrison 1966) that retardates be conceptualized as belonging to a common class. Therefore his work has been characterized by a relative lack of concern with the problem of etiology, and we have little way of assessing whether the differences he reports are a product of gross organic pathology or may actually reflect the cortical phenomena that Spitz postulates.

That one finds differences between normals and retardates matched on CA is not very surprising, since we are dealing with groups who are at different developmental levels (as defined by MA). One would be tempted to say that Spitz’s work has little relevance to the central issue of this section except for the fact that he has been quite explicit in his view that the differences he obtains are not developmental phenomena but reflect a physical deficit that should manifest itself even in comparisons with MA-matched normals.

The Lewin-Kounin formulation

The final defect position that we shall discuss is that of Lewin (1926-1933) and Kounin (1941). This position is different from the other defect views in that the defect postulated is one in the cognitive structure rather than the physical structure of the retardate. The Lewin-Kounin formulation has had considerable impact not only on our conceptualization of the retarded but also on the treatment and training practices that have been employed over the years. (For a more complete historical review and critique of the Lewin-Kounin formulation, the reader is referred to Zigler 1962.)

In Lewin’s general theory the individual is treated as a dynamic system with differences among individuals derivable from a diversity of (1) structure of the total system, (2) material and state of the system, or (3) meaningful content of the system. The first two of these factors play the most important role in Lewin’s theory of retardation. Lewin viewed the retarded child as having a less-differentiated cognitive structure, that is, having fewer regions or cells, than a normal child of the same CA. Thus, in terms of structure, the retarded child resembles a normal younger child. In relation to the material and state of the system, Lewin stated that even though a retarded child corresponded in degree of differentiation to a normal younger child, these children were not to be regarded as entirely similar. He considered “the major dynamic difference between a feebleminded and normal child of the same degree of differentiation to consist in a greater stiffness, a smaller capacity for dynamic rearrangement in the psychical systems of the former.” (Degree of differentiation was later operationally defined as MA.)

Although Lewin undoubtedly felt that lack of differentiation could lead to rigid behaviors (e.g., pedantry, fixation, stereotypy, inelasticity, perseveration), he was quite clear that this lack of differentiation was not what he meant by rigidity. To Lewin, lack of differentiation referred to the number of regions within the total system, while rigidity was defined in terms of the fluidity between regions. (By rigidity, Lewin was referring to the nature of the boundary between cells in the cognitive structure.) It follows from Lewin’s theory that an individual whose system is characterized by either lack of differentiation or rigidity, or both, is more likely to emit behaviors commonly referred to as rigid. The failure to draw a clear distinction between the meaning of rigidity as he employed it and rigid behaviors as such appears to be a major factor leading to the subsequent controversy in the area.

The clearest experimental support for the position that familial retarded individuals are more rigid than normal individuals having the same degree of differentiation is contained in the work of Kounin (1941; 1948). Kounin, building upon Lewin’s work, advanced the view that rigidity is a positive, monotonic function of CA. Again, it is imperative to note that by rigidity Kounin, like Lewin, referred to “that property of a functional boundary which prevents communication between neighboring regions” and not to rigid behaviors as such.

Kounin (1941) offered the findings of five experiments in support of his theory. In these experiments he employed three groups, older familial retarded individuals, younger familial retarded individuals, and normals. Noting the inadequacies of Lewin’s own experimental efforts, Kounin instituted certain experimental controls. He defined the degree of differentiation as the MA of an individual and controlled for this factor by equating the three groups on MA. He also attempted to reduce what he later referred to as “motivational factors (such as low success expectation and hesitance to enter unfamiliar regions) that might produce those very types of behavior that are sometimes lumped together in the pseudo-descriptive category of behavioral rigidity” (Kounin 1948). To control for these factors, Kounin attempted to make each subject feel confident and secure in the experimental tasks by having them engage in each of the activities prior to the experiment proper. As Kounin predicted, the three groups differed in certain instruction-initiated tasks (e.g., drawing cats until satiated and then drawing bugs until satiated, lowering a lever in order to release marbles and then raising the lever to release marbles). As predicted from the Lewin-Kounin formulation, the normals showed the greatest amount of transfer effects from task to task, the younger retarded a lesser amount of transfer, and the older retarded the least amount of transfer. That is, on the drawing task the retarded individuals drew longer on the second task following satiation of the first task than did normals, and the older retardates longer than the younger. On the lever-pressing task, the greatest number of errors, that is, lowering rather than raising the lever on part two, were made by the normals, the least number by the older retarded, with the younger retarded falling between these two groups.

One should note that on this last task the lesser rigidity, as defined by Lewin and Kounin, of the normals results in a higher incidence of a behavioral response often characterized as rigid (i.e., perseverative responses). Furthermore, this lack of influence of one region upon another in the performance of the retarded would only be predicted in those cases where the retarded individual is “psychologically” placed into a new region by employing an instructional procedure. In those instances where the individual must, on his own, move from one region to another, the LewinKounin formulation would predict that such movement would be more difficult for the retarded than for the normal individual. This prediction was also confirmed by Kounin in his concept-switching experiment in which the child was asked first to sort a deck of cards, which could be sorted either on the basis of color or form, and then to put the cards together some other way. Here the normals evidenced the least difficulty in shifting, the older retarded the most difficulty, and the younger retarded group again fell between the other two groups. Thus, when a movement to a new region is self-initiated, it is the retarded who evidence the higher incidence of perseverative responses.

The Lewin-Kounin theory of rigidity is a conceptually demanding one in that it sometimes predicts a higher and sometimes a lower incidence of “rigid” behaviors in retarded as compared to normal individuals. However, the fact that it generates specific predictions as to when one or the other state of affairs will obtain is a tribute to the theory. Kounin thus offered impressive experimental support for the view that, with MA held constant, the older or more retarded (or both) an individual is, the more will his behaviors be characterized by dynamic rigidity, that is, greater rigidity in the boundaries between regions.

This model and its experimental support was so impressive that until fairly recently very few further experimental tests were attempted. However, recent explicit tests of the model (Balla & Zigler 1964; Stevenson & Zigler 1957; Zigler & Unell 1962) have failed to provide support for it. Much evidence now indicates that the differences found by Kounin were not a product of the inherent rigidity of retardates, but rather reflected a number of motivational differences between normal children and institutionalized retardates of the same MA. These motivational factors will be discussed in the final section.

Motivational and emotional factors

A recurring theme in the present article has been the importance of a variety of nonintellective factors as determinants of the level at which the retarded functions. We shall never comprehend the behavior of the retarded if we assume that every behavior he manifests is the immutable product of his low intelligence. Furthermore, we must go beyond the overly simplistic theories that have been advanced, such as the view that all retardates manifest a highly similar pattern of behavior which is determined by their common defect. In-deed, a striking feature encountered when groups of retardates are observed is the variety of behavior patterns displayed. Clearly, we are not dealing with a homogeneous group of simple organisms. Once we concern ourselves with the total behavior of the retarded child, we find him an extremely complex psychological system. To the extent that his behavior deviates from the norms associated with his MA, he is even more difficult to understand than the normal individual.

It is unfortunate that so little work emanating from a personality point of view has been done with the retarded. Some progress has been made, however, and much of the recent work supports the view that it is not necessary to employ constructs other than those used to account for the behavior of normal individuals in explaining the behavior of the familial retarded. It appears that many of the reported differences between retardates and normals of the same MA are a result of motivational and emotional differences which reflect differences in environmental histories and are not a function of innate deficiencies.

That personality factors are as important in the retardate’s adjustment as are intellective factors has been noted (e.g., Penrose 1949; see also Windle 1962 for an especially comprehensive review of the importance of nonintellective factors in the prognosis of mental retardation). Many of the early workers in this country felt that the difference between social adequacy and inadequacy in that large group of borderline retardates was a matter of personality and character rather than intelligence. A number of studies have confirmed this view (see Windle 1962). Perhaps the best of these is the comprehensive study by Weaver (1946) of the adjustment of 8,000 retardates inducted into the U.S. Army, most of whom had IQs below 75. Of the total group, 56 per cent of the males and 62 per cent of the females made a satis-factory adjustment to military life. The median IQs of the successful and unsuccessful groups were 72 and 68 respectively. Weaver concluded that “personality factors far overshadowed the factor of intelligence in the adjustment of the retarded to military service.”

This tendency to overemphasize the importance of the intellect in adjustment has been made clear by Windle (1962). On the basis of a survey, he found that most institutions presume that intelligence is the critical factor in adjustment after re-lease. Windle goes on to point out that the vast majority of studies (over 20) on outcome “after release from institutions have reported no relation between intellectual level and later adjustment.” In examining this literature we find that the factors which led to poor social adjustment include anxiety, jealousy, overdependency, poor self-evaluation, hostility, hyperactivity, and failure to follow orders even when requests were well within the range of intellectual competence.

It is hardly surprising that retardates evidence such difficulties in light of their atypical social histories. The specific atypical features of their socialization histories and the extent to which they are atypical may vary from child to child. Two sets of parents who are themselves familially retarded may provide quite different socialization histories for their children. At one extreme we may find a familially retarded child who grows up in an abysmal home environment and who is ultimately institutionalized, not because of lack of intelligence but because his own home represents such a poor environment. That many borderline retardates are institutionalized for just such reasons has been con-firmed by Kaplun (1935) in a study of 642 high-grade retardates; Zigler’s recent finding (1961) that a positive relationship exists between the institutionalized familial retardate’s IQ and the amount of preinstitutional deprivation he experienced provides further support for this claim. This latter finding does not indicate that social deprivation produces greater intelligence but rather that our institutions contain borderline retardates who would not be institutionalized except for their extremely poor home environments.

At the other extreme, the familially retarded set of parents of an institutionalized child may have provided him with a relatively normal home, even though it might differ in certain important respects (e.g., values, goals, and attitudes) from the typical home in which the families are of average or superior intelligence. In the first example the child not only experiences a quite different socialization history while still living with his parents, but he also differs from the child in the second situation to the extent that institutionalization affects his personality structure. Given the penchant of many investigators for comparing institutionalized retardates with children of average intellect who live at home, the factor of institutionalization becomes an extremely important one. One cannot help but wonder how many differences discovered in such comparisons reflect some cognitive aspect of mental retardation as opposed to the effects of institutionalization, the factors that led to the child’s institutionalization, or some complex interaction between these factors and institutionalization.

To add even more complexity, the socialization histories of both institutionalized and noninstitutionalized familial retardates differ markedly from the history of the brain-damaged retardates. The brain damaged do not show the same gross differences in the frequency of good versus poor environments as do familials. In the face of such complexity, we need not consider the problem unassailable, nor need we assert that each retarded child is unique and that it is therefore impossible for us to isolate the ontogenesis of those factors which we feel are important in influencing the retardate’s level of functioning. Once we conceptualize the retardate as occupying a position on a continuum of normality, we can allow our knowledge of normal development to give direction to our efforts.

This does not mean that we ignore the importance of the lowered intelligence per se, since per- sonality traits and behavior patterns do not develop in a vacuum. However, in some instances the personality characteristics of the retarded will reflect environmental factors that have little or nothing to do with intellectual endowment. For example, many of the effects of institutionalization may be constant regardless of the person’s intelligence level. In other instances, we must think in terms of an interaction; that is, given his lowered intellectual ability, a person will have certain experiences and develop certain behavior patterns differing from those of a person with greater intellectual endowment. An obvious example is the greater amount of failure which the retardate typically experiences. But again what must be emphasized is that the behavior pattern developed by the retardate as a result of such a history of failure will not differ in kind or ontogenesis from those developed by an individual of normal intellect who, by some environmental circumstance, also experiences an in-ordinate amount of failure. By the same token, if the retardate can somehow be guaranteed a more typical history of success, we would expect his behavior to be more normal, independent of his intellectual level. Within this framework, the author will discuss the personality factors which have been known to influence the performance of the retarded.

Caution is needed in evaluating the role of motivational and emotional factors in the performance of the retarded. Performance on a task is most appropriately conceptualized as a function of two types of factors, intellective (i.e., cognitive) and nonintellective (i.e., motivational). The contribution of each factor will vary with the nature of the task. Motivational factors will more readily influence a perseveration task (e.g., how long a retardate will continue to put marbles into a box) than they will a discrimination-learning or concept-formation task. It has been demonstrated that the performance of retardates on tasks of the latter type is also influenced by motivational factors (Butterfield & Zigler 1965a; Zigler & deLabry 1962), but this should not be interpreted as evidence that basic intellectual capacity has been changed. Rather, these demonstrations suggest ways in which one may help the mentally retarded to utilize their intellectual capacity optimally. As such, they should not be viewed as manipulations which can make the retarded “normal” in their intellectual functioning.

Anxiety

Considerable evidence has now been collected indicating the importance of anxiety on performance for a wide variety of tasks (Taylor 1956). The attenuating effects of anxiety on performance appear to be a function of both the task-irrelevant defensive responses employed by the person to alleviate his anxiety and the drive features of anxiety itself. The drive approach to anxiety (Taylor 1956), which has received considerable confirmation, conceptualizes high anxiety as beneficial on extremely nondemanding tasks (e.g., classical eyelid conditioning) but detrimental on complex tasks where a variety of responses are available to the person. The higher anxiety level of retardates, as compared to normals, has now been noted by several investigators (e.g., Garfield 1963) who have either demonstrated or suggested that the heightened anxiety level of retardates could well have produced certain of the differences between retardates and MA-control normals reported in the literature. Work with retardates that has either focused on anxiety or raised the anxiety issue in a post hoc manner is of considerable value “in that it applies concepts and techniques to the study of retarded individuals, which for the most part had not been applied or seen as relevant for this group” (Garfield 1963, p. 594). [SeeAnxiety.]

The facts that anxiety level affects the performance of retardates much as that of normals and that retardates might have higher levels of anxiety than normals tell us little about the ontogenesis of anxiety in retardates. To understand their atypical anxiety levels, we must examine the relatively atypical experiences of the retarded, as well as a variety of other motivational states which influence their performance.

Social deprivation

It has now become increasingly clear that our understanding of the performance of the institutionalized familial retardates will be enhanced if we consider the inordinate amount of preinstitutional social deprivation they have experienced (Clarke & Clarke 1954; Kaplun 1935; Zigler 1961). A series of recent studies (Green & Zigler 1962; Zigler 1961; Zigler & Williams 1963) has indicated that one result of such early deprivation is a heightened motivation to interact with a supportive adult. (In the process of conducting these studies, a social deprivation scale was constructed which promises to bring some added objectivity to the social deprivation concept.) These studies suggest that, given this heightened motivation, retardates exhibit considerable compliance with instructions when the effect of such compliance is to increase or maintain the social interaction with the adult. Compliance is apparently reduced in those instances where it leads to terminating the interaction.

It now appears that the perseveration so frequently noted in the behavior of the retarded is primarily a function of this motivational factor rather than the inherent cognitive rigidity suggested by Lewin (1926-1933) and Kounin (1941). Evidence on this latter point comes from findings indicating that (1) the degree of perseveration is directly related to the degree of preinstitutional deprivation experienced (Zigler 1961) and (2) institutionalized children of normal intellect are just as perseverative as institutionalized retardates, while noninstitutionalized retardates are no more perseverative than noninstitutionalized children of normal intellect (Green & Zigler 1962). The finding that institutionalization (or the social history factors leading to institutionalization) is the crucial factor in determining the child’s response to social reinforcement on a simple task has also been found by Stevenson and Fahel (1961). The heightened motivation to interact with an adult, stemming from a history of social deprivation, would appear to be consistent with the often-made observation of certain behaviors in the retarded, such as seeking attention and wishing for affection (Doll 1962).

It is impossible to place too much emphasis on the role of overdependency in the institutional familial retarded and on the socialization histories that give rise to such over dependency., Given some minimal intellectual level, the shift from dependence to independence is perhaps the single most important factor necessary for the retardate to become a self-sustaining member of our society. It appears that the institutionalized retardate must satisfy certain affectional needs before he can cope with problems in a manner characterized by individuals whose affectional needs have been relatively satiated. These affectional needs can best be viewed as ones which often interfere with certain problem-solving activities. Because the retardate is highly motivated to satisfy such needs through maximizing interpersonal contact, he is relatively unconcerned with the specific solution to these problems. Of course the two goals will not always be incompatible, but in many instances they will be. Some evidence that this attenuating aspect of retarded behavior can be overcome has been presented by McKinney and Keele (1963), who found improvement in a variety of behaviors in the men-tally retarded following an experience of increased mothering.

Zigler and Williams (1963) have provided some evidence on the interaction between preinstitutional social deprivation and institutionalization in influencing the child’s motivation for social inter-action and support. It was found that although institutionalization generally increased this motivation, it was increased much more for children coming from relatively nondeprived homes than for those coming from more socially deprived backgrounds.

Change in IQ scores. An unexpected finding of the Zigler and Williams study was that a general decrease in the IQs of retardates was discovered between the administration of two IQ tests, the first of which occurred at the time of admission five years prior to this follow-up study. This change in IQ, discovered in the context of a study employing the amount of preinstitutional social deprivation as an independent variable, is reminiscent of a finding by Clarke and Clarke (1954). These investigators found that changes in the IQs of retardates following institutionalization were related to their preinstitutional histories. They discovered that children coming from extremely poor homes showed an increase in IQ which was not observed in children coming from relatively good homes. Zigler and Williams, however, found that the magnitude of the IQ change in their subjects was not significantly related to preinstitutional deprivation. Although this finding appears inconsistent with that of Clarke and Clarke, it should be noted that some support for a relationship was suggested, since the only subjects in the Zigler and Williams study who evidenced an increase in IQ were the highly deprived group. The failure of Zigler and Williams to replicate the findings of Clarke and Clarke may be due to two factors: the subjects used by Clarke and Clarke were older and had been institutionalized at a later age than the retardates employed by Zigler and Williams and the IQ changes reported by Clarke and Clarke took place during two years of institutionalization, while the IQ changes reported in the Zigler and Williams study were based on five years of institutionalization. This latter factor becomes increasingly important in view of E. C. Jones and Carr-Saunders’ finding (1927) that normal institutionalized children show an increase in IQ early in institutionalization and then a decrease in IQ with longer institutionalization.

The work of Clarke and Clarke, Jones and Carr-Saunders, and others, dealing with changes in IQ following institutionalization has given central importance to the degree of intellectual stimulation provided by the institution in contrast to that provided by the original home. This orientation suggests that it is the actual intellectual potential of the person which is altered. The Zigler and Williams study, however, suggests that the change in IQ reflects a change in the child’s motivation for social interaction. That is, as social deprivation, resulting from increased length of institutionaliza- tion, increases, the desire to interact with the adult experimenter increases. Thus, for the deprived child the desire to be correct must compete in the testing situation with the desire to increase the amount of social interaction. This argument would appear to provide the conceptual framework for Clarke and Clarke’s finding that highly deprived subjects evidence an increase in IQ with relatively short institutionalization, while the less deprived subjects demonstrate no greater increase than a test-retest control group. One would further expect that with continued institutionalization all children would exhibit a decrease in IQ, the phenomenon found by Jones and Carr-Saunders (1927) and one that appears in the Zigler and Williams study. Direct support for this view comes from the finding in the Zigler and Williams study of a positive relationship between the magnitude of the decrease in IQ and the child’s motivation for social interaction.

It should be noted that the Jones and Carr-Saunders (1927) study involved institutionalized children of approximately average intellect, thus indicating that the dynamics under discussion here are the same for both normal and retarded children. Furthermore, the position advanced here is quite consistent with the findings for normal children obtained by Barrett and Koch (1930); these investigators found that the greatest increase in IQ was obtained by children who showed the greatest improvement in their personality traits or by children who evidenced a marked change in the nature of their relationship with the examiner. Conversely, what must be emphasized in respect to lowered IQs is not the lowered test scores per se but rather that the factors which attenuate these test scores will, in all probability, reduce the adequacy of many problem-solving behaviors performed in a social situation.

Although there is considerable observational and experimental evidence that social deprivation results in a heightened motivation to interact with a supportive adult, it appears to have other effects as well. Again, the nature of these effects is suggested in those observations of the retarded that have emphasized their fearfulness, wariness or avoidance of strangers, or their suspicion and mistrust. The experimental work done by Zigler and his associates on the behavior of the institutionalized retarded has indicated that social deprivation results in both a heightened motivation to interact with supportive adults (positive-reaction tendency) as well as a reluctance and wariness to do so (negative-reaction tendency). That both of these tendencies are influenced by the quality of past relationships with adults and are amenable to experimental manipulations has been demonstrated in recent studies employing both normal and retarded children (e.g., Shallenberger & Zigler 1961). However, little work has been done on the range of behaviors that might be influenced by the negative-reaction tendency.

Failure and performance . Another factor frequently mentioned as a determinant in the performance of the retarded is their high expectancy of failure (Cromwell 1963). This failure expectancy has been viewed as an outgrowth of a lifetime characterized by frequent confrontations with tasks for which the retarded are intellectually ill-equipped to deal. That failure experiences and the failure expectancies to which they give rise affect a wide variety of behaviors in the intellectually normal has now been amply documented. Of special interest to workers in the area of mental retardation is Lantz’s finding (1945) that a relatively simple failure experience prevented children from profiting by practice which ordinarily leads to improvement on intelligence-test scores [seeAchievement Motivation].

The results of experimental work employing the success—failure dimensions with retardates are still somewhat inconsistent. The work of Cromwell and his students (reviewed in Cromwell 1963) has lent support to the general proposition that retardates have a higher expectancy of failure than do normals. This results in a style of problem solving for the retardate which causes him to be much more motivated to avoid failure than to achieve success. However, the inconsistent research findings suggest that this fairly simple proposition is in need of some further refinement. One investigator found that retardates performed better following success and poorer following failure as compared to a control group. Another investigator (Heber 1957) found that the performances of normals and retardates were equally enhanced following both a failure and a success condition, although in the success condition the performance of retardates was enhanced more than that of normals.

Conversely, Kass and Stevenson (1961) found that success enhanced the performance of normals more than that of retardates. Another study also found that failure had a general enhancing effect for both normals and retardates but that failure enhanced the performance of normals more than that of retardates (Gardner 1958). In a recent study by Butterfield and Zigler (1965a), one factor which may have produced this type of inconsistency was isolated. These investigators found that both normal and retarded children reacted differentially to success and failure experiences as a function of their responsivity to adults, that is, their desire to gain an adult’s support and approval. The nature of the difference between normals and retardates in their reaction to success or failure experiences appeared to be determined by this desire for approval. Among high-responsive subjects, failure, as compared to success, attenuated the performance of retarded subjects while improving the performance of normal subjects. Among low-responsive subjects, failure, as compared to success, attenuated the performance of normals while improving the performance of retardates. Debilitating effects of prolonged failure on the performance of the retarded have been found by Zeaman and House (1962). These investigators discovered that following such failure, retardates were unable to solve a simple problem, although they had previously been able to do so. Assuming a failure set in retardates, Stevenson and Zigler (1958) confirmed the prediction that retardates would be more willing to “settle for” a lower degree of success than would normal children of the same MA. The fear of failure in the mentally retarded also appears to be an important factor in differences that have been found between normals’ and retardates’ achievement motivation (Jordan & DeCharms 1959).

Recent studies (Green & Zigler 1962; Turnure & Zigler 1964) have indicated that the high incidence of failure experienced by retardates generates a cognitive style of problem solving characterized by outer-directedness. That is, the retarded child comes to distrust his own solutions to problems and therefore seeks guides to action in the immediate environment. This outer-directedness may explain the great suggestibility so frequently attributed to the retarded child. Evidence has now been presented indicating that, compared to normals of the same MA, the retarded child is more sensitive to verbal cues given by an adult, is more imitative of the behaviors of both adults and peers, and engages in more visual scanning. Furthermore, certain findings (Green & Zigler 1962) suggest that the noninstitutionalized retardate is more outer-directed in his problem solving than the institutionalized retardate. This makes considerable sense if one remembers that the noninstitutionalized retardate does not reside in an environment adjusted to his intellectual shortcomings and should therefore experience more failure than the institutionalized retardate.

Turnure and Zigler (1964) have suggested that the distractability so frequently encountered in the retarded reflects, in part, this outer-directed style of problem solving. This interpretation is of particular interest, since distractability has often been viewed as a neurophysiologically determined characteristic of the retarded rather than the reflection of a style of problem solving emanating from the particular experiential histories of such children. Work on the outer-directedness of the retarded also appears related to the locus of control work done by Cromwell and his associates (Cromwell 1963). These investigators found that retardates, as compared to normals, manifest an external locus of control, that is, they attribute certain events caused by their own behavior to outside forces over which they have little control. (This internal-control versus external-control dimension has been employed by Cromwell to bring some further order to the inconsistent findings in the success—failure literature.)

The reinforcer hierarchy . Another nonintellective factor important in understanding the behavior of the retarded is the retardate’s motivation under various types of incentives. That performance by normals and retardates on a variety of tasks is influenced by the nature of the incentive is certainly well documented. The social deprivation work discussed earlier in this section indicates that retardates have an extremely high motivation for attention, praise, and encouragement. Several investigators (e.g., Cromwell 1963; Zigler 1963) have suggested that in normal development the effectiveness of attention and praise as reinforcers diminishes with maturity and is replaced by the reinforcement inherent in the information that one is correct. This latter type of reinforcer appears to serve primarily as a cue for the administration of self-reinforcement [seeLearning, article on REINFORCEMENT].

Zigler and his associates (Zigler 1962; Zigler & deLabry 1962; Zigler & Unell 1962) have argued that a variety of experiential factors in the history of the retarded cause them to be less motivated to be correct for the sake of correctness than normals of the same MA. Stated somewhat differently, these investigators have argued that the position of various reinforcers in the reinforcer hierarchies of normal and retarded children of the same MA differ. To date, the experimental work of this group has centered on the reinforcement which inheres in being correct. It is this reinforcer that is the most frequently dispensed, immediate incentive in most real-life tasks. Furthermore, it is a frequently used incentive in many experimental cognitive and perceptual tasks on which retardates and normals are compared, and it also seems to be the most important incentive in the typical test situation. When such an incentive is employed in experimental studies, one wonders how many of the differences found are attributable to differences in capacity between retardates and normals rather than to differences in performance which result from the different values that such incentives might have for the two types of subject.

Clearest support for the view that the retardate is much less motivated to be correct than is the middle-class child, so typically used in comparisons with the retarded, is contained in a study by Zigler and deLabry (1962). These investigators tested middle-class, lower-class, and retarded children equated on MA on a concept-switching task (Kounin 1941) under two conditions of reinforcement. In the first condition, similar to that employed by Kounin, the only reinforcement dispensed was the information that the child was correct. In the second condition, the child was rewarded with a toy of his choice if he switched from one concept to another. In the “correct” condition these investigators found, as Kounin did, that retardates were poorer in their concept switching than were middle-class children. That this was not a simple matter of cognitive rigidity was indicated by the finding that lower-class children equated with the middle-class children on MA were also inferior to the middle-class children. In the toy condition this inferiority disappeared, and retarded and lower-class children performed as well as the middle-class children. This study highlights an assumption that has been noted as erroneous by many educators; namely, that the lower-class child and the retarded child are motivated by the same incentives that motivate the typical middle-class child. An intriguing avenue of further research is the degree to which the position of various reinforcers in the hierarchy can be manipulated.

General effects of institutionalization . No discussion of motivational factors in the performance of the retarded would be complete without some mention of the effects of institutionalization. The institutionalization variable has probably contaminated more research in the area of mental retardation than any other single variable. Given our general lack of knowledge concerning the effects of institutionalization on human behavior, the extent of this contamination cannot be determined. That the effects of institutionalization on the behavior of retardates are considerable has been suggested by several investigators (e.g., McCandless 1964; Windle 1962). In view of the general consensus concerning the importance of institutionalization, it is amazing that more work has not been done to investigate its effects on retarded children.

Some fairly clear findings with retardates have demonstrated that institutionalization causes a decrement in the quality of language behavior (Lyle 1959), reduces the level of abstraction on vocabulary tests (Badt 1958), interferes with the ability to conceptualize an emotional continuum (Iscoe & McCann 1965), and increases the child’s orientation toward punishment (Abel 1941). These studies, though suggestive, have shed little light on the specific aspects of institutionalization which affect such behaviors or on the exact nature of the process through which behaviors are affected. Whether the deficiencies in the behavior of the institutionalized retardate are motivational in nature or reflect an actual change in intellectual capacity is still an open question.

Evidence that institutions for the retarded differ in their effects on behavior has recently been reported by Butterfield and Zigler (1965b). It was found that children residing in a cold, restrictive institution showed a higher motivation for adult support and approval than children residing in an institution having a warm, accepting social climate. These investigators are presently conducting a cross-institutional longitudinal study of six state schools for the retarded in an effort to isolate the institutional factors and psychological processes underlying such effects.

Much of this work on motivational and emotional factors in the performance of the retarded is very recent. The research conducted on several of the factors discussed in this section is more suggestive than definitive. It is clear, however, that these factors are extremely important in determining the retardate’s general level of functioning. Furthermore, these factors seem much more open to environmental manipulation than do the cognitive processes discussed earlier. An increase in knowledge concerning motivational and emotional factors and their ontogenesis and manipulation holds considerable promise for alleviating much of the social ineffectiveness displayed by that sizable group of persons who must function at a relatively low intellectual level.

Edward Zigler

[Directly related are the entriesAchievement TESTING; Intelligence AND INTELLIGENCE TESTING; other relevant material may be found inCreativity, article on GENIUS AND ABILITY; Infancy, article onTHE EFFECTS OF EARLY EXPERIENCE; Systems ANALYSIS, article on PSYCHOLOGICAL SYSTEMS; Thinking;and in the biographies ofBinetandMontessori]

BIBLIOGRAPHY

Abel, Theodora M. 1941 Moral Judgments Among Sub-normals. Journal of Abnormal and Social Psychology 36:378–392.

Badt, Margit 1958 Levels of Abstraction in Vocabulary Definitions of Mentally Retarded School Children. American Journal of Mental Deficiency 63:241-246.

Balla, David; and Zigler, Edward 1964 Discrimination and Switching Learning in Normal, Familial Retarded, and Organic Retarded Children. Journal of Abnormal and Social Psychology 69:664-669

Barrett, Helen E.; and Koch, Helen L. 1930 The Effect of Nursery-school Training Upon the Mental Test Performance of a Group of Orphanage Children. Journal of Genetic Psychology 37:102-122.

Burt, Cyril; and Howard, Margaret 1956 The Multi-factorial Theory of Inheritance and Its Application to Intelligence. British Journal of Statistical Psychology 9:95-130.

Butterfield, Earl C.; and Zigler, Edward 1965a The Effects of Success and Failure on the Discrimination Learning of Normal and Retarded Children. Journal of Abnormal Psychology 70:25-31.

Butterfield, Earl C.; and Zigler, Edward 1965b The Influence of Differing Institutional Social Climates on the Effectiveness of Social Reinforcement in the Men-tally Retarded. American Journal of Mental Deficiency 70:48-56.

Clarke, A. D.; and Clarke, A. M. 1954 Cognitive Changes in the Feebleminded. British Journal of Psychology 45:173-179.

Cromwell, Rue L. 1963 A Social Learning Approach to Mental Retardation. Pages 41-91 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Doll, Eugene E. 1962 A Historical Survey of Research and Management of Mental Retardation in the United States. Pages 21-68 in E. Philip Trapp and Philip Himelstein (editors), Readings on the Exceptional Child. New York: Appleton.

Ellis, Norman R. 1963 The Stimulus Trace and Behavioral Inadequacy. Pages 134-158 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Ferguson, George A. 1956 On Transfer and the Abili-ties of Man. Canadian Journal of Psychology 10:121-131.

Fremming, Kurt H. (1947) 1951 The Expectation of Mental Infirmity in a Sample of the Danish Population: (Based on a Biographical Investigation of 5,500 Persons Born in the Years 1883-1887). London: Eugenics Society. ? First published in Danish.

Gardner, William I. 1958 Reactions of Intellectually Normal and Retarded Boys After Experimentally Induced Failure: A Social Learning Theory Interpretation. Ann Arbor, Mich.: University Microfilms.

Garfield, Sol L. 1963 Abnormal Behavior and Mental Deficiency. Pages 574-601 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Garrison, M. (editor) 1966 Cognitive Models and Development in Mental Retardation. American Journal of Mental Deficiency 70, no. 4 (Monograph Supplement).

Goldstein, Kurt 1943 Concerning Rigidity. Character and Personality 11:209-226.

Green, Calvin; and Zigler, Edward 1962 Social Deprivation and the Performance of Retarded and Normal Children on a Satiation Type Task. Child Development 33:499-508.

Hebb, Donald O. 1949 The Organization of Behavior: A Neuropsychological Theory. New York: Wiley.

Heber, Rick F. 1957 Expectancy and Expectancy Changes in Normal and Mentally Retarded Boys. Ann Arbor, Mich.: University Microfilms.

Heber, Rick F. 1959 A Manual of Terminology and Classification in Mental Retardation. American Journal of Mental Deficiency 64, no. 2 (Monograph Supplement ).

Heber, Rick F. 1962 Mental Retardation: Concept and Classification. Pages 69-81 in E. Philip Trapp and Philip Himelstein (editors), Readings on the Exceptional Child. New York: Appleton.

Hirsch, Jerry 1963 Behavior Genetics and Individuality Understood. Science 142:1436-1442.

Iscoe, Ira; and Mccann, Brian 1965 Perception of an Emotional Continuum by Older and Younger Mental Retardates. Journal of Personality and Social Psychology 1:383-385.

Jervis, George A. 1959 The Mental Deficiencies. Volume 2, pages 1289-1313 in American Handbook of Psychiatry. Edited by Silvano Arieti. New York: Basic Books. ? See especially Table 1 on page 1290.

Jones, E. Caradog; and Carr-saunders, A. M. 1927 The Relation Between Intelligence and Social Status Among Orphan Children. British Journal of Psychology 17:343-364.

Jones, Harold E. (1946) 1954 The Environment andMental Development. Pages 631-696 in Leonard Car-michael (editor), Manual of Child Psychology. New York: Wiley.

Jordan, Thomas E.; and Decharms, Richard 1959 The Achievement Motive in Normal and Mentally Retarded Children. American Journal of Mental Deficiency 64:457-466.

Kaplun, David 1935 The High-grade Moron: A Study of Institutional Admissions Over a Ten Year Period. Journal of Psychoasthenics 40:69-91. ? Now called the American Journal of Mental Deficiency.

Kass, Norman; and Stevenson, Harold W. 1961 The Effect of Pretraining Reinforcement Conditions on Learning by Normal and Retarded Children. American Journal of Mental Deficiency 66:76-80.

Kounin, Jacob S. 1941 Experimental Studies of Rigidity. Character and Personality 9:251-282. ? Part 1: The Measurement of Rigidity in Normal and Feeble-minded Persons. Part 2: The Explanatory Power of the Concept of Rigidity as Applied to Feeble-mindedness.

Kounin, Jacob S. 1948 The Meaning of Rigidity: AReply to Heinz Werner. Psychological Review 55:157-166.

Lantz, Beatrice 1945 Some Dynamic Aspects of Success and Failure. Psychological Monographs 59, no. 1; Serial no. 271.

Lashley, K. S.; Chow, K. L.; and Semmes, Josephine 1951 An Examination of the Electrical Field Theory of Cerebral Integration. Psychological Review 58:123-136.

Laurendeau, Monique; and Pinard, Adrien 1963 Causal Thinking in the Child: A Genetic and Experimental Approach. New York: International Universities Press.

Leahy, Alice M. 1935 Nature-Nurture and Intelligence. Genetic Psychology Monographs 17:236-308.

Lewin, Kurt (1926-1933) 1935 A Dynamic Theory of Personality: Selected Papers. New York: McGraw-Hill.

Luria, A. R. 1963 Psychological Studies of Mental Deficiency in the Soviet Union. Pages 353-387 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Lyle, J. G. 1959 The Effect of an Institutional Environment Upon the Verbal Development in Imbecile Children. 1: Verbal Intelligence. Journal of Mental Deficiency Research 3:122-128.

Mccandless, Boyd R. 1964 Relation of Environmental Factors to Intellectual Functioning. Pages 175-213 in Harvey A. Stevens and Rick F. Heber (editors), Mental Retardation: A Review of Research. Univ. of Chicago Press.

Mcclearn, Gerald E. 1962 The Inheritance of Behavior. Pages 144-252 in Leo Postman (editor), Psychology in the Making. New York: Knopf.

Mckinney, John P.; and Keele, Tina 1963 Effects of Increased Mothering on the Behavior of Severely Retarded Boys. American Journal of Mental Deficiency 67:556-562.

Maher, Brendan A. 1963 Intelligence and Brain Damage. Pages 224-252 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill. Mental Retardation in the Soviet Union. 1964 Canada’s Mental Health Supplement no. 42.

Milgram, Norman A.; and Furth, Hans G. 1963 The Influence of Language on Concept Attainment in Educable Retarded Children. American Journal of Mental Deficiency 67:733-739.

O’Connor, N.; and Hermelin, B. 1959 Discrimination and Reversal Learning in Imbeciles. Journal of Abnormal and Social Psychology 59:409-413.

Penrose, Lionel S. (1949) 1963 The Biology of Mental Defect. 3d ed. London: Sidgwick & Jackson.

Phillips, Leslie; and Zigler, Edward 1964 Role Orientation, the Action-Thought Dimension, and Out-come in Psychiatric Disorder. Journal of Abnormal and Social Psychology 68:381-389.

Piaget, Jean (1932) 1948 The Moral Judgment of the Child. Glencoe, III.: Free Press. ? First published in French.

Rieber, Morton 1964 Verbal Mediation in Normal and Retarded Children. American Journal of Mental Deficiency 68:634-641.

Shallenberger, Patricia; and Zigler, Edward 1961 Rigidity, Negative Reaction Tendencies, and Cosatiation Effects in Normal and Feebleminded Children. Journal of Abnormal and Social Psychology 63:20-26.

Siegel, Paul S.; and Foshee, James G. 1960 Molar Variability in the Mentally Defective. Journal of Abnormal and Social Psychology 61:141-143.

Spitz, Herman H. 1963 Field Theory in Mental Deficiency. Pages 11-40 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Spivack, George 1963 Perceptual Processes. Pages 480-511 in Norman R. Ellis (editor), Handbook of Mental Deficiency: Psychological Theory and Research. New York: McGraw-Hill.

Stevenson, Harold W.; and Fahel, Leila 1961 The Effect of Social Reinforcement on the Performance of Institutionalized and Noninstitutionalized Normal and Feebleminded Children. Journal of Personality 29: 136-147.

Stevenson, Harold W.; and Zigler, Edward 1957 Disscrimination Learning and Rigidity in Normal and Feebleminded Individuals. Journal of Personality 25: 699-711.

Stevenson, Harold W.; and Zigler, Edward 1958 Probability Learning in Children. Journal of Experimental Psychology 56:185-192.

Taylor, Janet A. (1956) 1963 Drive Theory and Manifest Anxiety. Pages 205-222 in Martha T. Mednick and Sarnoff A. Mednick (editors), Research in Personality. New York: Holt. ? First published in the Psychological Bulletin.

Thorpe, Louis P. (1946) 1955 Child Psychology and Development. 2d ed. New York: Ronald Press.

Turnure, James; and Zigler, Edward 1964 Outer-directedness in the Problem Solving of Normal and Retarded Children. Journal of Abnormal and Social Psychology 69:427-436.

U.S. President’s Panel On Mental Retardation 1963 A Proposed Program for National Action to Combat Mental Retardation. Washington: Government Printing Office.

Weaver, Thomas R. 1946 The Incidence of Maladjustment Among Mental Defectives in Military Environment. American Journal of Mental Deficiency 51: 238-246.

Wheeler, Lester R. 1942 A Comparative Study of the Intelligence of East Tennessee Mountain Children. Journal of Educational Psychology 33:321-334.

Whitney, E. Arthur. 1956 Mental Deficiency: 1955. American Journal of Mental Deficiency 60:676-683.

Windle, Charles 1962 Prognosis of Mental Subnormals. American Journal of Mental Deficiency 66, no. 5 (Monograph Supplement).

Zeaman, David; and House, Betty J. 1962 Approach and Avoidance in the Discrimination Learning of Retardates. Child Development 33:355-372.

Zigler, Edward 1961 Social Deprivation and Rigidity in the Performance of Feebleminded Children. Journal of Abnormal and Social Psychology 62:413-421.

Zigler, Edward 1962 Rigidity in the Feebleminded. Pages 141-162 in E. Philip Trapp and Philip Himelstein (editors), Readings on the Exceptional Child. New York: Appleton.

Zigler, Edward 1963 Social Reinforcement, Environmental Conditions and the Child. American Journal of Orthopsychiatry 33:614-623.

Zigler, Edward; and Delabry, Jacques 1962 Concept-switching in Middle-class, Lower-class, and Retarded Children. Journal of Abnormal and Social Psychology 65:267-273.

Zigler, Edward; and Unell, Earl 1962 Concept-switching in Normal and Feebleminded Children as a Function of Reinforcement. American Journal of Mental Deficiency 66:651-657.

Zigler, Edward; and Williams, Joanna 1963 Institutionalization and the Effectiveness of Social Reinforcement: A Three Year Follow-up Study. Journal of Abnormal and Social Psychology 66:197-205.