Keywords

In this book (Jessor, Graves, Hanson, & Jessor, 1968) we have sought to inform the reader about what we have done and how we went about our work. We described the initial problem which we assumed as the explanatory objective, the concepts we used in a theoretical formulation, the stance we took with respect to field research, the measures devised and the studies in which they were employed, and, finally, the results that were obtained. A review of these various aspects should be useful at this point; discussion of problems and issues raised by the research and some of its limitations and implications will follow.

A Brief Overview

The research began with the task of accounting for the differential rates of occurrence of problem behavior, especially heavy alcohol use, among three ethnic groups in a small rural community in southwestern Colorado. Although members of the community were ready with their own “explanations,” and although there were several obvious vantage points from which an investigation might begin, it seemed clear that the situation provided a natural laboratory in which a general theory of deviance-conformity might be developed and put to test. The accomplishment of such an objective required a conceptual analysis of the behavior involved and of the factors, in both the person and the situation, which might be important influences upon it.

Our first concern was to recognize the essential continuity between deviance and conformity and to assume that an explanation focused at either pole must have implications for the other. What this implied was that deviant behavior, like any other, was best treated as learned, purposive, goal oriented, in short, as adaptive action requiring no special principles to account for its occurrence beyond those required for social behavior in general. The central issue became that of explaining not deviance per se but the occurrence of deviant rather than some other, that is, conforming, behavior. Stated otherwise, the problem was to account for selection or choice among possible adaptive alternatives. In this light, the selection of deviant behavioral adaptations, despite the ultimate possibility of negative sanctions, appeared to be more likely when other alternatives had come to be seen by the actor as promising him little in the way of success. The view that it would be useful to interpret deviance as reflecting, at least in part, the failure of conformity was adopted.

Our second concern was to recognize the complexity inherent in the concept of deviance. One source of complexity stems from the diversity of behavior which the concept subsumes. Any number of actions, all significantly departing from normative prescription, can be learned as alternative adaptations when conforming behavior fails to secure personal goals. The recognition of this source of complexity suggested the need to go beyond a concern with heavy alcohol use alone to include other adaptations which might be functionally equivalent. A test of a theory of deviant behavior seemed to us to require an assessment of the class of deviant alternatives rather than a focus upon any particular one. A second source of complexity is that which inheres in all social behavior: the fact that deviance and conformity represent the outcome of multiple influences and determinants in both the person and his situation.

These views about deviant behavior directed our analysis of both personality and the sociocultural environment, an analysis demanding multiple determinants likely to reflect the failure of conformity and the availability of deviant alternatives. For the major personality formulations, we drew upon Rotter’s social learning theory (1954); for the sociocultural concepts, we borrowed from Merton (1957), and Cloward and Ohlin (1961). Complementary, analogous conceptualizations of the person and of the environment, systematically coordinated to each other and to deviance, were consequently developed.

The sociocultural environment was articulated as a system made up of three major structures: the opportunity structure, the normative structure, and the social control structure. Location in the opportunity structure was construed as a source of pressure toward the adoption of illegitimate means, with low access to opportunity implying a relatively high degree of pressure. Locations in both the normative structure and the social control structure were defined as sources of controls against the adoption of illegitimate means, with high anomie in the normative structure and high access to illegitimate means in the social control structure implying a low degree of control. Deviance as a sociocultural outcome was, therefore, considered to be the resultant of sociocultural pressure and controls, neither factor alone yielding a sufficient explanation.

The same conceptual approach was applied to the person. Personality was construed as a system made up of three structures parallel to those constituting the sociocultural system: a perceived opportunity structure, providing a source of pressure toward deviance, and a personal belief structure and a personal control structure, both representing personal controls against engaging in deviance. The resultant of personal pressure and control was seen to determine, now at the personality level, the likelihood of deviant behavior, neither factor alone being sufficient.

Although these two conceptual systems, sociocultural and personality constitute separate theories of deviance, their formal similarity and parallel concepts enabled their assimilation into a single, interdisciplinary, explanatory scheme. This was our basic objective: to construct a field theory of deviant behavior in which the interaction of sociocultural and personality determinants could be dealt with systematically.

To make the field theory a more convincing synthesis, it seemed important to consider how the sociocultural environment comes, over time, to influence the development of personality. To this end, socialization was construed as a system lying at the interface between society and the person. It also was articulated into three structures, ones likely to reflect the sociocultural system and, at the same time, to be relevant to the personality system. The structures of the socialization system, analogues of the structures in the other two systems, were the parental reward structure, the parental belief structure, and the parental control structure.

A test of the adequacy of the explanatory formulation was seen to rest on its ability to yield, simultaneously, an account of differences between ethnic groups in rates of occurrence of deviance and an account of individual differences within the community and within ethnic groups in the occurrence-nonoccurrence of deviance. The logic of the approach was that the same factors used to explain individual behavior could be applied to explaining differences between groups of individuals, in our case, ethnic groups. The implication of this logic for the meaning of the concept of ethnic status will be elaborated later on.

Second to, and influenced by, our concern for theory was our commitment to developing a research methodology appropriate to the testing of theory in field studies. Although field studies are usually seen as part of the context of exploration and discovery, it was our view that they are also appropriate to the context of justification. To fulfill this latter role, to be relevant to the testing of propositions, field studies require design in which consistent efforts to minimize inferential ambiguity are made. For us this meant, beyond the usual concern for standardization and reliable measurement, the logical derivation of measuring procedures from concepts, the development of multiple measures of concepts, and the use of multiple, converging studies, each independent of the others but converging upon the validity of the over-all social-psychological framework. To the extent that such a methodological orientation could be successfully implemented, to that extent, it seemed to us, could theory-testing be compelling in field research.

Three separate and independent studies were carried out in the community. The first of these was a community survey study in which data were collected by individual interview from a random sample of adults between the ages of twenty and sixty-five, stratified by sex and ethnic group. The interview included measures of the sociocultural system, the personality system, and the behavior system, that is, deviance-conformity and alcohol use. Information on the latter was supplemented by an exhaustive search of relevant court records but, in the main, this study provided a self-contained test of the theory based upon self-report interview data. The second study focused upon a younger age group in the community, the students in the local high school. It involved a wide range of procedures, including self-report group questionnaires, interviews, sociometrics, behavior tests, teacher ratings, and school records. Yielding sociocultural, personality, and behavior measures, it enabled a second, independent test of the over-all framework.

The third study dealt with socialization and was an attempt to study the linkage between the sociocultural system, on the one hand, and the personality and behavior systems, on the other. In the socialization study, data were collected by individual interview from the mothers of the students in the high school, and measures based upon these data were used to predict the personality and behavior measures independently obtained from their children in the high school.

All three studies provided some degree of support for the theory guiding the research, and the convergence of the findings from the separate studies constitutes a strong basis for inference. Among the ethnic groups, the Anglos were shown to occupy the most favorable position in the opportunity structure: They have the greatest objective access to valued goals by legitimate means and are, consequently, under the least pressure to adopt alternative, often illegitimate, means. With respect to the normative structure, they were also shown to occupy the most theoretically favorable position: they have the greatest degree of consensus around group norms, that is, the least anomie, and are subject, therefore, to the greatest normative control against the adoption of illegitimate means. Finally, with respect to the social control structure, their position is also the most favorable: they have least access to illegitimate means and are, therefore, subject to the strongest social controls. Taken together, the findings show the Anglos to occupy the point of intersection in sociocultural “space” which should be theoretically least conducive to deviance, the point, relative to the other two ethnic groups, at which pressures toward deviance are lowest and controls against deviance are highest. These findings are consistent with the data showing that the Anglos, among the three ethnic groups, make the least contribution to deviance rates in the community.

Considering the other two ethnic groups, the Spanish and the Indians, the findings are more complex and make clear the utility of a theory which deals simultaneously with both pressures and controls. In terms of pressures toward deviance, the Indians actually have a somewhat more favorable position in the opportunity structure than the Spanish, at least when access to opportunity is defined exclusively in terms of socioeconomic status. With respect to controls, however, the Spanish clearly occupy the theoretically more favorable position. The measure of normative controls suggests that anomie is more pervasive and generalized among the Indians than among the Spanish; and with respect to social controls, the picture is sharpest, the Spanish having least exposure to deviant role models and being mapped into solidary sanctioning networks such as the Catholic Church, the family, and informal groups significantly more than are the Indians. Despite equal or even greater pressures toward deviance, the Spanish are subject to the operation of much stronger and more consistent controls than the Indians. These findings are consonant with the data showing the Indians to contribute most to the deviance rates in the community, with the Spanish intermediate between them and the Anglos. The intermediate position of the Spanish rates is actually much closer to that of the Anglos than it is to that of the Indians, suggesting the possibly more important role played by social controls, relative to pressures, in influencing the occurrence of deviance.

The findings just described support the sociocultural aspect of the theory in dealing with ethnic group differences in deviant behavior. The results bearing on the personality aspect of the theory are also supportive. The Anglos have the greatest perception of opportunity, that is, the highest expectations of achieving goals, or the least personal disjunctions. They are also least alienated with respect to the personal belief structure, and they have the strongest personal controls. The trend with respect to the two minority groups is for the Indians to hold the more deviance-prone position on the personality measures compared to the Spanish, although the relative position of the two groups is not clearly established on all of the measures. With respect to personal control measures, however, as was the case with social controls, the Spanish, despite their low position in the economic opportunity structure, are closer to the Anglos than they are to the Indians.

The support provided for the over-all social-psychological framework by the outcome just described gains reinforcement from the fact that it emerges from two independent studies in the community, one dealing with adults and the other with adolescents of high school age. These two studies, using different age groups, different measures, and different settings, yet generating congruent empirical data, yield the kind of convergence toward which the methodological planning was oriented. Further convergence stems from the socialization study. Ethnic group differences in parental reward structure, parental belief structure, and parental control structure measures follow from the sociocultural position of the family and accord with the personality and behavior measures of the high school children.

Showing that ethnic group differences in deviance rates were consonant with the relative positions of the three ethnic groups on the theoretical variables in the sociocultural and personality systems was a major objective. A second major objective was to provide an account of individual differences by reference to the same theoretical framework. This involved a more direct test of the theory, an assessment of the direct relation of sociocultural and personality measures to the occurrence of deviant behavior.

Each of the measures in the sociocultural and personality systems was related, singly, to each of the multiple criterion measures of deviance and alcohol use. What these analyses showed was that, with some exceptions, the measures related in the direction which the theory implied but that the degree of relationship of each measure was generally small. Given our conceptualization of deviance as the complex outcome of both pressures and controls, low, single-measure correlations were not unexpected. The next step, therefore, was to deal simultaneously with multiple measures, and this was done by a pattern analysis procedure in which each individual was characterized by the “syndrome” of scores he had obtained with respect to measures of both pressures and controls. This pattern analysis procedure, which captures the intent of the theoretical interpretation, was more successful.

Considering the community as a whole, strong linear relations were shown to obtain between the sociocultural syndrome (which included measures of objective opportunity and social controls) and various criteria of deviance. The same was shown to be true for the relations between the personality pattern (which included measures of perceived opportunity, alienation, and personal controls) and deviance and drinking criteria. Most impressive, however, was the “field” pattern, which incorporated both sociocultural and personality measures (objective opportunity, social controls, perceived opportunity, and personal controls) into a single predictor pattern. The relation of the field pattern to the several criteria was shown to account for variance not accounted for by either the sociocultural or the personality patterns alone, and it tended consistently to yield the best prediction of the various deviance criteria in both the Community Survey Study and the High School Study. Treating the community adults as a whole and the high school students as a whole, the multivariable pattern analyses provided compelling evidence in support of the theoretical framework.

The final step in the direct assessment of the theory was to examine it, as above, within each of the ethnic groups. The pattern analyses retained the predictiveness they had shown for the community as a whole when they were applied within both the Anglo and Spanish groups, and for both sexes, but they were strongly attenuated in their ability to predict deviance within the Indian group. This attenuation seemed to be due in part to the high deviance rate characterizing the Indians and making differential within-group prediction extremely difficult. More will be said on this point in the discussion of limitations of the research.

To conclude this overview, a further point needs to be made. The measurement of deviance and deviance-prone behavior, such as heavy alcohol use, proved to be a task of great complexity, but one clearly meriting the attention it received. By retaining separate measures of various aspects of deviance and drinking behavior, we were able not only to assess the interrelations among them but also to use them as multiple, separate criteria in theory-testing. Further, by constructing a global index of deviance which combined various separate measures, we were able to approach most closely the kind of criterion the theory was directed at. This global deviance criterion was best and most consistently predicted by the sociocultural and personality measures.

In summary, these three studies in the Tri-Ethnic Research Project, the Community Survey Study of adults, the High School Study of adolescents, and the Socialization Study, yielded convergent, empirical support for the theoretical framework as an explanation of both group and individual differences in deviance and alcohol use. Such convergence suggests that inferences about the nature of deviance in this community, inferences of the sort contained in the theory, can be drawn with some degree of security.

Limitations of the Research

The preceding overview has focused upon the larger pattern of our findings. There were, however, a number of important limitations in our work stemming from the nature of the situation in which it was done, from our approach, and from our findings; these merit at least brief discussion at this point.

Perhaps the most salient question has to do with the character of the research context—a small, rural, southwestern community—and the constraints which it imposes upon generalizing from the findings. As with the study of any community, its representativeness of some universe is always in doubt, especially when the community has been selected, as in our case, for its particular ethnic composition. While there are obviously many other communities like the one in which we worked, it is difficult, on any sampling basis, to lay claim to inferences which go beyond its boundaries. This means, in the most severe terms, that our explanation of deviance applies only to this community, or perhaps to others which are demonstrably similar, and greater generalizability must wait upon extension and replication.

An aspect of this limitation which should be emphasized, and which illustrates the point, is that deviance in the research community cannot be said to be institutionalized in any formal sense. While there are informal peer groups, there do not appear to be gangs of the sort that characterize large urban centers, nor the formalized criminal organizations among adults frequently found in cities. Although the measures of deviance and deviance-prone behavior were comprehensive, they dealt with behavior which is most accurately described as only informally structured. The applicability of the findings is, therefore, in question where deviance can be shown to be a relatively institutionalized, formally supported and rewarded pattern of behavior.

A second major limitation has to do with the fact that the entire theory was not available at the outset of the research, but was in part developed during the process of investigation. This is most true of the social control formulations which, although considered from the beginning, were not specified in sufficient detail to guide the initial data collection. As the importance of social controls began to emerge, data already in hand were used to measure those concepts. While this was generally a feasible thing to do, it was obviously only poorly accomplished in certain respects—for example, in the measurement of opportunity to engage in deviance. Despite the support for the final social control formulations, support which obtained in the separate studies, their partially post hoc nature leaves this portion of the theory somewhat less securely established.

A third shortcoming is that certain of our measures simply did not work out. A primary example was the predictive failure of the internal-external control measure. A great deal of effort had gone into the development of this procedure, since the concept of internal-external control seemed theoretically important in linking personality with deviance, and since it was the kind of concept which could be readily coordinated with sociocultural variation. Its failure remains unclear to us, especially since it failed in both the High School and the Community Survey studies to show the expected relationships. These negative findings are not in accord with the success which the same or very similar measures have had in other studies (Rotter, 1966; Seeman, 1963; Wood, Wilson, Jessor, & Bogan, 1966), or with the importance which social-psychological analysis would assign to the concept. Further work is called for, and our data in this regard are disconcerting.

An additional shortcoming which we regret very much is our failure to explore in greater depth the role of peer groups and the impact of peer socialization . As informal influences conducing to or constraining against deviance, it is obvious they exert a strong influence. In the High School Study, sociometric data could have been employed to this end rather than serving simply as a criterion measure of deviance. In the Socialization Study, inquiry about peer associations could have been profitably undertaken. In both cases, the peer group network in which each youth is embedded would have been available for analysis, and the relation of peer support to the occurrence of deviance might have been better understood.

The inability of our approach to predict the particular form of deviant adaptation engaged in is a further limitation of note. The importance of measuring access to illegitimate means, especially the exposure to deviant role models component, was argued as the direction to take in coping with precisely this problem: with why, for example, one person may adapt to failure and frustration by mental illness, another by narcotics use, and a third by crime or heavy alcohol use. What would seem to be required is an analysis of exposure to various, alternative, deviant adaptations, an appraisal of the possibility of learning them through modeling, and an assessment of the conditions which may endow them differentially with the likelihood of success. In our own work, this type of detailed analysis was not made. Instead, we relied on a crude measure of exposure to any form of deviance, and this limitation precluded a contribution to the understanding of the selection of specific forms of deviance. Our approach, dealing with the class of deviant behaviors, was relevant to our concern with testing a general theory, but it meant that an important problem was not confronted.

The fact that our predictor measures in both the sociocultural and personality systems, when dealt with singly, generally had only low or moderate relationships to the criterion measures should also be commented on at this point. How much this reflects inadequate development of measures is difficult to say. Despite a tremendous expenditure of resources on the construction and refinement of measures, it is clear to us that we could have benefited from even more effort in this respect.

The recognition that, despite our efforts, the measures remained relatively crude, supported the decision to dichotomize the measures in devising the pattern analysis procedure. Although that decision seems to discard the possibility of greater discrimination, it does acknowledge the crudeness and attempts, by dichotomizing, to assure that variation on each predictor is securely established. It was the latter which was of primary concern to us in that our aim at this stage of theory development was to establish parameter relevance in prediction rather than to estimate parameter values.

A further shortcoming relates to the fact that the empirical findings do not clarify the issue of the conceptual unity of the separate theoretical structures posited within both the sociocultural and the personality systems. Alternative measures within each structure sometimes related better to measures in other structures than they did to each other; also, structures within one system sometimes related to structures in the other system which were not their conceptual analogue better than to the one that was. These data are not clear-cut: they support both a generality point of view and a specificity point of view about the functional unities implied by the separate structures. This issue will require further empirical and conceptual analysis. Empirically, there is a need to develop minimally overlapping measures which represent most precisely the conceptual content of each particular structure. Conceptually, the issue turns on the degree to which the structures within a system can be argued to be theoretically uncorrelated. The latter situation would be difficult to defend, either for the sociocultural system or the personality system, since the very notion of system implies a correlation among structures; and this is generally what we found. Whether relations within systems can obtain without jeopardizing the specificity of relations between analogous structures in different systems will only be known when further data are in hand. In the meantime, the theoretical structures postulated for the environment and for the person have thus far been of major heuristic value.

A final limitation to which attention must be called was the relatively poor prediction of deviance within the Indian group. While the pattern analyses showed effectiveness in accounting for variation in deviant behavior within both the Anglo and Spanish groups, only directional trends were obtained for the Indians. There are a number of possible explanations for this. It is possible, for example, that the interview procedures (and the reliance on non-Indian interviewers) were less appropriate to the Indian respondents than to the others in obtaining information on values, expectations, beliefs, and attitudes.

It is also possible that the measures were insufficiently sensitive to important factors differentially operative in the different ethnic groups. For example, while measures of values showed no ethnic group differences of any magnitude, it could be the case that measures focused upon other aspects of the orientation of minority groups to the dominant group could have revealed important differences. One such aspect, of obvious concern to the anthropologist, is acculturation. This concept did not enter directly into our theoretical framework; yet, a secondary analysis of some of our data interpreted in terms of the process of acculturation does seem to enhance prediction within both minority groups (see Graves, 1967). The analysis is post hoc, and the measures of acculturation are possibly tenuous, but the results attained are consistent and coherent, and they certainly suggest that the concept of acculturation may have utility in a community such as the one studied here.

One apparent problem in the attempt to account for differences in deviance within the Indian group is the fact of the high rate of Indian deviance, which makes discrimination an extremely difficult task. Given the general clustering of the Indian group at the upper end of the deviance criteria and at the deviance-prone end of the predictors, better prediction of individual differences within the Indian group would require predictors and criterion measures of much greater sensitivity. It is possible, too, that such a high rate of deviance, with the attendant patterning of learning and modeling which it implies, tends to make most of our predictors relatively unimportant determinants of deviant behavior. This difficulty in prediction within the Indian group should not, however, obscure the fact that, at the group level, an association between sociocultural and personality deviance proneness, on the one hand, and rates of deviant behavior, on the other, has been shown to exist.

The shortcomings of the research which have been noted do not constitute an exhaustive list. They are meant to represent some of the kinds of limitations which attenuate the strength of the findings and to alert the reader to the necessary caution in considering their implications, a few of which can be considered now.

Some Conclusions and Implications

Not too long ago, Henry Murray expressed the opinion that “no theoretical system constructed on the psychological level will be adequate until it has been embraced by and intermeshed with a cultural-sociological system” (1959, p. 20). At the most general level, this view was a starting point for us, and the outcome of our work impels us to return to it as a conclusion. The explanatory usefulness of a field theory of behavior has been shown to extend beyond the limits of its psychological and sociocultural component systems. When the latter are “intermeshed,” certain previously recalcitrant problems become more docile, problems such as why everyone at the same social location does not behave the same way, or why the epidemiology of certain behaviors is patterned in a particular way in the social structure. Beyond these practical consequences, however, there is the fundamental gain of being able, through reliance upon a field theoretical system, to generate more detailed intelligibility about social behavior.

To make this point about the over-all theory is not to diminish the importance of the particular concepts within the embracing systems. Conclusions about certain of the concepts we have used can, as a matter of fact, be drawn with a fair degree of confidence. It is clear, for example, that the notion of expectation constitutes one of our most powerful concepts for describing persons with respect to deviance proneness. Differences in values were relatively minor, whether those values were defined in the common language referring to success or whether they were defined in motivationally relevant terms. What emerged as crucially important were differences in expectation for achieving what was valued. That expectations play a central role in the selective course of human behavior seems clear from the data. This conclusion has greater impact when it is realized that the measures of expectation were conceptually remote from the behavior at issue; unlike another of the psychological measures, tolerance of deviance, which also turned out to be an important predictor, but which dealt directly with deviance, the expectations measures never implicated deviance in the actual measurement procedure.

The conclusion about the important role of expectations in social behavior is in accord with the literature and would seem to have pervasive implications for efforts to deal with problem behavior. Social intervention focused upon raising expectations that socially desirable behavior can lead successfully to valued goals would be a tactic consonant with those implications. That such expectations follow, in part at least, from the position occupied in the opportunity structure is suggested by the research and makes the latter a prime target for concentration in organizing efforts at remediation.

To move to a consideration of the concepts employed to describe the sociocultural environment, certainly social controls have emerged as central. The critical part played by social controls was nowhere clearer than in the differentiation it yielded between the two minority ethnic groups. Both groups are subjected to strong pressures toward deviance, yet the Spanish, embedded in a persisting structure of religious, family, and interpersonal sanctions, contribute far less to the deviance rates than the Indians, for whom the control structure is fragmented or weak. The meaning of controls in our research is in large part the degree to which a person is mapped into solidary groups that reward conforming behavior, punish departures from group norms, and provide relatively few models for deviance. Other things equal, the strengthening of family and both formal and informal group ties would seem to be relevant to efforts toward reducing deviance.

The latter point is probably also relevant to increasing the degree of normative consensus or agreement about what is appropriate in the way of behavior. The research has demonstrated the possibility of relatively direct measurement of anomie and, although the data derive, unfortunately, from only one study, suggests that such lack of normative consensus may be relevant to deviance. The strengthening of social ties, and the corollary development of interpersonal communication, may be as important in reducing anomie as in strengthening social controls.

The analyses of deviance have been instructive in themselves. The use of multiple raters (for example, teachers, or peers) and multiple sources of data (for example, self-reports, records, teacher ratings, or peer nominations) enable increased reliability and convergent validity to overcome the possible shortcomings of self-reports. It should be noted, in regard to the latter, that our own experience justifies reliance upon self-report data, especially where the concern is with rank-ordering a group of subjects on degree of deviance. The convergence, in our research, between self-report data and external criteria of deviance was impressive.

The findings about alcohol use make the complexity of deviance most evident. Comprehensive understanding of alcohol use required knowledge of how it was learned, the context of its use, the amount drunk, the meaning or psychological functions of drinking, and the consequences of its use. Two persons drinking moderate amounts of alcohol may be doing so in very different ways: one by himself, as a way of overcoming a feeling of being nervous or tense; and the other in a group, as a way of expressing his feeling of community with his companions. The difference between these two patterns is not only likely to have different consequences but to be differentially related to sociocultural and personality pressures and controls as well.

What the research has indicated is that at least several aspects of alcohol use are explicable in terms of the sociocultural and personality concepts in the over-all scheme. This was most apparent where the alcohol use was heavy and for personal-effects, problem-solving reasons (although the scheme did show predictiveness, even for the amount of alcohol drunk). The demonstration that excessive alcohol use is related to differential pressures and controls, both sociocultural and personality, has important implications, not only for understanding this form of problem behavior but also for possible remedial measures.

A final implication of the research requiring mention has to do with the meaning of the concept of ethnicity. Our work has contributed to a social-psychology of ethnicity or ethnic status. Instead of dealing with ethnic status in terms of some set of unique traits or in terms of a peculiar cultural legacy distinctive of a particular group, we have dealt with it as representing a position in social-psychological space. A unitary set of sociocultural, personality, and behavior measures has been applied to all three ethnic groups. The result of this approach was an analysis of ethnic status in the language of the theoretical scheme. Thus, what it means to be an Indian in this community is to have limited opportunity, to be confronted by relatively pervasive anomie, to be subject to weak social controls, and the like. Ethnic status, then, is merely a descriptive term, but one probabilistically implying a bundle of theoretical attributes.

This interpretation of the concept of ethnicity is analogous to what Oscar Lewis has urged with respect to another descriptive concept: poverty. In developing the notion of “the culture of poverty,” Lewis has gathered together a number of traits often considered to represent distinctive characteristics of ethnic, national, or regional groups and has argued that these characteristics are:

“…both an adaptation and a reaction of the poor to their marginal position in a class-stratified, highly individuated, capitalistic society. It represents an effort to cope with feelings of hopelessness and despair that arise from the realization by the members of the marginal communities in these societies of the improbability of their achieving success in terms of the prevailing values and goals.” (1966, p. 21).

Lewis’ distinction between poverty and the culture of poverty is analogous to the distinction we are suggesting between ethnic status and the social-psychology of ethnic status. It happens that poverty and ethnic or marginal status often go together; that is probably what accounts for the fact that the content of Lewis’ culture of poverty and the content of our social-psychology of ethnic status have so much in common.

The implications of this discussion seem important. They suggest that the deviance rates of the three ethnic groups characterize them not by virtue of their ethnic status, but largely by virtue of their social-psychological status; place Anglos in the situation of Indians, and deviance rates should increase markedly. Such a view departs sharply from that part of the community psychology which, for example, considers deviance and drunkenness an inherently Indian trait. Further, this view would seem important in any considerations of change; insofar as the problem is seen not as a problem of ethnicity but as a problem of the attributes associated with it, the latter become the obvious target of change efforts.

Finally, this perspective provides a rationale for considering, as a single unit or as a whole, communities which are made up of different ethnic groups. Precisely this has been done in some of the major analyses presented here. Although the groups are descriptively different, the rationale suggests that they can be treated homogeneously in terms of their position on a set of variables applicable to all members of the community, variables which, in essence, summarize much of the social and psychological meaning of ethnic status.

A final word: Our work has captured, obviously, neither the quality of daily life nor the succession of events which pattern the course of time in the community; that task belongs to the sensitive ethnographer. Instead, the path we chose to follow was an abstract one, and it is now possible to see where it has led. The ideas developed and the data generated have given us a beginning sense of understanding. Hopefully, they may have application to other social problems as well and, ultimately, may contribute in a small way to the amelioration of the human condition. It would be difficult, at this stage, to ask more of behavioral science.