Behavior Analysis as a Framework for a Multidisciplinary Approach to Social Change

vision of a just society based on personal values--values either acknowledged as subjective or asserted to be empirically derived. What is missing? Why can't effective programs be designed-by us or others? As an applied behavior analyst who teaches and supervises clinical graduate students, I am often confronted with a problem which seems to me to have the same basic form: Students wonder why a particular client fails to change, why theoretically sound interventions fail to alter behavior and attain goals . My typical response emphasizes a basic tenet of behavior analysis: the client, like the pigeon, is always right-he or she is not 'resisting' . I ask students to recall the elaborate functional analysis described by Kanfer and Saslow (1969) which conceptualized all of the variables required for an accurate assessment into seven basic data categories . I stress that if the intervention is not succeeding, then the therapist has not performed an adequate functional analysis, and must assume responsibility for the lack of change. The functional analysis must serve three general purposes, though students sometimes forget the third one. First, it must clearly identify and operationalize the relevant target behaviors and goals . Second, it must clearly identify the relevant antecedent and consequent controlling variables. And third , it must identify potentials and limitations of the current environment as they influence, first , the extent of change possible, that is the goal, and second, the intervention methods available for use (see Kanfer & Saslow's (1969] emphasis on an analysis of the physical, cultural, and social environment). Once these potentials and limitations are incorporated in the analysis, often through the utilization of data from other disciplines (e.g., sociology, Kanfer & Grimm, 1980), a treatment plan which has a reasonable likelihood of success can be developed. · It appears that an analogous functional analysis is necessary when we attempt to change social systems , institutions, and communities, and that our ability to design effective interventions is dependent upon an accurate analysis . Our functional analyses on the societal level seem to be strongest in identifying the specific antecedent and consequent events maintaining the status quo, but woefully weak in dealing with the identification of operationalized, specific target goals and the clarification of potentials and limitations of intervention . Unfortunately, these inadequacies in our analyses of systems will probably not be remediated through increased knowledge of behaviorism or behavior analysis . Knowledge acquired through a comprehensive multidisciplinary approach is neededknowledge drawn from the humanities, social sciences, natural sciences, and their applied technological offshoots . This assertion is not particularly startling or innovative-sociologists (e.g ., Kunkel, 1975), behavioral community psychologists (e.g., Fawcett, Mathews, & Fletcher, 1980), and even radical behaviorists (e.g., Michael, 1977) have all acknowledged the contributions other disciplines can make to a comprehensive behavior analysis. However, I would like to keep this multidisciplinary focus within the functional analysis framework-the one most familiar to applied behavior analysts-because it is a powerful way to organize and use the data. What it, the functional analysis, doesn't do in and of itself is provide the necessary data . An examination of the critical process of identifying specific, operationalized target goals may be facilitated by discussing the concrete processes involved in clinical intervention . When working with an individual client, the behavior therapist or behavior analyst selects a goal only after detailed consideration of his or her own values, those of the client and /or client advocate, those of the social institutions or systems involved, and any empirical data bearing on the ultimate outcomes of various alternative target goals (Goldfried & Davison, 1976; Krasner & Ullmann, 1973). The final selection will often be determined by the various value considerations, not the empirical ones, for there rarely are reliable data confirming the superiority of one goal over another, for this type of individual, in this particular context, at this point in time (Krasner & Ullmann , 1973). On other occasions, empirical determinants exist bu tare inadequate to influence the choice of goals. For example Davison (1976; 1978) has decided not to offer empirically validated behavior therapy, designed to alter sexual orientation, to male homosexuals requesting such intervention . Why? As he says, it is not a question of "Can we do it?", for we can, but rather, "Ought we do it?". And he believes the answer is "No"-on the basis of admittedly personal values, not empirical ones. He asserts that most gays request such therapy because of oppressive social pressure, not because of dissatisfaction with a homosexual object preference . Davison further argues that the very availability of therapy for that so-called problem reinforces th e labeling of it as a problem by both gays and straights. So he advocates that the only therapy offered to gays be one that teaches the counter*Revision of a paper presented at the annual meeting of the Association for Behavior Analysis, Milwaukee, Wisconsin, May 1981. Requests for reprints should be sent to: Richard F. Rakos, Department of Psychology, Cleveland State University, Cleveland, Ohio 44115. I would like to thank Kennee Beth Switzer for providing many valuable contributions to the arguments presented in this article. control skills, such as assertion skills, necessary to confront the oppressive society-to educate it, and to change it into a better, less prejudiced one. However, he has no data to support this decision . We do not know that homosexuals will be happier as social change agents. We do not know that they will be effective as social change a gen ts. We do not even know that society will be better off-however that is defined-should they succeed as social change agents. We only know that they can learn the counter-control skills (e.g., McKinlay, Kelly, & Patterson, 1978; Russel & Winkler, 1977). Davison's (1976; 1978) position is a provocative, controversial one (e.g ., Halleck, 1976; Sturgis & Adams, 1978) that dramatically illuminates each of the three sources of values inherent in any planned behavior change: the change agent's, the target population's, and the affected society's. There appears to be little difference between his position and that of a confirmed Marxist or Radical Skinnerian with a set view of what the ultimate world should look like . A truly classless , collectivist society, or an "Operant Society", are both exciting visions, but they are based on personal values rather than the empirical scientific data asserted by their respective proponents. If, for example, we make the unsubstantiated assumption that most people would agree with Skinner (e.g., 1971) that cultural survival is our primary or ultimate goal (and not everyone is ready to, e.g., Edelstein, 1977), we will still not agree on what type of culture is to survive-what specific behaviors and secondary reinforcers should be taught in that culture (Roger, 1977). Virtually all intellectuals except some radical behaviorists and some philosophers agree that empirical data cannot answer questions which are fundamentally ones involving values. Arguments in this vein include those that assert that we cannot isolate the values inherent in the selection of the goals, methodology, and practical applications of scientific investigation from the "basic science" itself (Bressler, 1972; Krasner & Ullmann, 1973) to the view that scientific facts are only transient conceptualilzations of the world (Gergen, 1973; 1977; 1978), to the assertion that the facts of social science are products of cultural conditioning which impact upon the society they describe and thereby change it (Gergen, 1973), or are culture-specific (Gergen, 1978), and therefore lack transcultural, "transtemporal" validity except perhaps for statements of the most abstract principles (Gergen, 1977). In other words, social science is really social history. If these views are correct, and empiricism alone cannot determine values and hence goals, we need to study other disciplines for guiding stimuli. Our predominant values must be challenged, tempered, and strengthened by the wisdom and knowledge from these additional sources of data. Philosophy, political science, history, anthropology, religion, medicine, economics, art, literature, sociology, and experimental psychology may all have much to offer in answering the questions before us: what available data (descriptive or experimental), what alternate conceptualizations, what differing value systems shed light on the target problem? What are the potential target goals? What intermediate goals are possible and desirable? Of course, these other disciplines are unlikely to use behavioral language (Michael, 1977), but that does not mean that behaviorists cannot find a variety of ways to specify and operationalize goals and values distilled in global terms by others. For example, Sarason (1978) believes that a careful study of social history would sensitize social activists to the existence of three transcultural probA . MULTIDISCIPLINARY APPROACH I Richard F. Rakos t 13 !ems that humans have always attempted to deal with: the individual's sense of aloneness; the individual's attempt to form viable communities; and the individual's attempt to justify living with the knowledge of death. He asserts that social history also demonstrates that each society defines and deals with these differently, that these problems are not created by humans nor will they ever be eliminated, and that any social change effort that ignores the history and dynamics of these three problems will ultimately exacerbate them. Goldenberg (1978), from a similar-perspective in an excellent short book entitled Oppression and Social Intervention, articulates three experiences which humans seem to universally strive for; a sense of self-identify, connected to one's past group heritage; a sense of personal impact, directed toward present behavioral freedom; and a sense of transcendence, in which one goes beyond immediate experiences to understand the human experience, its meaning, and its diversity. While these experiences are not described in rigorous behavioral terms, they can nevertheless be operationalized. They are manifested as overt and convert behaviors, and when problematic, can be quantified as behavioral excesses and deficits. Their "truth", or validity, and perhaps most important, utility, is a separate issue, but we might, for example, eventually understand the failure of various Marxist, Leninist, Trotskyist, and Maoist socioeconomic theories and revolutions to reflect a lack of appreciation of what reinforcers--other than material and economic ones-potently affect the behavior of human beings. The People's Republic of China and Hungary provide two examples of countries which modified orthodox political theory (the unacceptability of private ownership) in the face of economic and social realities. This leads to my next argument: though the values of the interventionist must be very influential in the selection of goals for social change, the alternatives cannot be generated independently of the larger social context with which one is dealing . And, obviously, the larger the social context, the greater the heterogeneity of and conflict between members-which leads to decreased consensus regarding possible goals and decreased control over relevant controlling variables. As with the functional analysis for individuals outlined by Kanfer and Saslow (1969), the social activist must understand this larger social context from several perspectives independently of his or her own values. First, there must be an analysis of the interrelationship of various institutions, agencies, governing bodies, etc. Second, the potentials and limitations inherent in the current sociocultural-physical environment must be identified. And third, the analysis must isolate more than the stimuli that are maintaining the problem behavior; it must identify the various reinforcing and aversive stimuli functioning in the social context which affect the individual's behavior in general. This is one important aspect of what we call "culture". A reinforcer analysis on a societal level may deal with specific tangible and social stimuli. But more profitable may be an analysis of values, which are, essentially, conditioned reinforcers (Krasner & Ullmann, 1973; Skinner, 1971). These will vary among societies, subcultures, geographic regions, racial and ethnic groups, etc. But they must all be understood completely, for as reinforcing and aversive stimuli, they do control the behavior of individuals acting both within and outside of social institutions. One major controlling stimulus is what anthropologists and historians call "Weltanschauung" the world view-