Book Review INFORMAL LOGIC XV.3, Fall 1993 The Person and the Situation: Essential Contributions of Social Psychology by R.E. Nisbett and L. Ross JOHN D. GREENWOOD City College, CUNY Nisbett, Richard E. and Ross, Lee (1991). The Person and the Situation: Essential Contributions of Social Psychology. Phila- delphia, PA: Temple University Press. Pp. xvi + 286. ISBN 0-87722-851-5. US $34.95. In this work Nisbett and Ross aim to provide a broad but selective overview of social psychology, focusing on topics that they hold to be "general, cumulative and important" -critical to any understanding of the potential and limitations of theoreti- cal and applied social psychology. They aim to provide an answer to the serious student who asks "What have we really learned from social psychology?" (p. xv). The text is basically organized around three themes, which are identified by Nis- bett and Ross as the most significant CQn- tributions of social psychology to date: the recognition of the power and subtlety of situational influences on behavior; the rec- ognition that situational influences are me- diated by subjective construals of the situa- tion; and the recognition that individuals and social groups form 'tension systems'. The authors successfully communicate their enthusiasm for their subject and "re- newed pride" in their field. One of the unu- sual and exciting things about this work is that it argues for a broader conception of social psychology-notably in the Intro- duction and Chapter 6 on the Social Psy- chology of Culture-a conception that views social psychology as internally linked to other disciplines such as sociolo- gy and anthropology. Thus the authors look forward to the emergence of disci- plines such as "cultural psychology" and "cognitive anthropology" (0' Andrade, 1981); Stigler, Shweder, & Herdt, 1990), and backwards to the "golden age" of so- cial psychology in the holistic or relational tradition of Kurt Lewin. Unfortunately too much of this gets lost in the central Chapters 2-5, which seem exclusively concerned with the intel- lectual demolition of dispositional theories of personality and behavior-as if this were the only significant achievement of contemporary social psychology. Chapter 2 samples some of the classic studies that demonstrate the power and subtlety of situational influences on behav- ior, and the failure of lay persons to recog- nize these influences: lay persons steadfastly maintain explanations of their own and others' behavior in terms of stable personality dispositions. Thus, for exam- ple, although many laboratory and field studies of 'bystander apathy' (Latane and Darley, 1970) have demonstrated that in- tervention in an apparent emergency varies as a function of the number of other by- standers present (more persons intervene and intervene more quickly when there are no other or fewer bystanders present), most individuals avow explanations of their own and others' behavior in such situations in terms of personality dispositions, even 232 John D. Greenwood when informed of the outcome of the ex- periments. Similar points apply to the other classical studies reviewed with enthusiasm by Nisbett and Ross, such as Asch's stud- ies of conformity, Sherif's studies of inter- group competition and conflict, and Milgram's studies of destructive obedience. In Chapter 2 they argue that situations do not autonomously determine behavior, but only via subjective construals of the situation. Thus, for example, bystanders faced with an emergency do not respond in puppet-like fashion to the number of other bystanders present. Rather the number of other bystanders present influences their behavior via their representation of the sit- uation, by decreasing the likelihood that individuals will represent the situation as a genuine emergency and their personal re- sponsibility. In this chapter the authors de- tail the explanatory potential of references to differences in subjective construal, dis- associating themselves from behaviorists, who also emphasized situational factors but denied the legitimacy of references to subjective psychological factors. Nisbett and Ross also use this chapter to highlight one form of subject construal that is liable to a whole host of biases and errors, namely the so-called 'fundamental attribution error': the persistent tendency of persons to overestimate the influence of dispositional personality factors and un- derestimate the influence of situational factors in the explanation of their own and others' behavior (as in the case of lay ex- planations of 'bystander apathy', for ex- ample). Consequently, they insist that the recognition that situations are only causally potent via subjective construals does not undermine their claim about the causal po- tency of situations, or provide any support for dispositionalist accounts of behavior. On the contrary, Nisbett and Ross claim that individuals regularly commit the funda- mental attribution error by mistakenly sup- po:iing that subjective differences in con- struals are themselves a product of disposi- tional as opposed to situational factors. Chapters 4 and 5 engage the disposi- tional thesis: that consistencies in the be- havior of the same individuals in different situations, and differences in the behavior of different individuals in the same situa- tion, are largely determined by, and thus can be predicted by reference to, stable personality dispositions or traits such as impUlsiveness, honesty, conscientiousness, dependency, and the like, that are differen- tially distributed among individuals. In Chapter 4 they document the really quite impressive empirical evidence against cross-situational consistency: against the view that persons act in the same way in different situations because of stable dispositions. Few correlations be- tween individual pairs of behavioral meas- ures or between personality scale scores and individual behavioral measures gener- ate coefficients higher than .30-and usu- ally they are much lower. In Chapter 5 they describe studies that suggest that layper- sons do hold beliefs about the consistency and predictability of behavior that are grounded in dispositional theories, and document the variety of forms of bias, dra- matic overconfidence, and other attribu- tional and predictive errors to which the layperson appears to be prone. However, and much to their credit, Nisbett and Ross are not satisfied with this efficient hatchet job on lay personality theories couched in terms of dispositions, and are not glibly dismissive of common sense or 'folk' theories of behavior. On the contrary, one of the major virtues of this work is that it makes a genuine attempt to grasp the nettle of a paradox familiar to most consumers of social psychological research, be they psychologists, philoso- phers, or laypersons. Nisbett and Ross rec- ognize that these conclusions based upon experimental studies conflict with our common sense and common experience, including their own. Although many em- pirical studies appear to demonstrate that we are regularly inaccurate and unjustifia- bly overconfident with respect to our explanation and prediction of the behavior of others, this is intuitively hard to accept, since most of us seem to get on pretty well most of the time in forms of social interac- tion that are based upon our understanding and anticipation of the actions of others. Nisbett and Ross refuse to dismiss their own everyday experience as simply wrong (p. 145), and make a genuine and poten- tially fruitful attempt to resolve this para- dox, in a way that they hope enables scientific social psychology not only to "challenge" and "reform" common sense, but also to "expand" it. Their solution to the paradox is to ar- gue that the real degree of stability and predictability of human behavior is not a reflection or product of stable underlying personality traits, and is not in fact depend- ent upon cross-situational consistency: "real world behavioral consistency need not be a reflection of personality traits" and "predictability need not depend on cross situational consistency" (p. 147). Rather the limited consistency and predict- ability of human behavior is based upon social situational factors, such as social role and social relational demands and ex- pectations' the presence of a social audi- ence, and social commitments. Much of the consistency of human behavior is a product of the characteristic demands of social roles, and many differences in behavior are a product of the different demands of different social roles (p. 147). This explains why there is little cross- situational consistency, since individuals may fulfil different social roles in different social situations, or be bound by different social relations or commitments in differ- ent situations. The reason why traditional experimental studies are blind to this real degree of consistency and predictability is because they focus on behavioral consist- ency independently of role-demands, which are either held constant or allowed to vary randomly in experimental studies, and are rarely treated as independent varia- bles whose effects are compared and Review of R. E. Nisbett and L. Ross 233 contrasted. Such studies demonstrate the myth of naive dispositionalism and cross- situational consistency, but obscure the very real degree of stability and predicta- bility based upon social situational and social role factors (p. 147). In Chapter 6 they develop this thesis by suggesting that many constancies and dif- ferences in behavior can be explained by reference to cultural factors, in particular by reference to culturally mediated forms of construal that are tied to different social roles and forms of cultural collectivity. This 'solution' does go some way to resolving the apparent conflict between the results of experimental studies and our common sense and experience. However, by characterizing social roles as 'situation- al' factors, and by rhetorically narrowing the focus of the debate to 'situational' versus 'dispositional' accounts of behav- ior, their presentation of this 'solution' is seriously misleading. Dispositionalism does not appear to get a fair shake. All theories that cite per- sonality characteristics or dispositions are treated as variants of stable trait theories that posit cross-situational behavioral con- sistency. Despite their emphasis on social and cultural factors, the authors seem to studiously ignore social presentational the- ories of personality, which suggest that be- havioral consistency and predictability are grounded in dispositions that relate to the self-presentational demands of different roles located within different social collec- tives (Harre, Clarke, & DeCarlo, 1985), as in Goffman's (1959, 1961) analyses of 'impression management' within the 'mor- al careers' of persons in the social roles of doctors, schoolchildren, hospital patients, and soldiers. The account offered by Nis- bett and Ross differs from such accounts only insofar as Nisbett and Ross call their account 'situational'. Yet social roles are rather more like dispositions than situa- tional factors: for most people most of the time, they are stable and enduring. and ap- pear to be the locus of evaluations of 234 John D, Greenwood behavior by actor and observer, And although social roles per se are not of course features of persons, it is only by vir- tue of a person's acceptance of, commit- ment to, or acquiescence in a social role that it plays a causal role with respect to his or her behavior, One way of making this point is that even if one insisted on calling such stable and enduring dispositional factors 'situa- tional', it is quite clear that they are signifi- cantly different from the types of situational factors detailed in Chapter 2, and which are presented in Chapters 2 to 5 as the ex- planatory alternative to dispositional fac- tors, For one of the salient features of the situational factors documented in Chapter 2 is that they are appear to hold with re- spect to a wide variety of persons occupy- ing quite different social roles, Thus, for example, Milgram's experiment is often commended on methodological grounds because it employed subjects drawn from a wide variety of walks of life, and "engaged in a wide variety of occupations" (Mil- gram, 1970, p, 83), and the influences on bystander apathy are presumed to apply to priests, policemen, mothers, students, and the like more or less indifferently, This creates a problem for Nisbett and Ross, one that is perhaps best illustrated by consideration of a study to which they de- vote a deal of attention, namely the 'good samaritan' study of Darley and Bateson (1973), Students at Princeton theological seminary were advised to prepare them- selves to give a brief extemporaneous talk in another building, One group of students in the 'late' condition were asked to hurry over to the other building, since they were expected a few minutes ago, Another group of students in the 'early' condition were advised that although they were not expected for a few minutes, they might as well start making their way across, On their way across, students in both groups came upon a man "slumped in a doorway, head down, coughing and groaning", 10 percent of those in the late condition offered assistance, as opposed to 63 per cent of the subjects in the 'early' condition, Nisbett and Ross claim that "these findings tell us little if anything about the personal dispositions of seminarians but a great deal about the situational determi- nants of altruism" (p, 49), However, the claim that the behavior of such subjects is determined by situational factors not only undercuts accounts based upon traditional dispositions but also any account- including the authors' own-based upon dispositions grounded in the demands of social roles, Of course, the experiment itself tells us nothing of the influence of different social role demands, In order to do so, the experi- menters would have had to vary the social roles of the subjects, to compare the ac- tions of student priests as opposed to stu- dent doctors, policemen, manual workers, and the like, in both early and late condi- tions. Now if this had been done, there might have been no significant differences in the behavior of different social role sub- jects in the late condition: most student doctors, policemen and manual workers might also have failed to offer assistance. However, there might have been signifi- cant differences in the behavior of different social role subjects in the early condition-the majority of manual work- ers might have offered assistance, but the majority of student doctors and policemen might not. Indeed, this form of variance is precisely what would be expected according to Nisbett and Ross's own account of the stability and predictability of behaviour based upon role-demands and expectations, Consequently, their interpretation of the implications of the good samaritan experi- ment begs the question, and begs the ques- tion against their own account (p, 13)): The Darley and Bateson experiment thus, in a sense, replieates but amends the lesson of the parable of the Good Samaritan. Their experiment invites LIS to surmise that all the priests and Levites who passed by on the other side of the road were simply running behind schedule. Perhaps this is the case, and the experi- ment certainly does suggest that this is a possible explanation of the behavior of the priests. However, it is also possible that it does not apply to the Levites. Given the role demands of being a Levite, most of the Levites might not have helped anyway, even if they were on schedule or ahead of schedule. Such considerations might also lead one to question Nisbett and Ross's charac- terization of situational factors as determi- nants of human behavior. For if they are correct in supposing that a great deal of the stability and predictability of human be- havior derives from role-demands and ex- pectations, then it might be better to conceive of situational factors such as late- ness, the number of other bystanders present, and the like as interferences: as factors that constrain behavior-and the predictability of behavior-in accord with role demands and expectations. If this is correct, then the lay picture may not be entirely wrong-indeed it may only be based upon certain inaccurate indi- vidualistic assumptions about those rela- tively enduring social role based dispositions that play the primary role in the explanation of human behavior, and which generally enable us to anticipate and predict human behavior in the absence of countervailing and local situational fac- tors. Moreover, this may be ultimately -all that the fundamental attribution 'error' amounts to: the supposition-which may not in fact be an error-that the primary determinant or locus of human behavior is the socially located person. The authors also appear to be blind to some of the reflexive implications of their general analysis of behavior in terms of sit- uational influences and social role de- mands and expectations, insofar as these apply to the types of experiments that they take to be supportive of their theoretical position. For example, there is no consid- eration of the possibility that the results supportive of the fundamental attribution Review of R. E. Nisbett and L. Ross 235 'error' are themselves an artifact of the role demands of peculiar experimental sit- uations, as suggested by studies in which situational or dispositional 'errors' can be produced as required via manipulations of the experimental situation (Quattrone, 1982). These sorts of considerations might very well lead one to doubt whether lay folk really are committed to CIOSS- situational behavioral consistency. Studies of cross-situational consistency usually in- volve asking subjects to predict whether a described hypothetical person will behave honestly or impulsively in some different past situation, with no mention of the so- cial role or relations of the hypothetical stranger. Whatever the ecological validity of such experiments, one wonders whether subject predictions of cross-situational consistency really are reflections of their implicit personality theories, or merely a consequence of self-presentational pres- sures to appear consistent to experiment- ers. One of the ironies of Nisbett and Ross's deflational analysis of lay accounts of personality is that it is based upon stud- ies that naively take subjects' accounts at their face value, ignoring their rhetorical and self-presentational elements. Thus although this work does point to a broader social and relational conception of persons and situations, it does not go near- ly far enough. Consequently, although Nisbett and Ross themselves seem to be clearly aware of the possibility of social theories of personality rooted in the differ- ent conventions and construals of social collectives that may vary cross-culturally and transhistorically, this important impli- cation of their work is likely to be missed by all but the most discerning reader. Too many readers-and particularly philosoph- ical readers-will simply see this work as providing further grist to the mill for con- temporary critiques of lay or folk psychology-the unhappy philosophical fate of the authors' earlier work on Human Inference (Nisbett and Ross, 1980). 236 John D. Greenwood In this review, I have concentrated on the issue that dominates this work: the con- flict between dispositional versus situation- al theories of behavior. The authors also claim to have demonstrated the utility of treating social and psychological phenom- ena as 'tension systems', as fields of im- pelling and constraining forces in states of dynamic equilibrium (such as the relations between the West and the Eastern Bloc during the period of the Cold War). It is not clear that they make a success of this, since the concept of a 'tension system' is applied to so many diverse phenomena (rivers, so- cial influence, conformity to group norms, cognitive dissonance and the like) that it is hard to make any concrete sense of it. Having said this, the actual explana- tions advanced by the authors of a variety of social phenomena under the rubric of 'tension systems' -in terms of situational influences, subjective construal, role re- quirements, cultural systems, and channel factors (factors that serve as critical facili- tators for changes in a system via disrup- tion of the eqUilibrium of forces, such as the liberalization of the Communist party in the Soviet Union that led to its recent and spectacular transformation) are gener- ally plausible and always challenging. This is particularly true of their final chapter on the potential and limitations of applied so- cial psychology, in which they offer expla- nations of why some large scale interventions fail and why some modest in- terventions can be spectacularly success- ful. Whatever the virtues of their general analysis, this particular chapter provides a realistic but optimistic estimate of the potential application of social psychology. References D' Andrade, R.G. (1981) The cultural part of cog- nition. Cognitive Science, 5, 179-195. Goffman, E. (1959) The Presentation of Self in Everyday Life. New York: Doubleday-Anchor. Goffman, E. (1961) Asylums. New York: Double- day-Anchor. Harre, R., Clarke, D., & De Carlo, N. (1985) Motives and Mechanisms. Methuen: New York. Latane, B., & Darley, 1.M. (1970) The Unrespon- sive Bystander: Why Doesn't He Help? New York: Appleton-Century-Crofts. Milgram. S. (1972) Some conditions of obedience and disobedience to authority. In A.G. Miller (Ed.) The Social Psychology of Psychological Research. New York: Free Press. Nisbett, R.. & Ross, L. (1980) Human Inference: Strategies and Shortcomings of Social Judge- ment. Englewood Cliffs, Nl: Prentice-Hall. Quattrone, G.A. (1982) Overattribution and unit formation: When behavior engulfs the person. Journal of Personality and Social Psychology, 43, 593-607. Stigler, 1.W., Shweder, R.A., & Herdt, G. (1990) Cultural Psychology: Essays on Comparative Human Development. New York: Cambridge. JOHN D. GREENWOOD DEPARTMENTS OF PHILOSOPHY AND PSYCHOLOGY CITY COLLEGE, CUNY NEW YORK, NY 10031 0