PSY 301 - 2 Discussion Questions

Group Dynamics; Theory, Research, and Practice 2000.

Vol. 4. No.

1.68-80 Copyright 2000 by the Educational Publishing Foundation 1089-2699/0055.00 DO1: 1O.1O37//1O89-2699.4.1.68 Milestones in the Psychological Analysis of Social Influence William D. Crano Claremont Graduate University Social influence research has been, and remains, the defining hallmark of social psychology. The history of this preoccupation is reviewed selectively, and important contributions to social influence and persuasion are discussed. The central thesis of the presentation is that a return to a consideration of the social group, a critical source of identity and individuality, pays major dividends in understanding the processes of social influence. Moscovici's insistence on the importance of minority influence processes is seen as a harbinger of the return of the group to social influence.

Finally, the leniency contract is proposed as a model that integrates these insights with important features of social identity, the elaboration likelihood model, and considerations of structural attitude theory in developing a predictive device that accounts for immediate and persistent majority attitude change as well as indirect and delayed focal change attributable to minority persuasion.

More than 100 years ago, Triplett (1898) published a study whose effects remain evident, even in today's analyses of social influence processes. In his experiment, Triplett demon- strated that the mere presence of coacting human beings could have a powerful impact on people's behaviors. He found that, when others were present in the research context, partici- pants worked harder and faster than when they worked alone, and he attributed this productivity enhancement to the fact that "the bodily presence of another contestant participating simultaneously in the race serves to liberate latent energy not ordinarily available" (Triplett, 1898, p. 533). Zajonc's (1965, 1980) view that the mere presence of others operated as a nonspecific, energizing stimulus helped make sense of the phenomenon, and Baron's (1986) later work showed how conflict or distraction produced by the copresent actors could produce performance gains or losses. Although social facilitation research continues to be pursued (e.g., Aiello & Svec, 1993), the field today is This research was supported in part by National Institute on Drug Abuse Grant R01-DAI2578-0I, for which I am most grateful.

I acknowledge the kindness of Rod Bond, who provided me a prepublication draft of his meta-analytic work on group size and conformity.

Correspondence concerning this article should be ad- dressed to William D. Crano, Department of Psychology, Claremont Graduate University, 123 East 8th Street, Claremont, California 91711.

Electronic mail maybe sent to [email protected].

more inclined to study the impact of actors who take a more directive role in shaping beliefs and actions in the interpersonal context. In this review I trace some of the major developments in directed social influence, from Sherif to today's studies of majority and minority influ- ence.

Along the way, 1 consider important questions about the processes that might result in different influence outcomes and suggest possibilities that may foster understanding of this fundamental activity. This review is not a complete rendering of the literature, nor is it intended to be. Rather, it highlights some of the studies that have had an inordinate influence on the directions the field has taken. Finally, the review will trace the disappearance and reappear- ance of the social group in social influence and document the significance of this reemergence for integrating developing theories.

The importance that social psychology at- taches to directed social influence can be inferred from the field's continuing preoccupa- tion with questions surrounding this issue, and the preeminence of those who have labored on their solution. Of the great names in social psychology's pantheon—-Allport, Asch, Camp- bell, Festinger, Hovland, McGuire, Moscovici, Jones, Kelley, and Sherif—all have devoted at least a portion of their considerable talents to developing a better comprehension of social influence. Much of the early research under- taken by this distinguished company of scholars was devoted to two tasks: uncovering factors SPECIAL ISSUE: SOCIAL INFLUENCE 69 that enhanced social influence, and developing frameworks in which the interplay of these factors could be understood and mapped onto existing empirical data. I consider some of these frameworks over the course of this review, but before doing so it is reasonable first to consider some of the early studies on influence that gave rise to social psychology's intense interest in the first place.

Sherif s Autokinetic Research Series Solomon Asch is typically acknowledged as the father of social influence research, and there can be little doubt that his impressive series of reports in the 1950s called attention to the intriguing and not easily explicable phenom- enon of compliance. However, to ignore the equally impressive studies conducted by Muza- fer Sherif 20 years before Asch's is to proceed in extreme peril. Evidence of Sherifs creative genius is well distributed throughout the litera- ture of social psychology, but nowhere is it more obvious than in his autokinetic illusion research (Sherif, 1935, 1936). The illusion relies on a pervasive perceptual shortcoming of human beings. It requires only that a perceiver fixate on a small pinpoint of light in an otherwise completely dark room.1 After focusing on the light for a short period of time, it appears to move. Even experienced judges are not immune to this compelling illusion. Sherif (1936) took advantage of this well-established but poorly understood human foible to study norm forma- tion, a theme to which he would return and augment in the classic Robbers Cave experiment (Sherif, Harvey, White, Hood, & Sherif, 1961).

In his initial autokinetic study, Sherif (1936) required participants merely to estimate the length of (illusory) movement of the light over a series of judgment trials. The impressive over- lap among participants' judgment patterns laid the foundation for all of his important work on social influence. The first study revealed enor- mous variation in judgments on early trials; but as perceivers gained task experience, within- participant oscillations abated. It was as if each participant had arrived at a self-defined con- sensus, which, once developed, guided later judgments.

The across-trials damping of variation in individual judgments suggested an interesting possibility. If one could develop a consensus around one's own judgments, could not the same process of consensus-based norm formation be accelerated when responding in concert with others? Triplett's (1898) earlier work suggested clearly that the implicitly and inadvertently directive actions of coacting participants in an autokinetic experiment would have profound effects. To test this possibility, Sherif (1936) paired naive respondents in a second autokinetic judgment study. Over a series of autokinetic trials, their assignment on each was to estimate in tandem the movement of the light. Response order was fixed, and no confederates were used.

Sherif found that the naive participants influ- enced one another. The influence was attribut- able to more than mere presence because each participant's response bore directly on that of his or her partner. Attenuation of response variation was rapid. Participants used one another as models of reality and moved expeditiously to a mutually acceptable judgmental accommoda- tion over the course of their brief interaction.

A third study revealed that this accommoda- tion was not undertaken in the service of peaceful coexistence. Recall that each partici- pant reported a distance estimate on each trial.

Conceivably, a person who actually "saw" the light move 5 feet might be reluctant to report this perception if his or her partner had reported a mere 2 inches. To maintain a more harmonious situation, the second respondent might report a much-diminished light movement. The rapid accommodation of paired participants' re- sponses then would be attributable to public acquiescence (of the second responder in the pair) rather than a rational weighting of perceptual and socially supplied information.

To investigate this possibility, Sherif (1935) conducted a third study, in which participants who had responded with a given partner were reassigned to new response groups. Sherif believed participants formed response norms in the first session. If so, then subsequently switching participants into new groups should have little effect. At least over the short term, they would continue to respond as they had at the end of the initial norm-formation session.

Conversely, if the (second-responding) partici- pants had merely acquiesced to the implicit demands of their response partners, if nothing 1 The impact of the illusion is enhanced if perceivers do not know the true dimensions of the setting. 70 CRANO other than simple accommodation were actually represented in their behavior, then a similar form of accommodation would be found in the second session, when each responded in tandem with a new partner. Such a pattern of acquies- cent accommodation, however, was decidedly absent in the second session of Sherif's third experiment. Participants maintained the re- sponse norm they had formed on the initial paired trials and remained resistant to the information/implied influence provided (in the form of a perceptual judgment) by their new partners.

This series suggests strongly that the partici- pants had learned something in the norm- formation studies and that this learning had affected either their verbal reports or their eyesight. It is not possible to choose between these two alternatives on the basis of the research reported to this point, but a study published nearly three decades after the initial autokinetic studies provides grounds for interest- ing speculation. In this experiment, Hood and Sherif (1962) paired a naive participant with a confederate in an autokinetic light-movement estimation task. In the initial phase of the experiment, the confederate made a series of light movement judgments. The naive partici- pant simply observed. These confederate-based judgments were said by the authors to be either consistently high or consistently low.2 After this observational session, the naive participant made an independent series of judgments after the confederate had left the setting.

The results of this treatment are as important as they are heuristic. The judgments of naive participants studied in the condition in which the confederate responded with high estimates were significantly greater than those of participants paired with a low-responding accomplice. Hood and Sherif (1962) argued that their results suggested that people pattern their responses so as to be in accord with those of an influence source, and when they do so they are not necessarily responding to pressure. They argued that interpersonal pressure is probably not a reasonable term to use in describing the force of an influence source in an unstructured situation.3 In their experiment, there was no obvious pressure laid on respondents to adopt the response norm of their partner. Indeed, by the time the participant had begun to voice a response to the apparent light movement, the confederate had left the building. Sherif argued that simple compliance cannot provide a reason- able explanation of this result; there must be more to the process than acquiescence. He suggested an approach to understanding confor- mity that relied on a more rational conception of social action, a view of people as data- dependent information processors who used the responses of others in deciding on the proper action or the proper judgment, especially in unusual or novel contexts.4 Asch's Line Judging Research It is important to appreciate the historical context of Hood and Sherif's (1962) arguments in defense of their presumption that participants rationally used socially supplied inputs in coming to a judgment of belief or appropriate action. By this time, the conformity research of Solomon Asch (1951, 1952, 1955, 1956) had stimulated widespread attention. Asch's findings did not appear amenable to Sherif's desire (or Asch's, for that matter; see Campbell, 1990) to characterize humankind as rational information processors, as he had argued years before. The Asch line judgment research series is so well established that its description is unnecessary; however, for heuristic purposes, certain features of this approach deserve emphasis. Recall that Asch required participants to determine which of three comparison lines matched a stimulus line presented over a series of judgment trials.

When studied in isolation, participants' judg- 2 How one can be sure that an illusory judgment was perceived by the watching participant as high or low was not explained compellingly, given the impossibility of knowing the actor's perception of an illusion, but even the most critical reader can accept the fact that the estimates made in the "high-responding confederate condition" were substan- tially greater than those made in the low-responding confederate treatment.

3 The term "unstructured situation" is used advisedly. By il, Hood and Sherif (1962) implied that the judgment processes of participants placed in contexts in which they have little prior experience may be very different from those exhibited in familiar, well-learned settings (see Gorenflo & Crano, 1989, for a parallel discussion of the role of judgment type on social comparison processes).

4 As Sherif and Hovland (1961) suggested in their classic monograph, human actors were not seen as completely rational. Biases bom of prior beliefs, especially highly ego-involved beliefs, were expected to affect perceptions. SPECIAL ISSUE: SOCIAL INFLUENCE 71 ments, for all practical purposes, were perfect.

The task was so simple, the correct answer so inescapable, that no mistakes were made. As in Sherif's series, Asch's initial study was a mere preliminary. A perceptual, rather than social- psychological, exercise was used to establish unequivocally that the judgments required of participants were so transparent that any devia- tions from the obvious represented more than mere misperception.

Such deviations, of course, are the stuff of which Asch's reputation is made. He found that naive participants often made egregious errors of judgment when responding in concert with two or more trained confederates who had been coached, unwaveringly, to give the wrong answer on specified judgment trials. Two important findings from Asch's research deserve comment: Confederates' impact was evident only when they were unanimous; further, enlarging the size of the (unanimous) confeder- ate majority beyond 3 had little effect. Appar- ently, the compliance-inducing effect of the unanimous peer group quickly reached asymp- tote.

Asch's research suggested that 15 unani- mous stooges were no more persuasive than 3 were.5 The failure to discover a group size effect could not easily be made compatible with the rationalist position of Sherif.

Clearly, if 15 people reported a perception at odds with one's own, the combined weight of their views should be greater than that of only 3 other perceivers, if a rational weighting decision-making process were occurring. That such findings did not emerge appeared not to support to Sherif *s views.

Even worse, the simplicity of Asch's judgment task, the near-inevitability of a cor- rect response under nonitifluenced conditions, as was evident in the control participants' near-perfect scores, rendered difficult an optimis- tic, Rousseau-like view of the noble naive perceiver. A more cynical interpretation ap- peared justified, one in which perceivers were viewed as easily swayed by the dictates of a largely disinterested majority, who apparently could affect even the most fundamental judg- ments by mere surveillance and the implied threat of disapproval. Milgram (1974, 1977) expanded on this theme in a later and very influential series of studies.

Early Attempts at Integration The implications that might be drawn from the pioneering work of Asch and Sherif appeared very different, and the research that followed did not provide an instant solution to the apparent lack of fit between the two significant series. Insko, Smith, Alicke, Wade, and Taylor (1985) found that the size of the apparent majority did have an impact on judgments in a color-judging task. This result was magnified when participants were led to believe that a verifiable (right-wrong) decision could be made. The issue of verinability is important in that it might signal to participants that their judgments may be adjudged valid or invalid. If verifiability is at issue, there must be some consensually agreed-on answer. This probably is not the impression Asch's respon- dents had, if we give credence to the results of his control group participants. Other research also produced findings at odds with Asch's contention that absolute majority size was irrelevant (e.g., Gerard, Wilhelmy, & Conolley, 1968; Kumar, 1983), and contemporary models of social impact or social influence have resolutely denied Asch's contention while simul- taneously attempting to come to grips with his results (Latane & Wolfe, 1981; Tanford & Penrod, 1984).

The field's response to this variation in results was well considered and persistent. Four years before dissonance, Festinger (1953) published a theoretical article that called for the recognition of the differentiation between public compliance and private acceptance. Three years later, Deutsch and Gerard (1955) argued the distinc- tion between normative and informational social influence, and Kelman's (1958) classic tripartite distinction of compliance, identification, and intemalization followed close on the heels of this work. These approaches were all attempts to make theoretical sense of the wide variations in response to social influence that had been seen in the literature and whose conceptual bound- aries were defined by the findings of Asch and Sherif.

5 This result should not be overinterpreted. Research suggests that in persuasion settings involving more complex issues or that admit to more varied persuasive arguments than "Line B is the right match," more sources result in greater influence (e.g., Harkins & Petty, 1983, 1987). 72 CRANO The difficulty with Festinger's, Deutsch and Gerard's, and Kelman's attempts at organization and synthesis is that the explanatory mechanism developed to answer the central question, "Under which conditions will one process (compliance vs. acceptance; informational vs.

normative influence; compliance, identification, or internal!zation) occur?" were either ill formed or incomplete. For all three models, the answer to the question was usually couched in terms of variations in source qualities. Source expertise, competence, attractiveness, skill, abil- ity, and the like, made for strong influence effects, which often persisted. Brute force— number, surveillance, strong social pressure— also made for influence, but the effects did not persist, nor did they follow a pattern that suggested learning or cogent integration of information.

The problem with all three of these explana- tory devices is that the postulated source effects sometimes are reversed; expert sources some- times produce short-lived (or no) change, and brute force effects sometimes persist. The classic models have problems with such rever- sals.

Apotential explanation of this variability in source effect seems to lie in the interaction of source and task (which I take to include variations in message strength). If the task is one on which the correct answer is inescapable, there would seem to be little possibility for information-based influence, socially supplied or otherwise. Such a task description character- izes Asch's line judgment paradigm, if we believe his control group results. However, it is not necessarily true of work undertaken on a Crutchfield-type apparatus, a methodological variant of Asch's paradigm that sometimes tellingly produced results quite at odds with Asch's usual result, often more in line with the results of Sherif's research (Crutchfield, 1955).

The apparent difference between these two approaches is that Crutchfield mechanized Asch's confederates. In most other critical aspects, Asch's and Crutchfield's research con- texts were identical. The less apparent but more important difference is that, in mechanizing Asch's procedure, Crutchfield incidentally en- larged the arena of judgments that could be, and were, required of participants across various research endeavors. Asch seemed more or less stuck with a task in which the answers to all questions were immediately obvious. That participants sometimes gave the nonobvious answer was the feature of his studies that made the research noticeable. Imagine the lack of interest his research would have generated had the vast majority of Asch's participants resisted influence. Crutchfield's apparatus was applied in a much wider variety of judgment contexts.

Far from being constrained to obvious percep- tual judgments reported under great social pressure, Crutchfield's device was used to test participants' responses on a host of issues, ranging from perceptual judgments to answers to obscure factual items.

The importance of this variation of contexts is that the required judgments varied in terms of the apparent inescapability of a valid answer.

Research suggests that the more difficult or uncertain the judgment, the more likely are informational social influence and private accep- tance (e.g., Crano, 1970; Endler, 1965). With issues of a more objective nature, in which participants are more certain of the correct answer, we find (short-term) normative influ- ence but little, if any, evidence of informational influence (see Gorenflo & Crano, 1989). The perception of inevitability plays a role in the process that will be activated. Considering the interactive combination of source and context materially enhances explanatory power.

Campbell's Framework This more interactive view suggests a new way of conceptualizing the influence process.

Who better to systematize this insight than Campbell (1961, 1963), who developed a gen- eral framework based on fundamental psy- chological and social-psychological principles that could engage the most extreme forms of acquiescent behavior, from compliance to con- formity to identification to independence. Camp- bell's view acknowledged and made creative use of long-established principles to produce a framework that predicted new outcomes while simultaneously mapping onto existing data. The vision he presented in his general model of social influence, proposed in an important set of theoretical articles in 1961 and 1963, extended conceptually well beyond the bound- aries proposed in the theory of social compari- son (Festinger, 1954). These articles speak for SPECIAL ISSUE; SOCIAL INFLUENCE 73 themselves and need little expansion. In brief, Campbell theorized that people differentially use three broad classes of data when deciding on a course of action: (a) history or prior experi- ence, (b) contextual information directly per- ceived, and (c) socially supplied information about a person or context that people may or not have directly experienced themselves. Combin- ing these factors in an action- or decision- making context can account for variations in behavior ranging from complete compliance to complete anticonformity (Brewer & Crano, 1994; Nail, 1986). Paradigmatically, most social influence research negates the first source of behavior-relevant information and assesses the strength of the factors that affect the relative weighting of the second and third sources (i.e., information directly perceived or socially sup- plied by an external agent). This framework for understanding provides a useful structure for organizing the myriad factors that may impinge on human social action (Crano, 1970). Camp- bell's view requires a departure from a main effects orientation. By its very structure, it demands that features of the source and context be consulted in determining the probable outcome of an attempt at social influence.

Persuasion and Message-Based Attitude Change One might have predicted that this multifacto- rial framework for conceptualizing features that affect social influence might have useful and widespread application to a range of research endeavors, especially those involved in the study of people's actions under pressure. How- ever, a parallel development in social psychol- ogy in some ways counteracted Campbell's insight, at least for a while. At the time of great fomentation in social influence research, the study of persuasion or attitude change also was moving into high gear.6 Much of this work was guided by Lasswell's (1948) mantra, "Who says what to whom, how, and under which circum- stances?" a formula that was meant to capture essential components for a comprehensive theory of attitude change. Lasswell's is a good incantation, and it has served us well over the past 50 years. However, it does admit to a serious shortcoming because the interdependent pieces of the formula can be studied perfectly well in splendid isolation and, moreover, they can be tested in circumstances that largely deny the reality that social influence is often best conceptualized as an interpersonal, rather than as a solitary, intrapersonal phenomenon. Thus, if we were not careful, (a) we might have developed a literature on source credibility (the "who" part of the prescription) that was devoid of considerations of other parts of the formula (the "what," "whom," "how," and perhaps, most importantly, the "under what circum- stances"), (b) we might have attempted to develop a psychology of susceptibility indepen- dent of considerations of source or issue or message or context, (c) we might have at- tempted to forge an understanding of message factors (the "what") that was independent of the "who" or "whom" and the context in which the what was delivered, and (d) we might even have developed a literature of social influence in which the interaction of influence source, target, message, and context was not considered a relevant term in the change equation. In view of Campbell's framework and the events leading up to its development, it is clear that these attempts would have been ill-advised, but until recently (with the advent of the dual-process models of attitude change) they are largely what we have been doing in persuasion research. This is not to say that we have been treading water for the past 40 years. Over this span, we have identified many factors that positively affect people's tendency to weigh socially supplied inputs relative to independent perceptions.

These factors have to do with the qualities of the person or group that is laying on the influence, the personal qualities of the individual receiving the (socially supplied or directly experienced) information, and the social-psychological con- text (external and internal) in which the information is conveyed or in which the critical behavior is to take place.

We have learned, for example, that sources that possess greater expertise, competence, confidence, and other socially valued traits typically induce greater influence than those that do not. By the same token, we have learned that target individuals who lack expertise, compe- tence, and confidence or who in some other 6 Currently, persuasion and social influence processes are viewed as guided by the same fundamental principles (see Chaiken, Wood, & Eagly, 1996), but history suggests that this view has not long been in the majority. 74 CRANO ways perceive themselves to be subordinate or inferior to an influence source will prove more susceptible to the source's (socially supplied) information. We have learned that beliefs of relatively little importance, or of low vested interest, are easy to influence but have few implications for behavior, whereas those that have important personal consequences are resistant but not impossible to change, and when they do change, they are strongly directive of action (Crano, 1995, 1997; Petty & Krosnick, 1995).

In addition, we have learned that contexts involving difficult or ambiguous judgments tend to incite targets to overweigh socially supplied information relative to their own direct percep- tual inputs. Arrangements that anticipate these empirical regularities are built into all contempo- rary theoretical models of social influence.

Further, these models typically require us to consider more than one variable at a time when gauging the likelihood and extent of persuasion or social influence. This more multi factored view of the persuasion process has bled into our considerations of social influence, conformity, and compliance, although possibly not as thoroughly as we might have hoped (Petty & Cacioppo, 1986a, 1986b; Wood, in press).

Moscovici's Reconsideration Obviously, we still have far to go in adopting a more complex view of social influence, but on the whole the movement of our conceptualizing is in the right direction. Recent theoretical treatments of social influence are considerably more complex than their forebears, and the complexity makes sense. They are grounded on reasonable interpretations of regularities in the literature, and they allow prediction of complex variations in the data pattern that are both common and reproducible. This general movement toward a more multifactorial and interactive conceptualization of social influence and the processes that underlie it is a consistent feature of contemporary theory and research in social influence with one apparent and important exception. That exception is seen in the work of Moscovici, whose conflict-based theory, along with the dual-process attitude models, is largely responsible for the revitaliza- tion of psychology's interest in social influence.

Moscovici's model focuses almost entirely on source and target characteristics and devotes little attention to contextual factors, no attention to message effects, and even less to the inter- action of these features of the influence context (Moscovici, 1974, 1976, 1980, 1985a, 1985b; Moscovici, Lage, & Naffrechoux, 1969; Mos- covici & Personnaz, 1980, 1991). This is not to say that the model cannot be used. Its very unidimensionality invites additions and modifi- cations, and social influence researchers have shown themselves ready to append interactive features to the fundamental conflict theory that both enrich its predictions and expand the realm of phenomena to which it might apply. Bringing to bear the insights born of the intersecting histories of social influence and persuasion, current models of majority and minority social influence illustrate the best of integrative science, the construction of new insights from earlier regularities.

At its heart, Moscovici's model is concerned with factors that affect a minority group's power to influence the majority or, from the opposite perspective, with the majority's power to move the minority. The model holds that majorities and minorities instigate distinct belief-change processes. Influence targets finding themselves in disagreement with the majority are thought to focus on the negative interpersonal ramifica- tions of their deviance, which can include ostracism and other sanctions. These negative ramifications stimulate apparent movement in the direction of the majority, but the movement is primarily undertaken in the service of avoiding censure. It is not based on a well- considered elaboration and appreciation of the logic of the source's position. The majority persuades because it possesses coercive power, the capacity to monitor and to punish misbehav- ior. Overt compliance, but not conversion or private belief change, resolves source-target conflict in such cases. Conversely, by virtue of the unexpectedness of their position, minorities stimulate targets to try to understand why they hold a particular view. The outcome of a target's quest for understanding can result in minority- based social influence. However, owing to a hypothesized reluctance to be identified with the minority, Moscovici's theory assumes that minority influence will be delayed or that it will be seen on altitudes associated with, but not identical to, the focus of persuasion (e.g., Perez &Mugny, 1987, 1990).

By focusing on influence sources specifically SPECIAL ISSUE: SOCIAL INFLUENCE 75 designated as being of majority and minority status, Moscovici explicitly obligated research- ers to include consideration of the social group as an essential element to understanding social influence.

He was not primarily concerned with the internal workings of the individuals under persuasive stress, a common theme of much of mainline social psychology, especially North American social psychology.

His emphasis on the group exposed a shortcoming of the excessively intraindividualistic orientation to social influence that had come to characterize much of the field (see Turner, 1991).

To appreciate the way in which Moscovici's insight opened the door for progress, consider the prototypic minority influence study.

In the standard study, participants are exposed to a persuasive communication that is contrary to their established beliefs, attributed to a majority or (typically, in-group) minority. Premeasures sometimes are taken, and posttest measures on the targeted or focal attitude are administered immediately and sometimes after a delay of a week or 2.

Occasionally, associated beliefs are assessed as well, and, infrequently, message quality also is manipulated. When the source represents the majority, Moscovici predicted immediate change on the targeted (focal) attitude, with little persistence and no diffusion of effects to associated attitudes. Resistance is the predicted response to persuasive messages attributed to minority sources; however, change on associated beliefs (sometimes termed indi- rect change) is expected, as is delayed change on the focal belief itself.

Leniency Contract The problem with this set of predictions is that it often falls short: Majority effects sometimes persist, and minorities occasionally produce indirect change without delayed focal change. Moscovici's theory was not sufficiently developed to digest these rich findings, but models that appeal to the combined literatures of social influence, social identity, and interper- sonal relations support provocative and poten- tially critical insights.

Let us consider one such integration in which these considerations are commingled in a theoretical model:

the leniency contract (Crano & Chen, 1998).7 The potential utility of the leniency model is that it integrates into a single predictive framework important considerations derived from the social influence, persuasion, and social identity traditions.

As such, it offers a means by which the rich insights of the past 100 years may be opportunistically drawn on in developing a model that is at once inclusive and parsimonious.

The initial assumption of the leniency con- tract is that the targets of persuasive attacks consider the self-relevance of the issue under debate before deciding on a course of response.

Relevance may be adjudged in terms of the issue's perceived vested interest (Crano, 1995, 1997; Crano & Alvaro, 1998; Sivacek & Crano, 1982) or the implications of compliance or resistance for place maintenance in the social group.

The leniency model assumes that social identity concerns (Tajfel, 1982; Tajfel & Turner, 1986) become prominent as a consequence of the mere specification of source as being of majority or minority status. Unlike earlier research on source credibility (e.g., Hovland, Janis & Kelley, 1953), this particular character- ization renders the context interpersonal and relevant to considerations of the self.

A majority source is a representative of the modal opinion on an issue in an assemblage of individuals that the target uses as a consequential feature of his or her self-definition. A source who is a numeric minority on some subjectively important charac- teristic (e.g., race, religion, ethnicity) or who propounds a minority point of view is deviant, or holds beliefs that are deviant, from the majority of the reference or membership group.

As such, a description of an influence source as being of majority or minority status is relational; it suggests an association or connection between target and group that is of some consequence for the target. Depicting a source as of minority or majority status not only describes features of a source but also suggests a relationship with the 7 This is only one of a number of new models of social influence that might have been considered. Models devel- oped by De Dreu and De Vries (1996) and Mugny's research group (e.g., Mugny, Burera, Sanchez-Mazas, & Perez, 1995; P^rez & Mugny, 1996) might have been profiled just as profitably, but the pull of paternity is strong, and the leniency model makes explicit reference to persuasion-based factors as well as those whose effects usually are examined in the social influence laboratory and thus facilitates my theoretical point better than the others. 76 CRANO target vis-a-vis the group on an issue or characteristics of potential importance. If this were not the case, the source and target classification of majority or minority could not be expected to have much impact. Such a characterization would serve as a heuristic rather than as a systematic cue (see Bohner, Frank, & Erb, 1998; Chaiken, 1987; De Dreu & De Vries, 1993, 1996; Erb, Bohner, Schmalzle, & Rank, 1998). In the leniency model, majority/ minority source characterization may induce systematic as well as heuristic processing.

A counterattitudinal communication from the in-group thus presents a relational threat that may jeopardize the target's relationship with a potential source of self-identity. A threat of this nature is ignored at great peril.

Reactions to such threats can take many forms, but at a minimum we would expect the target to consider the issue that is the source of contention with the identity-endowing source. The leniency contract suggests that reactions to a message will vary as a function of features of both the source and its position relative to the well-being of the group (Alvaro & Crano, 1996, 1997; Crano, 1994; Crano & Chen, 1998; Crano & Hannula-Bral, 1994).

If the issue under discussion is not one on which the majority is seen as having a legitimate position—one on which it has no legitimate voice—its impact will be greatly attenuated.

Group pressure will be adjudged inappropriate, and the group will lose stature. If the issue is relevant to the group, as, for example, a con- sideration of comprehensive examinations or tuition would be for a group of students of the same university, the majority is seen as having a legitimate voice (Baker & Petty, 1994; Mackie, 1987).

In this circumstance, its message will be considered carefully. If the message is weak and unpersuasive, the model suggests the majority might have an impact, but it will be short lived.

If the message is strongly argued, however, it may have both immediate and long-term effects.

Considerable research supports these proposi- tions (e.g., Baker & Petty, 1994; Crano & Chen, 1998; De Dreu & De Vries, 1993,1996; Mackie, 1987).

Thus, depending on the relevance and strength of its message, a majority may have no effect, a short-lived result, or a lasting impact.

This complex predictive pattern is not a feature of any other theory of majority influence, but it is completely consistent with the theoretical expectations derived from the leniency contract.

As is shown, this model also provides a theoretical explication of minority influence that accounts for both the delayed focal change and the (immediate) indirect change effects that the meta-analysis of Wood and her colleagues found associated with minority sources (Wood, Lundgren, Oueilette, Busceme, & Blackstone, 1994).

In the case of minority-based persuasion, the leniency contract assumes that uneasiness or perceptions of threat do not arise in targets that find themselves at odds with the counterattitudi- nal pronouncements of an in-group minority, unless the minority threatens the very essence of the group, the group's raison d'etre. In that case, the in-group becomes out-group and is marginal- ized, ostracized, or ignored. In cases in which the in-group minority's position does not threaten the continued existence of the group, however, the leniency model holds that the minority will receive courteous and polite treatment from the majority. Owing to concerns with group solidarity, maintenance, cohesion, and stability, in-group minorities are not dero- gated for holding the position that creates their minority status in the first place. On these issues of less critical moment, the minority's message is elaborated with little counterargument and little source derogation. From considerable research on persuasion, we know that an ideal formula for change involves the active elabora- tion of a strong message coupled with little counterargumentation and no (or minor) source derogation. This is precisely the formula sug- gested by the leniency contract in detailing the majority's presumed reaction to in-group devi- ants, at least when the critical issue is not central to the viability of the group. Such open-minded and poorly defended elaboration would be expected to result in continuous oscillation of the majority's position and continual instability in the group's beliefs. Yet we know from considerable research that the majority position often is quite persistent and resistant to change.

How, then, can the theoretical claims of the leniency model be sustained? How, in this system, can the group maintain its integrity and defend the status quo, operations that Moscovici claimed are central to the majority?

To respond to this legitimate issue, we must consider one additional feature of the leniency contract. As with all contracts, the leniency contract specifies a quid pro quo. In recompense SPECIAL ISSUE; SOCIAL INFLUENCE 77 for the open-minded elaboration of the minori- ty's appeal and the non-derogatory treatment that is provided, the model posits a cost. That cost is paid in terms of an implicit understanding that no change will ensue from the persuasive interchange. This implicit agreement is a pivotal theoretical feature of the model. It allows for the maintenance of the majority's core beliefs while assuaging or, at a minimum, not alienating a significant factor in the group, the in-group minority. The model provides an ecologically appurtenant prescription for group survival. It allows for considerable attitudinal variation within groups on all issues except those that put the viability of the group into jeopardy. Re- search on intragroup relations suggests that such apparent beneficence is a common feature of cohesive groups. The leniency model thus provides the means for the group to tolerate some degree of freedom of expression while simultaneously protecting and maintaining the status quo. At least on the surface, such a complex process would appear to offer the best of all possible worlds, except for those indi- vidual group members who, as minority voices, are truly intent on influencing the character of the majority.

All is not lost for the minority, however.

Although the contractual feature of the leniency model appears to account for stability in established groups, it also offers the means by which the minority can move the majority. The mutually agreed-on understanding that majority group members will not change as a result of an open-minded elaboration of the minority's position does not negate the fact that strong change pressures have been introduced as a function of the manner in which the minority's message is elaborated. How is this pressure diffused? The leniency model suggests that, by spreading activation, beliefs in close cognitive contiguity to the targeted issue are put at risk as a consequence of the majority's leniency (Anderson, 1983), because, although targeted (focal) beliefs are strongly defended in this system, related beliefs are not. The change pressure experienced as a result of the hypoth- esized lenient responses to minority influence is diffused to these related beliefs, which, unde- fended, are easy prey for attitude change pressures. This process suggests a means by which indirect attitude change, a relatively consistent feature of minority influence (see Wood et al., 1994), comes about. The model also suggests a mechanism for delayed focal change, another consistent feature of the literature.

The leniency model's account of indirect attitude change requires acceptance of the possibility that attitudes are not held in isolation but rather are related to a greater or lesser extent to all other beliefs that constitute the cognitive system. The model assumes that the greater the cognitive contiguity, the greater is the possibil- ity for indirect change, and research supports this expectation (e.g., Alvaro & Crano, 1997; Crano & Chen, 1998). This structural account of indirect change also provides a means of understanding delayed focal change. If an atti- tude is changed, then those that are (or were) contiguous in cognitive space will be put under pressure to change as well, if attitudes are linked in some form of structure, as proposed. Thus, if indirect attitude change occurs in circumstances compatible with those described by the leniency contract, we would expect delayed focal change if the attitude altered by indirect attitude change processes was sufficiently moved to affect the equilibrium of the overall structure of beliefs.

In other words, when a minority source induces major indirect attitude change, we would expect subsequent (delayed) change on the focal issue, especially if the indirect change were a conse- quence of a strongly argued focal message. This is precisely the result discovered by Crano and Chen (1998). The leniency model, then, can account for both indirect, minority-induced attitude change and delayed focal change. It provides a theoretically plausible account of the manner in which minorities and majorities persuade while simultaneously supplying a reasonable method of accounting for prior results in minority/majority influence. Perhaps most importantly, the leniency contract is in close accord with a changing emphasis in the field of social influence, a return to earlier days when the centrality of the social group in influence was widely acknowledged.

Over the years, social influence had departed from its roots as an interpersonal process. The integration of considerations of group identity, intergroup process, persuasion, belief structure, and message elaboration, central building blocks of the leniency contract, all combine to move our consideration of social influence back to its heritage. Although the model clearly makes use of recent results in social cognition, the leniency 78 CRANO contract also acknowledges—indeed is depen- dent on—intergroup processes to provide a better and more precise account of the ways in which majorities and minorities persuade.

References Aiello, J. R., & Svec, C. M. (1993). Computer monitoring of work performance: Extending the social facilitation framework to electronic pres- ence.

Journal of Applied Social Psychology, 23, 537-548.

Alvaro, E. M., & Crano, W. D. (1996). Cognitive responses to minority or majority-based communi- cations: Factors that underlie minority influence.

British Journal of Social Psychology, 35, 105-121.

Alvaro, E. M., & Crano, W. D. (1997). Indirect minority influence: Evidence for leniency in source evaluation and counterargumentation. Journal of Personality and Social Psychology, 72, 949-964.

Anderson, J. R. (1983). The architecture of cognition.

Cambridge, MA: Harvard University Press.

Asch, S. E, (1951). Effects of group pressure on the modification and distortion of judgments. In H.

Guetzkow (Ed.), Groups, leadership, and men (pp.

177-190). Pittsburgh, PA: Carnegie Press.

Asch, S. E. (1952). Social psychology. Englewood Cliffs, NJ: Prentice Hall.

Asch, S. E. (1955). Opinions and social pressure.

Scientific American, 193, 31-35.

Asch, S. E. (1956). Studies of independence and conformity:

A minority of one against a unanimous majority. Psychological Monographs, 70(9), (Whole No.

416).

Baker, S. M., & Petty, R. E. (1994). Majority and minority influence: Source advocacy as a determi- nant of message scrutiny. Journal of Personality and Social Psychology, 67, 4-19.

Baron, R. S. (1986). Distraction/conflict theory:

Progress and problems. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol.

19, pp.

1^40).

New York:

Academic Press.

Bohner, G., Frank, E., & Erb, H.-P. (1998). Heuristic processing of distinctiveness information in major- ity and minority influence. European Journal of Social Psychology, 28, 185-206.

Brewer, M. B., & Crano, W. D. (1994). Social psychology. Minneapolis, MN: West Publishers.

Campbell, D. T, (1961). Conformity in psychology's theories of acquired behavioral dispositions. In I. A. Berg & B. M. Bass (Eds.), Conformity and deviation (pp. 101-140). New York:

Harper.

Campbell, D. T. (1963). Social attitudes and other acquired behavioral dispositions. In S. Koch (Ed.), Psychology: A study of a science:

Vol.

6. Investiga- tions of man as socius (pp. 94-172). New York:

McGraw-Hill.

Campbell, D. T. (1990). Asch's moral epistemology for socially shared knowledge. In I. Rock (Ed.), The legacy of Solomon Asch: Essays in cognition and social psychology (pp. 39-52). Hillsdale, NJ:

Erlbaum.

Chaiken, S. (1987). The heuristic model of persua- sion. In M. P. Zanna, J. M. Olson, & C. P. Herman (Eds.), Social influence: The Ontario symposium (Vol. 5, pp. 30-39). Hillsdale, NJ: Erlbaum.

Chaiken, S,, Wood, W., & Eagly, A.

(1996). Principles of persuasion.

In E. T. Higgins & A.

W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 702-742). New York: Guilford Press.

Crano, W. D. (1970). Effects of sex, response order, and expertise in conformity: A dispositional approach. Sociometry, 33, 239-252.

Crano, W. D. (1994). Context, comparison, and change: Methodological and theoretical contribu- tions to a theory of minority (and majority) influence. In S. Moscovici, A. Mucchi-Faina, & A.

Maass (Eds.), Minority influence (pp. 17-46).

Chicago: Nelson-Hall.

Crano, W. D. (1995). Attitude strength and vested interest. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences.

The Ohio State University series in attitudes and persuasion (Vol. 4, pp. 131-157). Hillsdale, NJ:

Erlbaum.

Crano, W. D. (1997). Vested interest, symbolic politics, and attitude-behavior consistency. Journal of Personality and Social Psychology, 72, 485- 491.

Crano, W. D., & Alvaro, E. M. (1998). The context/comparison model of social influence:

Mechanisms, structure, and linkages that underlie indirect attitude change. In W. Stroebe & M.

Hewstone (Eds.), European review of social psychology (Vol. 8, pp. 175-202). Chichester, England: Wiley.

Crano, W. D., & Chen, X. (1998). The leniency contract and persistence of majority and minority influence. Journal of Personality and Social Psychology, 74, 1437-1450.

Crano, W. D., & Hannula-Bral, K. A. (1994).

Context/categorization model of social influence:

Minority and majority influence in the formation of a novel response norm. Journal of Experimental Social Psychology, 30, 247-276.

Crutchfield, R. S. (1955). Conformity and character.

American Psychologist, 10, 191-198.

De Dreu, C. K. W, & De Vries, N. K. (1993).

Numerical support, information processing, and attitude change. European Journal of Social Psychology, 23, 647-662.

De Dreu, C. K. W, & De Vries, N. K. (1996).

Differential processing and attitude change follow- ing majority and minority arguments. British Journal of Social Psychology, 35, 77-90. SPECIAL ISSUE: SOCIAL INFLUENCE 79 Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influence upon individual judgment. Journal of Abnormal and Social Psychology, 51, 629-636.

Endler, N. S. (1965). The effects of verbal reinforce- ment on conformity and deviant behavior. Journal of Social Psychology, 66, 147-154.

Erb, H.-R, Bohner, G., Schmalzle, K., & Rank, S.

(1998).

Beyond conflict and discrepancy: Cogni- tive bias in minority and majority influence.

Personality and Social Psychology Bulletin, 24, 620-633.

Festinger, L. (1953). An analysis of compliance behavior. In M. Sherif & M. O. Wilson (Eds.), Group relations at the crossroads (pp. 232-256).

New York:

Harper.

Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-140.

Gerard, H. B., Wilhelmy, R. A., & Conolley, E. S.

(1968).

Conformity and group size. Journal of Personality and Social Psychology, 8, 79-82.

Gorenflo, D. W., & Crano, W. D. (1989). Judgmental subjectivity/objectivity and locus of choice in social comparison. Journal of Personality and Social Psychology, 57, 605-614.

Harkins, S. G., & Petty, R. E. (1983). Social context effects in persuasion: The effects of multiple sources and multiple targets. In P. Paulhus (Ed.), Basic group processes (pp. 149-175). New York:

Springer-Verlag.

Harkins, S. G., & Petty, R. E. (1987). Information utility and the multiple source effect. Journal of Personality and Social Psychology, 52, 260-268.

Hood, W. R., & Sherif, M. (1962). Verbal report and judgment of an unstructured stimulus. Journal of Psychology, 54, 121-130.

Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953).

Communication and persuasion. New Haven, CT:

Yale University Press.

Insko, C A., Smith, R. H., Alicke, M. D., Wade, J., & Taylor, S. (1985). Conformity and group size: The concern with being right and the concern with being liked. Personality and Social Psychology Bulletin, 11, 41-50.

Kelman, H. C. (1958). Compliance, identification, and internalization: Three processes of attitude change. Journal of Conflict Resolution, 2, 51-60.

Kumar, J. (1983). Conformity behavior as a function of confederates' age and size of confederate group.

Personality Study and Group Behavior, 3, 69-73.

Lasswell, H. D. (1948). The structure and function of communication in society. In L. Bryson (Ed.), Communication of ideas (pp. 37-51). New York:

Harper & Row.

Latane, B., & Wolfe, S. (1981). The social impact of majorities and minorities. Psychological Review, 88, 438-453.

Mackie, D. M. (1987). Systematic and nonsystematic processing of majority and minority persuasive communications. Journal of Personality and So- cial Psychology, 53, 41-52.

Milgram, S. (1974). Obedience to authority. New York: Harper & Row.

Milgram, S. (1977). The individual in a social world.

Reading, MA: Addison-Wesley.

Moscovici, S. (1974). Social influence: I. Conformity and social control. In C. Nemeth (Ed.), Social psychology: Classic and contemporary integra- tions (pp. 179-216). Chicago: Rand-McNally.

Moscovici, S. (1976). Social influence and social change. New York:

Academic Press.

Moscovici, S. (1980). Toward a theory of conversion behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13, pp.

209-239). New York:

Academic Press.

Moscovici, S. (1985a). Innovation and minority influence. In S. Moscovici, G. Mugny, & E. Van Avermaet (Eds.), Perspectives on minority influ- ence (pp. 9-52). Cambridge, England: Cambridge University Press.

Moscovici, S. (1985b). Social influence and confor- mity. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (Vol. 2, 3rd ed., pp.

347-412). New York:

Random House.

Moscovici, S., Lage, E., & Naffrechoux, M. (1969).

"Sleeper effect" and/or minority effect? Theoreti- cal and experimental study of delayed social influence. Cahier de Psychologie Cognitive, 1981, 199-221.

Moscovici, S., & Personnaz, B. (1980). Studies in social influence: V. Minority influence and conver- sion behavior in a perceptual task. Journal of Experimental Social Psychology, 16, 270-282.

Moscovici, S., & Personnaz, B. (1991). Studies in social influence: VI. Is Lenin orange or red?

Imagery and social influence. European Journal of Social Psychology, 21, 101-118.

Mugny, G., Butera, E, Sanchez-Mazas, M., & Pe*rez, J. A. (1995). Judgments in conflict. The conflict elaboration theory of social influence. In B.

Boothe & R. Hirsig (Eds.), Perception-evaluation- interpretation. Swiss monographs in psychology (Vol. 3, pp. 160-168). Goettingen, Germany:

Hogrefe & Huber.

Nail, P. R. (1986). Toward an integration of some models and theories of social response. Psychologi- cal Bulletin, 100, 190-206.

Perez, J. A., & Mugny, G. (1987). Paradoxical effects of categorization in minority influence: When being an out-group is an advantage. European Journal of Social Psychology, 17, 157-169.

P6rez, J. A., & Mugny, G. (1990). Minority influence:

Manifest discrimination and latent influence. In D.

Abrams & M. Hogg (Eds.), Social identity theory:

Constructive and critical advances (pp. 78-102).

London: Harvester Wheatsheaf. 80 CRANO Perez, J. A., & Mugny, G. (1996). The conflict elaboration theory of social influence. In E. H.

Witte & J. H. Davis (Ed.), Understanding group behavior (Vol. 2, pp. 191-210). Hillsdale, NJ:

Erlbaum.

Petty, R. E., & Cacioppo, J. T. (1986a). Communica- tion and persuasion: Central and peripheral routes to attitude change. New York:

Springer-Verlag.

Petty, R. E., & Cacioppo, J. T. (1986b). The elaboration likelihood model of persuasion. In L.

Berkowitz (Ed.)T Advances in experimental social psychology (Vol. 19, pp. 123-205). New York:

Academic Press.

Petty, R. E., & Krosnick, J. (1995). Attitude strength:

Antecedents and consequences. The Ohio State University series in attitudes and persuasion (Vol.

4).

Hillsdale, NJ: Erlbaum.

Sherif, M. (1935). A study of some social factors in perception. Archives of Psychology, 27(187), 1-60.

Sherif, M. (1936). The psychology of social norms.

New York: Harper & Row.

Sherif, M., Harvey, O. J., White, B. J., Hood, W.

R., & Sherif, C. W. (1961). Intergroup conflict and cooperation: The Robbers Cave experiment. Nor- man, OK: University of Oklahoma Book Ex- change.

Sherif, M., & Hovland, C. I. (1961). Social judgment.

New Haven, CT: Yale University Press.

Sivacek, J., & Crano, W.

D. (1982). Vested interest as a moderator of attitude-behavior consistency.

Journal of Personality and Social Psychology, 43, 210-221.

Tajfel, H. (1982). Social identity and intergroup relations. Cambridge, England: Cambridge Univer- sity Press.

Tajfel, H., & Turner, J. T. (1986). The social identity theory of intergroup behavior. In S. Worchel & W.

Austin (Eds.), Psychology of intergroup relations (pp.

7-24). Chicago: Nelson-Hall.

Tanford, S., & Penrod, S. (1984). Social influence model: A formal integration of research on majority and minority influence. Psychological Bulletin, 95, 189-225.

Triplett, N. (1898). The dynamogenic factors in pacemaking and competition. American Journal of Psychology, 9, 507-533.

Turner, J. C. (1991). Social influence.

Milton Keynes, England: Open University Press.

Wood, W. (in press). Attitude change: Persuasion and social influence. Annual Review of Psychology.

Wood, W., Lundgren, S., Ouellette, J. A., Busceme, S., & Blackstone, T. (1994). Minority influence: A meta-analytic review of social influence processes.

Psychological Bulletin, 115, 323-345.

Zajonc, R. B. (1965). Social facilitation. Science, 149, 269-274.

Zajonc, R. B. (1980). Comprescence. In P. B. Paulus (Ed.), Psychology of group influence (pp. 35-60).

Hillsdale, NJ: Erlbaum.