CULTURE AND THE STRUCTURE OF PERSONAL EXPERIENCE: INSIDER AND OUTSIDER PHENOMENOLOGIES OF THE SELF AND SOCIAL WORLD Dov ...
103 downloads
2559 Views
3MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
CULTURE AND THE STRUCTURE OF PERSONAL EXPERIENCE: INSIDER AND OUTSIDER PHENOMENOLOGIES OF THE SELF AND SOCIAL WORLD Dov Cohen Etsuko Hoshino‐Browne Angela K.‐y. Leung
This chapter argues for the importance of understanding the role of culture in structuring people’s personal phenomenological experience. Such an understanding is (1) important per se and (2) important for elucidating the feedback loops between culture and self, between macro‐level ideology and micro‐level experience. To illustrate, we contrast the ‘‘outsider’’ perspective on the self of Asian‐Americans with the ‘‘insider’’ perspective on the world for Euro‐Americans. We examine (1) the outsider versus insider perspective by looking at the phenomenology of memory imagery, online imagery, visualization and embodiment of narratives, and relational versus egocentric projection; (2) the implications for cultural diVerences in egocentric biases that derive from dwelling too much in one’s own internal experience; and (3) the emergence of developmental diVerences in characterizing the social world. We argue that the lessons of experience and cultural ideology cocreate each other, and we illustrate this by describing some ways that distinct phenomenological experiences are intimately tied to cultural norms, beliefs, and ideals.
I. Introduction To say that a person ‘‘sees herself through other people’s eyes’’ or ‘‘takes other people’s perspective’’ implies something about how people act in relation to others. Conversely, to say that someone is ‘‘self‐absorbed’’ or ‘‘lives in his own 1 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39001-6
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
2
COHEN, HOSHINO‐BROWNE, AND LEUNG
world’’ indicates something about people’s failure to think about others. The expressions are usually meant metaphorically, but they also have a more literal interpretation that implies something about the actual phenomenology that the person experiences. In the present chapter, we examine how people from diVerent subcultures may experience the self in ways that make a literal interpretation of these metaphors plausible; and in taking the metaphors literally, we try to address one of the important puzzles of cultural psychology, namely how micro‐level individual phenomenological experience and macro‐level cultural ideology can reinforce and recreate each other. Much research in cultural and cross‐cultural psychology has focused on content or the what of culture: What are the attitudes, beliefs, and values that diVerentiate between cultures? What are the contents of the independent versus interdependent self ? What are the scripts, norms, and expectations that drive behavior? On the other hand, a good deal of attention has been paid subsequently to the process or the how of culture: How do people process information about the social world diVerently (Kuhnen & Oyserman 2002; Nisbett, Peng, Choi, & Norenzayan, 2001)? How do people juggle multiple identities (Hong, Morris, Chiu, & Benet‐Martinez, 2000; Oyserman & Lee, 2007)? How do cultural practices get patterned by interpersonal interactions and aVordances in the environment (Cohen, 2001; Kitayama & Markus, 1999; Kitayama, Markus, Matsumoto, & Norasakkunkit, 1997; Miller & Goodnow, 1995)? In this chapter, we argue that studying people’s phenomenological experience of themselves in the world is another important route to understanding how psyches and cultures make each other up. To be a person in the world feels like something, and this feeling of the self in the world can arise from our cultural ideologies and in turn can help to recreate those ideologies. The perceptions, the imagery, the memories, the mental models, the salience (or lack of salience) of our own internal thoughts and physical sensations, and the (real or imagined) feeling of being in tune with another person all structure the feeling of being a self in the world. Here we focus on what might be called an ‘‘outsider’’ phenomenology of self versus an ‘‘insider’’ phenomenology. These are diVerences in the forms or structures of experience. In the ‘‘outsider’’ (or third person) form of experience, a person experiences himself or herself from the point of view of an outsider looking at the self. In the ‘‘insider’’ (or first person) form of experience, a person does not see himself or herself as others would; instead, the insider dwells in his or her own private, internal experiences and may end up either (1) projecting those experiences onto others or (2) mistaking the private, internal experience for something that is actually ‘‘out there’’ in the world. We argue that understanding the phenomenology of being a self in the social world is extremely important for understanding two diVerent
CULTURE AND PERSPECTIVE
3
subcultural systems. We examine this phenomenology for Asian‐Americans and Euro‐Americans and show how it manifests in predictable situations.1 As Nagel (1974) pointed out in a very diVerent context, one of the key features of consciousness is that consciousness feels like something. And similarly, being a cultural being in the world feels like something that is hard to reduce to a simple collection of beliefs, attitudes, or artifacts. It feels a certain way to be a self in one cultural system or another; and in itself, understanding these diVerences is an extremely important part of understanding culture.
II. A Sociofunctional Account of Perception A. HOW FORM IS CONTENT We also argue that the form or structure of our phenomenological experience is not generated haphazardly, but is systematically derived from the contents of our cultural beliefs and values and in turn aVects those beliefs and values. In this sense, the structure of experience—the form—is content. Embedded within the structure of our experience are cultural beliefs, attitudes, and values about the way the world is and should be. The content of our ideological conceptions shapes the phenomenological structure (the form) of our experience, and the form of our experience in turn leads us to certain beliefs about the world. In terms of outsider phenomenology, for example, the form of a third‐ person memory (as opposed to a first‐person memory) carries with it an implicit (content) understanding that the self is embedded in a social matrix where one is watched and watched over by others. Similarly, the form of having a third‐person online experience carries with it a tacit acknowledgment that we must be concerned with how others see us. Engaging in relational projection (for example, seeing others as showing contempt for us when we are feeling ashamed) carries with it an understanding that the self is interdependent with others and that emotions are essentially relational. At a very basic level, the very form of the third‐person experience gives us two messages: ‘‘think about how your actions look to other people’’ and ‘‘consider what the world looks like from other people’s points of view.’’ Behaving with propriety and achieving interpersonal harmony are the higher‐order cultural imperatives of these two basic messages. 1
We use the label ‘‘Asian‐Americans’’ as a shorthand way of saying North Americans whose ancestors came from East Asia. We use the label ‘‘European Americans’’ or ‘‘Euro‐Americans’’ as a shorthand way of saying North Americans whose ancestors came from Europe.
4
COHEN, HOSHINO‐BROWNE, AND LEUNG
On the other hand, consider what follows from the insider’s perspective. From the insider perspective, if I am so absorbed in my own internal experience that I do not consider others’ perspectives and do not understand that my perceptions of the world are simply perceptions from one viewpoint, this contains a license for a certain willful individualism. Further, if my experience is so dominated by my own internal thoughts and feelings that I cannot separate my internal experience from my perception of what is actually ‘‘out there’’ in the world, this too will create (1) an assuredness that I am right in my beliefs about the world, (2) an agentic readiness for action, and (3) a certain approach to forming relationships or alliances with others based on their assumed similarity to me. As we describe later, the insider perspective is well suited to fostering the mores of an individualistic culture and the types of side‐by‐side relationships that occur within it (Tocqueville, 2000). Thus, embedded in the structure of experience are a whole set of cultural beliefs about the self and how it should act in relation to other people. Though we may not realize it, the way we structure our phenomenological experience of reality (form) embodies a certain set of cultural attitudes, beliefs, and values (content). (And, as discussed later, because we often do not realize how the form embodies our beliefs and values, the implicit messages the form contains may be all the more powerful.) In the empirical section of the chapter, we examine how various aspects of experience can be similar or diVerent for Asian‐Americans and Euro‐ Americans. In the conclusion, we expand on what the consequences of some of these diVering experiences might be. And immediately below, we lay out the logic for why and in what situations Asian‐Americans and Euro‐Americans might have these diVerent sorts of phenomenological experiences.
B. PERCEPTION IN CULTURAL CONTEXT Our sociofunctional2 account of perception begins with the thought experiment: Suppose one had a human being B and could dictate the form (but not the content) of B’s phenomenological experience. How would one want B’s 2 Cultural traits and patterns are not always functional, of course. Thus, people hold superstitious, irrational beliefs (NemeroV & Rozin, 1989; Rozin, Markwith, & NemeroV, 1992; Rozin, Markwith, & Ross, 1990; Rozin, Millman, & NemeroV, 1986). They fail to pursue their goals optimally and sometimes engage in plainly self‐destructive actions (Edgerton, 1992). And some behavior patterns simply outlast their usefulness. [e.g., Fertility customs that were adaptive in agricultural societies persist even after a population has been urbanized and cities have become overcrowded (Triandis, 1994).] Nevertheless, in many cases, an assumption of functionality is a decent place to start when we examine cultural patterns.
CULTURE AND PERSPECTIVE
5
experience to be structured so that B could operate eVectively in a tight, interdependent milieu that valued face‐to‐face relationships? Conduct the same thought experiment with person C, who would operate in a more individualistic culture. How would one want C’s experience to be structured so that C could operate eVectively in a looser, more individualistic environment? It is our contention that B’s phenomenological experience would be structured such that B takes an outsider’s perspective on the self, whereas C’s experience would be structured more along the lines of an insider’s perspective, roughly speaking.
1. Tight and Loose Cultures In the case of Asian‐Americans and Euro‐Americans, it has been shown that Asian‐Americans tend to endorse more collectivistic ideologies whereas Euro‐Americans tend to endorse more individualistic ideologies—though the diVerence in explicit ideology may not be as great as cultural psychologists generally suppose (see Oyserman, Coon, & Kemmelmeier, 2002). That is, Asian‐Americans are more likely than Euro‐Americans to say that the self is defined by social roles, interpersonal duties, and in‐group–out‐group distinctions rather than by intrapsychic traits, attitudes, or personal expressivity. Research on Asian‐Americans and Euro‐Americans has tended to stress this broad diVerence, and it has often attributed Asian‐American and Euro‐American diVerences to diVerences in how individualist versus collectivist or independent versus interdependent the cultures generally are. However, if Asian‐American culture is more collectivistic than Euro‐ American culture, it is also tighter, in the sense that it has clear norms for proper social behavior and punishments for violations of these norms (Chan, Gelfand, Triandis, & Tzeng, 1996; Triandis, 1994; see also Heine, Lehman, Markus, & Kitayama’s fascinating contrast (1999) of Japanese and North American cultures). Generally speaking, tightness and collectivism tend to go together, but they are conceptually distinct; and in fact, Triandis (2004) views tightness–looseness as the more ‘‘basic’’ dimension from which other cultural syndromes (including collectivism) derive. Tightness plays an important role in deriving the hypotheses of this chapter (though as will be apparent, tightness and interdependence are bound together quite closely for the phenomena we describe). In many relatively tight cultures such as those originating in East Asia, it is important to maintain harmony and important not to personally stand out so that the self will not be a target of criticism (for failures) or sometimes envy and resentment (in the case of successes) (Gelfand et al., 2002; Kwan, Bond, & Singelis, 1997; Rothbaum, Pott, Azuma, Miyake, & Weisz, 2000).
6
COHEN, HOSHINO‐BROWNE, AND LEUNG
Negotiating a world where there are clear rules and where the opinions of others are quite important means that an individual needs to take stock of how he or she appears in another’s eyes (Ji, Schwarz, & Nisbett, 2000; Mesquita, 2001). It means one must in some sense look on the self as an outsider would. The East Asian conception of ‘‘face’’ (lian or mianzi in Chinese) is an apt metaphor for this because one’s face can only literally be seen from the outside. In contrast, in looser cultures where self‐esteem is valued and where the esteem of others is relatively less valued, it is far less necessary to take an outsider’s perspective on the self. Further, loose cultures definitionally allow individuals greater freedom, and thus they tend to put a premium on choice by the individual. In loose, independent cultures, the rational, agentic individual is supposed to be capable of (1) knowing what he or she wants and (2) going out and getting it (see Fig. 1; Campbell et al., 1996; Hoshino‐Browne et al., 2005; Iyengar & Lepper, 1999; Kitayama, DuVy, & Uchida, 2007; Morelli & Rothbaum, 2007; Schwartz, 2000; Schwartz et al., 2002). Finding and expressing our ‘‘authentic’’ self requires knowing ourselves and turning our focus to our internal states (Wang, 2004; Wang & Conway, 2004; Wang, Leichtman, & Davies, 2000). Our own thoughts and feelings become highly salient to us; and the focus of attention thus becomes more introspective as our own thoughts, feelings, and beliefs occupy consciousness. ‘‘Too much’’ introspection can easily slide into egocentrism. Attention is a limited resource, thus our focus on internal thoughts, feelings, and desires takes away from our ability to perceive the world (and perceive ourselves) from another’s perspective. Starting from early childhood, people get over their very crudest forms of egocentrism (believing that the rest of the world thinks, feels, and actually perceives the way they themselves do; Piaget & Inhelder, 1956). But it has been suggested that we are vulnerable to such egocentric thinking all our lives (Eibach, Libby, & Gilovich, 2003; Epley, Morewedge, & Keysar, 2004; Gilovich, Medvec, & Savitsky, 1998; Ross & Sicoly, 1979). The phenomenology of our own thoughts and feelings overpower us; we anchor on ourselves and then fail to adjust away when we think about what other people must think and feel (Epley, Keysar, Van Boven, & Gilovich, 2004; Gilovich et al., 1998). If living in a tight culture pushes us to take an outsider’s perspective on the self, then living in a looser one pushes us to take an insider’s perspective on the social world, leading us to dwell so much in our own thoughts that we eVectively become naı¨ve realists and also project our own internal states and feelings onto the external world. A loose, independent culture will not punish the egocentrism that comes from an insider’s perspective as harshly as would a tight culture that valued
CULTURE AND PERSPECTIVE
7
Phenomenological gestalt of the outsider’s perspective on the self, self as seen by others [Conducive to cultural imperatives of (1) behave with propriety and circumspection and (2) pursue harmony] • • • •
Online third person imagery. Third person memories. Generalized other looking at the self. Mental models of the self as seen by others; mental models constructed and embodied from the perspective of other people. • Simultaneous attention to one’s private thoughts and public appearances highlights disjunction between the two. • Relatively more attention focused on other people’s unobservable thoughts and feelings. • Attention to others potentially leading to empathic accuracy.
Phenomenological gestalt of the insider’s perspective on the world, self projected onto the world [Conducive to cultural imperatives of (1) know what you want and (2) go out and get it]
• • • •
Online first person imagery. First person memories. Mental models constructed and embodied from egocentric perspective. Attention to stimuli in one’s own head leads to confusing what is in one’s own head with what is “out there” in the world. • Projection of internal states onto other persons. • Feelings of knowing, empathy resulting from this projection. • Relative inattention to others’ states and characterization of others in terms of behavior immediately visible to the self. Fig. 1. Sociofunctional approach to perception. The cultural imperatives of tight versus loose cultures lead to diVerent phenomenological gestalts.
interpersonal harmony and individual circumspection. In fact, as we note later, the insider’s perspective can foster the willful individualism and sense of personal agency that can lead to success in a loose culture.
8
COHEN, HOSHINO‐BROWNE, AND LEUNG
C. THE IMPARTIAL SPECTATOR AND THE GENERALIZED OTHER Why is it that the look of another person looking at you is diVerent from everything else in the Cosmos? Walker Percy (1983) in Lost in the Cosmos O would some Power the gift to give us To see ourselves as others see us! It would from many a blunder free us, And foolish notion Robert Burns from To a Louse: On Seeing One on a Lady’s Bonnet at Church
Scholars (from the West) such as Adam Smith, Charles Cooley, and George Herbert Mead have discussed the notion of the ‘‘impartial spectator,’’ ‘‘the looking glass self,’’ and ‘‘the generalized other.’’3 In their conception, the self comes into being because it imagines how others would see it. As Smith (1759) wrote: Were it possible that a human creature could grow up to manhood in some solitary place, without any communication with his own species, he could no more think of his own character, of the propriety or demerit of his own sentiments and conduct, of the beauty or deformity of his own mind, than of the beauty or deformity of his own face. All these are objects which he cannot easily see, which naturally he does not look at, and with regard to which he is provided with no mirror which can present them to his view. Bring him into society, and he is immediately provided with the mirror which he wanted before. It is placed in the countenance and behaviour of those he lives with, which always mark when they enter into, and when they disapprove of his sentiments; and it is here that he first views the propriety and impropriety of his own passions, the beauty and deformity of his own mind.
The ‘‘generalized other’’ is a universal feature of the socialized human psyche. Mead (1934) and Smith (1759) in particular argued that the internalization of the ‘‘generalized other’s’’ perspective is the basis for social control. And in contemporary psychological work, researchers such as Baldwin and Holmes (1987), Diener and Wallbom (1977), Duval and Wicklund (1972), Fredrickson, Roberts, Noll, Quinn, and Twenge (1998), and Scheier and Carver (1977), have discussed phenomena such as self‐objectification, objective self‐awareness, and relational schemata that create private audiences. The contention in this chapter is thus not 3
Sartre’s notions of Being‐for‐others and Being‐for‐itself (versus Being‐in‐itself) are relevant to ideas about consciousness reflecting on itself, as are other existentialist notions. However, we avoid this terminology for the most part because of the ethical baggage Sartre’s terms entail (Sartre, 1965). We use the phrase‘‘being in the world’’ but not ‘‘Being‐in‐the‐world’’ to avoid the philosophical baggage associated with the latter expression (Heidegger, 1962).
CULTURE AND PERSPECTIVE
9
that the ‘‘impartial spectator’’ and the ‘‘generalized other’’ are important for Asian‐Americans but not for Euro‐Americans. Rather, the contention is that what Mead and Smith have said about the ‘‘generalized other’’ is more true for Asian‐Americans than it is for Euro‐Americans. That is, the self is more likely constructed and experienced for Asian‐Americans as something that is the object of gaze from another (Kitayama & Markus, 1999). Heine et al. (1999, p. 773), talking about the Japanese, argued that ‘‘rather than being seen as subjects, they may more aptly be viewed as imagined objects in the eyes of others.’’ In fact, Kitayama, Snibbe, Markus, and Suzuki (2004) have suggestive evidence that this imagined gaze of the other is what makes a situation psychologically real for Japanese participants. In a remarkable series of studies, they replicated Heine and Lehman’s finding (1997a) that Japanese participants do not show ‘‘normal’’ cognitive dissonance (rationalization) eVects after making a choice. However, they extended their work by showing that putting Japanese participants in the same room as a poster with simple line drawings of faces brings the dissonance eVects out in full force. The line drawings of ‘‘generalized others’’ looking at the participant give choices a real psychological impact; and now with the evaluative weight of imagined others looking on, ‘‘normal’’ dissonance eVects emerged among the Japanese.
D. PLAN FOR THE PAPER 1. The Outsider Perspective In Sections III to VI of this chapter, we try to take the metaphor of the outsider’s perspective on the self and show how it literally becomes instantiated in phenomenological experience. Thus, we flesh out how Asian‐Americans are more likely than Euro‐Americans to literally take an outsider’s perspective on themselves by examining the phenomenology of memories, online imagery, visualization, and relational projection. Importantly, we argue that the function of this outsider’s perspective is essentially social—in a tight society, it is to keep one from standing out. It gives one the circumspection needed to make sure one is following social norms. Thus, we would expect Asian‐Americans to take this outsider perspective (more than Euro‐Americans do) particularly in social situations or in situations where one is at the center of other people’s attention. Consistent with this functional interpretation, we would not expect an across‐the‐board diVerence such that Asian‐Americans would take more of an outsider’s perspective on themselves in all situations. Returning to our functionalist thought experiment: the ideal design for the structure of consciousness would involve flexibility so that a person might switch back and forth between a third‐person (outsider) experience and a
10
COHEN, HOSHINO‐BROWNE, AND LEUNG
first‐person (insider) experience, depending on the demands of the situation. The empirical studies below highlight this point, showing that Asian‐Americans are just as likely as Euro‐Americans to be in the ‘‘insider’’ mode for situations when they would not be the center of attention (Section IV), when others are not made salient (Section V), and when the situation is not social (Sections VI and X). 2. The Insider Perspective By taking the perspective of a spectator, Smith (1759) argued, ‘‘I divide myself, as it were, into two persons . . . I, the examiner and judge, represent a diVerent character from that other I, the person whose conduct is examined into and judged of.’’ Smith hypothesized that without being able to do such splitting, it would be extremely diYcult to separate oneself from one’s passions. In the continuation of the paragraph above, he argued: ‘‘To a man who from his birth was a stranger to society, the objects of his passions, the external bodies which either pleased or hurt him, would occupy his whole attention.’’ Not needing to control his behavior, this man would become absorbed with the thoughts and feelings developed from his own insider’s perspective. In Sections VII through IX, we examined this issue by looking at how taking too much of an ‘‘insider’s’’ perspective can distort one’s perceptions of the external world. A person’s inability to get outside of his or her own head makes him or her especially vulnerable to egocentric biases that can arise from dwelling too much in one’s own phenomenological experience. That is, when a person’s own thoughts and feelings are so salient, so overpowering, and occupy so much of their consciousness, these thoughts and feelings have great power in coloring a person’s judgments about either (1) what must be out there in the world or (2) what other people must be thinking or feeling. In fascinating research, Gilovich et al. (1998) and Vorauer and Ross (1999) have described some of these egocentric biases as arising from essentially an ‘‘anchoring and adjustment’’ process in which we come to anchor on our own thoughts and then adjust away from them insuYciently in thinking about what others must feel or in thinking about how our perceptions and ‘‘true’’ reality might diVer. The adjustment tends to be too close to the anchor (the self ’s experience), and sometimes adjustment does not occur at all. Thus, the confusion between what is in our own heads, what is in other people’s heads, and what is ‘‘out there’’ in reality is essentially an insider bias. And therefore, conversely, taking an ‘‘outsider’s’’ perspective on the self should allow one to get out of one’s own head enough so that one might avoid egocentric errors that derive from self‐absorbed thought. In Sections VII through IX, we examined various sorts of egocentric errors and illusions that those with insiders’ perspectives might fall prey to and those with outsiders’ perspectives might avoid.
CULTURE AND PERSPECTIVE
11
3. Characterizing the World Finally, we examine the issue of seeing through others’ perspectives in a developmental study (Section X). With Asian‐American and Euro‐American elementary school children, we examine how these children learn to describe the world. Our cross‐sectional study aims to assess how, as children age, they learn to either characterize the world in terms of actions they can see with their own eyes versus characterize it in terms of the (invisible) internal thoughts and feelings of others. 4. Phenomenological Gestalts The sociofunctional approach to perception thus argues that tight versus loose cultures lead to diVerent types of phenomenological gestalts. As noted in Fig. 1, these gestalts encompass a number of modalities (memory imagery, online imagery, mental models, embodied cognitions, feelings, and perceptions of the world and other people). In the case of tight cultures, the phenomenological gestalt of the ‘‘outsider’s perspective’’ serves the need for circumspection, whereas in loose cultures, the phenomenological gestalt of the ‘‘insider’s perspective’’ on the world facilitates the pursuit of individual goals and self‐ expression. We use the word ‘‘gestalts’’ to indicate that these perspectives are pervasive ways of being in the world that go beyond an isolated eVect in this or that realm. Rather, the eVects fit together in a coherent way such that they create (and are created by) a worldview that is necessary to operating in a tight, interdependent culture that values harmony or a loose, independent one that prizes individual selves. E. NOTES ON TERMINOLOGY AND CLAIMS 1. A Note on ‘‘Insider’’ and ‘‘Outsider’’ It is important to note that when we say ‘‘outsider’’ or ‘‘insider’’ perspective, we are using this as a shorthand way of saying ‘‘more outsider’’ or ‘‘more insider.’’ In most cases, it would seem relatively rare for someone to have a complete outsider’s perspective on themselves without any consciousness of their own internal sensations. Conversely, it would also seem relatively rare to have a complete insider’s perspective, without any hint of self‐consciousness or reflective thought.4 4 We do not deny that there may be complete cases of insider or outsider experiencing in some instances: For example, extreme outsider experiences might include some cases (1) where there is neurological injury that prevents sensation (Gallagher, 2004), (2) complete out of body sensations perhaps produced by stimulation of the right angular gyrus (Blanke et al., 2002), or
12
COHEN, HOSHINO‐BROWNE, AND LEUNG
2. A Note on Participants The point of this chapter is to illustrate the usefulness of a sociofunctional approach to perception by describing the phenomenological gestalts (the insider and outsider perspectives) of people in tight versus loose cultures. Thus, we compare people from Euro‐American and Asian‐American cultures. Euro‐American culture is typically described as relatively independent and loose, whereas Asian‐American culture is typically described as relatively interdependent and tight (Triandis, 1994). Of course, Asian‐American culture may be a relatively tight culture for at least two reasons. First, East Asian culture has generally been described as interdependent and tight, and Asian‐Americans may have preserved those traditions in North America. Second, Asian‐American culture may be tight, because Asian‐Americans occupy a minority status and sometimes an immigrant status in North America, and this status may produce a tight, circumspective culture (see also Dubois, 1903).5 The reasons why Asian‐Americans have a tight culture are neither here nor there for the purposes of the present chapter; rather, the tight culture of Asian‐Americans is merely taken as a (3) cases of extreme alexithymia that inhibit the feeling of emotion. On the other end, there may be cases that totally preclude self‐consciousness or reflective thought such as extreme cases of flow (Csikszentmihalyi, 1990), deprivation (the pangs of hunger may be the only thing felt by someone deprived of food), or passion. However, for these examples, it is not clear (1) that the ‘‘outsider’’ experiences would be accompanied by an attention to others in the external world and (2) that a lack of self‐consciousness will produce the sorts of projections and egocentric errors we describe in Sections VII through IX. 5 One could attempt to tease apart the eVects of minority or immigrant status by, for example, comparing Asian‐Americans with (1) Asians in Asia (though there would still be the self‐ selection bias that people who choose to immigrate may be systematically diVerent than those who do not), (2) Euro‐Americans in Asia (though again there is self‐selection bias in the choice to leave one’s country), (3) Asian‐Americans living in predominantly Asian‐American communities versus living in Euro‐American communities (though there is self‐selection bias in choosing where to live), (4) other immigrant or minority groups in North America (though no other nonwhite minority group has had as much mainstream economic and educational success as Asian Americans have), and so on. Even if one were prepared to accept the self‐selection and confound issues above, for studies that involved running participants at more than one site, there would still be possibilities of operational confounds in language, modes of administration, equipment and materials, personnel, and so on. In contrast, running all participants at one site allowed us to standardize operations (and importantly, standardize language) across participants. Finally, running participants at one site also allowed us to roughly equalize participants in terms of standard of living, ability and background knowledge, and so on. Ultimately, through converging evidence across various methods, one might indeed estimate how much of an eVect was due to Asian‐Americans having minority or immigrant status. However, as noted in the text, this is not the concern of the present chapter. We studied Asian‐Americans because Asian‐American culture is a tight culture. Why they have a tight culture is not the point here.
CULTURE AND PERSPECTIVE
13
starting point for contrasting the phenomenological gestalts of people from relatively tight versus relatively loose (Euro‐American) cultures. Also, it is worth adding that the labels of ‘‘Asian‐American’’ and ‘‘Euro‐ American’’ cover a massive degree of within‐group variation (Sanchez‐Burks, 2002; Snibbe & Markus, 2005; Triandis, 1994). As a label, ‘‘Asian‐Americans’’ refers to Americans whose ancestors come from East and Southeast Asian countries as diverse as Japan and Thailand and that collectively have 1.5 billion people. ‘‘Euro‐Americans’’ refers to Americans whose ancestors come from both northern and southern Europe and eastern and western Europe. Simply using the label ‘‘American’’ also collapses over a diverse group of people because we are using this as a shorthand for North American, as these studies were run in Canada and the United States—countries that have their own rich collections of regional subcultures (Levine, Martinez, Brase, & Sorenson, 1994; Plaut, Markus, & Lachman, 2002; Vandello & Cohen, 1999). And, of course, there is a great deal of variation due to individual diVerences (Cohen, 2007; Rozin, 2003). Nevertheless, we think that there are commonalities that make it reasonable to talk about an Asian‐American subculture and a Euro‐ American subculture that may be diVerent in some ways. Systematic variation within these subcultures due to the specific country of ancestral origin or socioeconomic status or age is a topic for further study (see Vandello & Cohen, 2003). In the present chapter, to help quantify the magnitude of between‐group and within‐group diVerences, we report the eVect size f for interactions. [Conventionally, fs in the range of 0.1, 0.25, and 0.4 are considered small, medium, and large eVects, respectively (Rosenthal & Rosnow, 1991).]
III. Relational Versus Egocentric Projection6 In his book, The Emotions, Sartre (1948) sketched out a ‘‘phenomenological theory’’ and described emotions as ‘‘a magic transformation of the world.’’ Our perceptions of the world change, depending on the emotional state we are in. Whether the world seems gloomy or wondrous, whether another’s grin was friendly or mocking, and whether a touch was playful or hostile, all depend greatly on the mood we are in when we perceive them. Psychologists have examined one way that emotions transform our perceptions when they have studied ‘‘projection.’’ Yet it is essential to understand that the label ‘‘projection’’ can embrace at least two very diVerent sorts 6 Because studies from Sections III and IV have been reported in more detail in Cohen and Gunz (2002), we give only brief methodological details for these two sections here.
14
COHEN, HOSHINO‐BROWNE, AND LEUNG
of phenomena. The first and most frequently studied type is what might be called classical or egocentric projection in which people take their own feelings and project them onto others (e.g., an angry person sees others as angry, a sad person sees others as sad, and so on). For a person experiencing emotion X, the faces of other people are magically transformed into expressions of X. However, there is a second type of projection that might be called relational projection. In relational projection, a person does not project their own feelings; rather, they project onto others the feelings that the generalized other would have in looking at them. That is, certain emotions come in complementary pairs in that they naturally co‐occur in social situations. Three of these pairs are: shame and contempt, anger and fear, sadness and sympathy: (1) A person who feels ashamed thinks that others are looking or would look at her with contempt; conversely, a person who feels contempt for another thinks that this person should feel ashamed of herself. (2) A person who feels afraid of another fears what this person might do in anger; conversely, a person who feels anger may do (or at least want to do) something that would cause the other to feel fear. Finally, (3) a person who feels sympathy for another experiences this emotion for someone who feels sad; conversely, a person who feels sad may expect (or want or anticipate) another’s sympathy. In the phenomenon of egocentric projection, a person would simply take their own internal emotions and project those same emotions onto others. However, in relational projection, one’s own emotions are a guide to what others are thinking—more particularly, they are a guide to how the generalized other would look at the self. Note that both types of projections are biases. They can lead one to correct predictions, but they can also lead to errors. The errors in one case simply derive from taking one’s insider’s perspective and projecting it onto the social world, whereas the errors in the other case derive from inappropriately projecting a generalized other’s outsider perspective. In Cohen and Gunz (2002), we induced Asian‐American and Euro‐ American participants to feel one of six emotions (randomly assigned). That is, supposedly as part of our memory study (see next section), some of our experimental participants were asked to write in detail about a time when they felt angry, afraid of another person, sad, sympathetic to another, ashamed of themselves, or contemptuous of another person. This assignment to emotional condition was a between‐subjects variable. After this ‘‘memory’’ study was completed, participants were told that we also needed their help rating pictures for an unrelated task. The pictures were 24 photographs showing a Euro‐American male, Euro‐American female, Asian‐American male, and Asian‐American female in various poses as well as images from 6 paintings, and participants were asked to rate the
CULTURE AND PERSPECTIVE
15
pictures for the emotions they displayed. Each participant’s egocentric projection score was computed as the average rating of emotion X when he or she had been induced to feel X himself or herself (e.g., the egocentric score was an average of how much a participant in the anger condition saw anger, how much a subject in the sad condition saw sadness, and so on). Each participant’s relational projection score was computed as the average rating of the emotion that would have been appropriate for the generalized other looking at the participant in the participant’s (induced) emotional state. For example, the relational projection score was an average of how much a participant in the shame condition saw contempt in another’s expression, how much a participant in the fear condition saw anger in another’s expression, how much a sad person saw sympathy in another’s expression, and so on. Figure 2 shows the significant ethnicity type of projection interaction, collapsing the results across mood induction condition. Consistent with taking an ‘‘insider’s’’ perspective on the social world, Euro‐American participants took their own emotions and projected them onto others, seeing in the faces the emotions that they themselves felt. However, consistent with an ‘‘outsider’’ perspective on the self, Asian‐Americans engaged in relational projection seeing in the faces the emotions appropriate for the generalized other looking at them. Again, Euro‐Americans and Asian‐Americans both transformed their perceptions of others based on their own emotions, but the former group took the internal emotions and projected them outward whereas the latter group took the perspective of the outsider looking in.
Projection scores
3.0
2.9
2.8
2.7
2.6 Euro-Americans Egocentric projection
Asian-Americans Neutral
Relational projection
Fig. 2. Egocentric and relational projection scores among Euro‐Americans and Asian‐ Americans.
16
COHEN, HOSHINO‐BROWNE, AND LEUNG
IV. Memory Imagery In terms of insider and outsider imagery, one useful area of research is memory. Nigro and Neisser (1983) have distinguished between field and observer memories. In the former, a person has the imagery of what they saw (or believed they saw) at the time. In the latter, a person takes the perspective of an outsider looking in on the scene—that is, this observer might see what the person saw, but the observer would also see the person in the scene. Thus, in an observer memory, a person would see himself or herself in the scene just as an outsider would. The experience of having first‐person imagery or third‐person imagery is relatively common in dreams, but Nigro and Neisser showed that memory imagery could be distinguished along these lines as well. In our memory study, we asked participants to think of specific incidents. We used the incidents from Nigro and Neisser (1983) and added some of our own to create five memories in which the participant would be at the center of attention in the scene and five memories in which participants would not be at the center of attention in the scene. This distinction is crucial because the essential function of the outsider perspective for Asian‐Americans is to create the circumspection necessary for correct behavior when under the scrutiny or in the gaze of others. It is not that Asian‐Americans always adopt the outsider perspective; rather, it is that the outsider’s perspective is a functional switching of perspectives that facilitates correct social behavior when one is the focus of attention. Examples of scenes in which the participant would be at the center of attention were scenes where the participant was giving an individual presentation or was being embarrassed; examples of scenes in which participants would not be at the center of attention were scenes where the participant was running for exercise or watching a scary movie. After participants recalled the 10 specific incidents, we described the terminology of first‐ and third‐person memories and for each incident, respondents rated the memories for their first‐person versus third‐person content, following Nigro and Neisser (1983). As may be seen in Fig. 3, the ethnicity center of scene interaction was significant. For scenes where the respondent would be at the center of attention, Asian‐Americans were more likely to report third‐person memories, both in comparison to Euro‐Americans and compared to scenes where the respondent would not be at the center of attention. These eVects held controlling for the memories’ vividness, emotion, and distance in the past. For scenes in which they were not the center of attention, Asian‐Americans were actually significantly less
CULTURE AND PERSPECTIVE
17
Memory imagery scores
6.5
6.0
5.5
5.0
4.5 Euro-Americans Participant not at center of attention
Asian-Americans
Participant at center of attention
Fig. 3. Memories for situations in which Euro‐Americans and Asian‐Americans were either at the center of attention or not at the center of attention. Higher numbers indicate more third‐ person imagery.
likely to have third‐person memories compared to Euro‐Americans, perhaps due to a loss of self‐consciousness in these types of situations (Weber, 1951). Finally, we also asked participants about what might be called ‘‘collective memories,’’ that is memories deriving from stories that are told and retold among family and friends. We asked respondents to think of four collective memories that they were not present for themselves. (Examples given included stories that were told and retold about the time dad went through the car wash with the windows down or about the time mom met Elvis, and so on.) Respondents were asked about the imagery associated with each of their collective memories and were asked to classify the memories into three categories. The first sort of imagery was imagery that was as if the respondent took the perspective of the friend or family member and experienced the same things that the friend or family member did. The second sort of imagery was imagery that was as if the respondent saw what happened to the friend or family member as an outsider looking onto the scene would. The third sort of imagery was imagery that was simply of the main character just retelling the story to the respondent. The first sort of imagery (where it was as if the respondent had experienced what the friend or family member did) was considered the highest form of empathic imagery, and it was found that Asian‐Americans were in fact more likely to enter into the scene from the friend or family member’s perspective. There was a main eVect of
18
COHEN, HOSHINO‐BROWNE, AND LEUNG
ethnicity here, though it was qualified by an interaction with gender, indicating that the eVect was driven by male participants. Although this interaction with gender creates some ambiguity, the findings for males at least complement the findings on memories involving the self. For scenes where they are at the center of attention, Asian‐Americans are more likely than Euro‐ Americans to have third‐person ‘‘outsider’’ imagery; however, for collective memories with scenes involving friends and family, Asian‐Americans are more likely to enter into the perspective of the other and take an ‘‘insider’s’’ perspective on that other’s experience.
V. Online Imagery As noted above, dreams and memories are places where people commonly report imagery in which they see themselves from an ‘‘outsider’s’’ point of view. However, this phenomenon can also exist in online experience (Blanke, Ortigue, Landis, & Seeck, 2002). That is, a person can have a third‐person experiencing of the event as it is actually happening in real time. Nigro and Neisser (1983), in fact, suggest that some third‐person memories may derive directly from such third‐person experiencing. They speculated that: it is also possible to have observer experiences. Both of us (the authors) can attest to the possibility of experiencing events from a ‘detached’ perspective as they occur. In such instances we are conscious of how the entire scene would appear (or does appear in fact) to an onlooker who sees us as well as our surroundings. It is not clear how these experiences are best interpreted—whether as a nonegocentric form of direct perception in Gibson’s (1979) sense or as the products of instantaneous reconstruction—but it is clear they exist (pp. 468–469).
Smith (1759) argued that taking the outsider’s perspective was the way one achieved self‐control and distanced oneself from one’s passions (see also Fujita, Trope, Liberman, & Levin‐Sagi, 2006; Libby & Eibach, 2003). If this is true, then it should be that such third‐person distancing would be especially likely when there were disturbing or unpleasant emotions or sensations to be contained. Section V thus took place as a putative study of ‘‘mental toughness’’ in which participants’ mental toughness was supposedly indexed by how long they could endure in a (physical) pain tolerance task. Before participating in this pain tolerance task, all participants were run through one of three experimental conditions. The first condition was designed to highlight the salience of others to one’s interdependent self. The second was designed to highlight the individualistic goals of an independent self.
CULTURE AND PERSPECTIVE
19
And the third was a control condition that did not involve a prime of any kind. The prediction was that as they endured the pain tolerance task, third‐person distancing or an ‘‘observer’’ online experience would be particularly prevalent for Asian‐Americans in the interdependent self condition. The evocation of the social group by the interdependent prime should encourage Asian‐ Americans to take an outsider’s perspective on themselves during the painful mental toughness task. Method. Participants were 48 Euro‐Americans and 62 Asian‐Americans. Participants in the interdependent prime condition were asked to bring in a picture of themselves with their family and when they arrived at the laboratory, they wrote a short essay on what it means to be a good son or daughter. Participants in the independent prime condition were asked to bring a picture of themselves with no one else in the photo, and they wrote a short essay on what it means to ‘‘be yourself ’’ by considering their special uniqueness (see also Trafimow, Triandis, & Goto, 1991). Participants in the control condition were not asked to bring in a photo and did not write an essay. After writing (or not writing) the short essay, participants were told they would be given a pain endurance task as a test of mental toughness. The task was presented such that enduring the pain was a matter of will or self‐control rather than physical sensitivity. The experimenter put the participant’s finger in the pain device and lowered a weight on top of it (for a more extensive description of the device and its pain‐inducing properties see Eastwood, Gaskovski, & Bowers, 1998). When the pain task ended, the experimenter immediately removed the participant’s finger from the pain device, and the participant was given the Cognitive Coping Strategy Inventory (Butler, Damarin, Beaulieu, Schwebel, & Thorn, 1989) and told to fill it out with reference to how he or she had just coped during the pain task. The inventory has 70 items, 7 of which were relevant to the use of third‐person distancing. These items included statements such as ‘‘I might attempt to imagine myself leaving my body and observing my pain in an impartial, detached manner,’’ ‘‘I might begin thinking about my pain as if I were conducting an experiment or writing a biology report,’’ ‘‘I might attend to the pain in much the same way that a sports announcer or reporter would describe an event,’’ ‘‘I might try and imagine myself ‘floating oV ’ away from the pain, but still realize that my body hurts,’’ and so on. Results. As may be seen in Fig. 4, the group that made most use of third‐ person distancing was the Asian‐American group in the interdependent prime condition. This was true both when this condition was contrasted with all five other conditions [t(110) ¼ 2.22, p < .03, eVect size f ¼ .21] as
20
COHEN, HOSHINO‐BROWNE, AND LEUNG
2.1
Third-person imagery
2.0
1.9
1.8
1.7
1.6 Euro-Americans No prime
Asian-Americans Individual prime
Collective prime
Fig. 4. Use of third‐person imagery during a self‐control task by Euro‐Americans and Asian‐Americans in individualism, collectivism, and no prime conditions. Higher numbers indicate more third‐person imagery.
well as when it was contrasted only with the Asian‐American independence and control conditions ( p < .02, eVect size f ¼ .26) or with the Euro‐Americans ( p < .06, eVect size f ¼ .21). Thus, the sort of observer perspective on the self that Asian‐Americans reported in the memories task also occurred in this online self‐control task. This third‐person ‘‘observer’’ imagery occurred when important others had been made salient to our Asian‐American participants through the interdependence prime. It did not occur in the independence prime or the control conditions, just as it did not occur in the memory study when the self was not the focus of attention in a scene. Thus, here too, third‐person imagery of an outsider’s perspective on the self was selectively recruited when it might usefully simulate the gaze of real or implied others. For our Euro‐American participants, none of the experimental conditions were particularly likely to produce third‐person imagery.
CULTURE AND PERSPECTIVE
21
VI. Mental Models of the Self and Others in Narrative In both Sections IV and V above, participants self‐reported whether they took a third‐person perspective on themselves in their memories and online imagery. It is hard to imagine how social desirability could create the observed interactions. And it is hard to imagine that participants cannot accurately report on the imagery in their own heads. Nevertheless, it would be good to have corroborating evidence from a third and diVerent method of assessing third‐person versus first‐person perspective. In research on mental models that people use to comprehend narratives, Bower, Morrow, and colleagues have developed various techniques that can be used to judge a participant’s point of view (Bower & Morrow, 1990). We chose a technique that infers the perspective a person takes in their mental model by measuring their reaction times as they read stories involving deictic terms. As Black, Turner, and Bower (1979, p. 188) wrote, ‘‘The term deixis (which is Greek for ‘pointing’) refers to the orientational features of language which are relative to the time and place of the utterance,’’ and such deictic terms establish a point of view for one’s mental model. The words come and go (along with words such as bring and take) are deictic words in that they establish whether actions are moving toward us (come, bring) or away from us (go, take). Once a person reading a story has a point of view, he or she expects that point of view to be maintained. Thus, a sentence that reads ‘‘Ted is enjoying his lunch when Sam comes to join him’’ is perfectly intelligible if, as readers, we have been taking Ted’s point of view. If, as readers, our perspective is the same as Ted’s perspective, what comes toward him also comes toward us. It would still make sense if we read that ‘‘Ted is enjoying his lunch when Sam goes toward him,’’ but this sentence requires a little more processing time because it implies a diVerent ‘‘camera angle’’ and we have to shift from Ted’s perspective to a more third‐ party perspective or to Sam’s perspective. This extra processing to change camera angles will slightly slowdown reading speed; and thus understanding how deictic words aVect reading speed will tell us something about the perspective in the mental models that people have already constructed in their heads. In sum, if a reader is identifying with a character and taking his or her perspective, sentences where another person or object comes toward this character will be processed faster than those where this other person or object goes toward the character. By contrast, if a reader is not taking the main character’s perspective and is instead taking the other person’s perspective or taking the perspective of an outside observer, then sentences where the other person goes toward the main character will be processed faster than those where the other comes toward the main character. In the
22
COHEN, HOSHINO‐BROWNE, AND LEUNG
study below, we examined how people constructed mental models when the self was the main character, relative to how they constructed such models when others (in this case, a close friend) were the main character. Participants read eight diVerent stories. In half of these stories, the main character was the participant. In the other half of the stories, the main character was a person who participants had identified as a friend they were very close to. Further, half the stories were social in content in that they involved the main character interacting with other people; the other half were nonsocial and did not involve the main character interacting with other people. The prediction was that Asian‐Americans would be more likely to take a third‐person perspective when the main character was the self rather than a close friend. (Or stated in terms of a first‐person perspective, Asian‐Americans would be more likely to take a first‐person perspective when the main character was a friend rather than the self.) For the Euro‐Americans, the prediction was the opposite: Euro‐Americans should be relatively more likely to take a first‐person perspective when the main character was the self rather than a friend. Further, this interaction was expected to hold for the social (as opposed to the nonsocial) stories (consistent with the eVects in both Sections IV and V). Method. Sixty‐six Euro‐Americans and 65 Asian‐Americans were invited to the laboratory and told that researchers were examining how narratives are processed. They were informed that they would read eight stories. The stories would come on the screen a sentence at a time (or if the sentences were long, a clause at a time). When readers were ready to go from one sentence to the next, they were to press a key on a keypad.7 Participants were told that before they read a given story, they would either be instructed to imagine the narrative involved them or to imagine the narrative involved a close friend. (The participant was to pick and write down the name of one male and one female friend who could serve as main characters for the latter narratives.) Two target sentences were embedded in each story with a number of ‘‘filler’’ sentences that did not involve deictic words. For each story, one target sentence stated that another person (in the social story) or an inanimate object (such as a wave or a racquetball in the nonsocial story) comes toward the main character. The other target sentence stated that another person or object goes toward the main character. Results. Reaction time data were log transformed, and the average reaction time for processing the ‘‘come’’ sentences was subtracted from that for processing the ‘‘go’’ sentences, so that higher numbers indicated a more 7 Participants were not told that reaction times for how long it took to read each sentence were being measured.
23
CULTURE AND PERSPECTIVE
Social stories RT of “go” relative to RT of “come”
0.10 0.08 0.06 0.04 0.02 0 Euro-Americans
Asian-Americans
Self as story subject
Friend as story subject
Nonsocial stories RT of “go” relative to RT of “come”
0.04 0.02 0 −0.02 −0.04 −0.06 Euro-Americans Self as story subject
Asian-Americans Friend as story subject
Fig. 5. Reaction times (RT) for Euro‐Americans and Asian‐Americans for stories involving themselves or a friend in social or nonsocial situations. Higher numbers indicate a more first‐ person perspective in the mental model. Note: The Y‐axis represents the log‐transformed reaction times for go minus come. Larger numbers thus imply a more first‐person perspective, that is, the action comes toward (rather than goes toward) the story’s subject.
first‐person perspective. As predicted, there was a significant three‐way interaction involving ethnicity social/nonsocial content main character, F(1,121) ¼ 4.91, p < .03 eVect size f ¼ .2. As shown in Fig. 5, for the social stories, Asian‐Americans were relatively more likely to take a third‐person perspective when the main character was the self rather than a friend, whereas Euro‐Americans were relatively more likely to take a third‐person perspective when the main character was a friend rather than the self.
24
COHEN, HOSHINO‐BROWNE, AND LEUNG
(Or stated conversely, Asian‐Americans were more likely to take a first‐ person perspective when the main character was a friend rather than the self, whereas Euro‐Americans were more likely to take a first‐person perspective when the main character was the self rather than a friend.) This pattern held only for the social stories (top panel of Fig. 5), and it reversed for nonsocial stories (bottom panel of Fig. 5).8
A. FURTHER RESEARCH ON MENTAL MODELS 1. Spontaneous Construction of Sentences Further work on mental models also supported the pattern of results above. In the above study, participants read scenarios and we measured reaction times as participants tried to comprehend the stories. In a subsequent study, we had participants imagine scenarios based on phrases or sentence fragments we presented to them and then participants spontaneously constructed sentences out of those phrases. For example, participants might read phrases such as, ‘‘party, beautiful woman, toward my friend, dance with, happy’’ or ‘‘restaurant, I, waved at, my friend, to my table.’’ The former would suggest a social scenario with the friend as the subject and the latter would suggest a 8
Every way of being in the world brings with it its own pleasures, pressures, and pains. Euro‐ American cultural imperatives associated with the insider way of being carry with them the burden of introspection and knowing the self, the responsibility and pressure to make the right choice (even for quotidian decisions), and the eVortful exertion needed to create and sustain an ‘‘authentic self’’ (Schwartz, 2000). Asian‐American cultural imperatives associated with the outsider way of being carry with them the anxieties of public self‐consciousness, the pressures of living up to others’ expectations, and the self‐discipline required for empathizing and harmonizing with others (Doi, 2002; Kityama & Markus, 1999; Kitayama et al., 2004; Reischauer & Jansen, 1995; Weber, 1951). The reversals in the nonsocial condition (in which Euro‐Americans take a more outsider perspective and Asian‐Americans take a more insider perspective) may reflect the need for a release from these pressures—with Euro‐Americans letting go and releasing themselves from the pressures of the insider way of being and Asian‐ Americans letting go and releasing themselves from the pressures of the outsider way of being. This sort of release phenomenon is nicely illustrated in a passage from Robert Penn Warren’s All the King’s Men (Warren, 1971): ‘‘Between one point on the map and another point on the map, there was the being alone in the car in the rain. They say you are not you except in terms of relation to other people. If there weren’t any other people, there wouldn’t be any you because what you do, which is what you are, only has meaning in relation to other people. That is a very comforting thought when you are in the car in the rain at night alone, for then you aren’t you, and not being you or anything, you can really lie back and get some rest. It is a vacation from being you. There is only the flow of the motor under your foot spinning . . . that nexus, which isn’t really there, between the you which you have just left in one place and the you which you will be when you get to the other place . . . .But meanwhile, there isn’t either one of them, and I am in the car in the rain at night’’ (pp. 128–129).
CULTURE AND PERSPECTIVE
25
social scenario with the self as the subject. In responding to the latter example, participants might imagine the scene and construct sentences such as, ‘‘At the restaurant, I waved at my friend and he came over to my table’’ or ‘‘At the restaurant, I waved at my friend and he went over to my table.’’ Eighty‐seven Euro‐Americans and 27 Asian‐Americans were given phrases designed to suggest social scenarios with the self as the subject, social scenarios with a friend as the subject, nonsocial scenarios with the self as subject, and nonsocial scenarios with a friend as subject. Analyses of the sentence constructions that participants gave in free response format replicated the pattern of the reaction time data above. For social scenarios in which the self was the subject, Euro‐Americans described the action as coming toward the self (implying a first‐person perspective) in 98% of the sentences they constructed. On the other hand, when their friend was the subject, Euro‐Americans described the action as coming toward the friend in only 79% of the sentences. This diVerential did not hold for Asian‐Americans. For social scenarios, Asian‐Americans described action coming toward the self 85% of the time when the self was the subject, but they also imagined action coming toward the friend 87% of the time when the friend was the subject. For these social stories then, the ethnicity main character (self versus other) interaction was significant, Z ¼ 1.96, p < .05, in predicting the implied first‐ versus third‐person perspective (see Fig. 6). And, consistent with previous results, this pattern did not hold for stories where the scenarios were not social. In fact, the pattern reversed for the nonsocial stories, again producing a replication of the three‐way interaction found in the reaction time data, Z ¼ 3.79, p < .0001. Thus, as indicated both by the sentences that participants spontaneously constructed and by the reaction time data, Euro‐ Americans and Asian‐Americans seemed to mentally model their worlds from diVerent points of view (Leung & Cohen, in press).9 2. Embodiment of the Self and Other’s Perspective In a third study, we extended the work further to the domain of ‘‘embodied cognition’’ and the sense of motion through the world (see Boroditsky, 2000, 2001, 2003; LakoV & Johnson, 1982; Niedenthal, Barsalou, Winkielman, 9 Additional data collected in this study allowed us to rule out the possibility that Euro‐ Americans were simply more familiar with the nuances of what ‘‘come’’ and ‘‘go’’ implied, as compared to Asian‐Americans. At the end of the study, we asked participants to complete two sentences, giving them a forced choice between ‘‘go’’ and ‘‘come’’: (1) From my window, I saw a man leave my apartment building and_______across the street and (2) From my window, I saw my roommate________toward my apartment building. The overwhelming majority of respondents (92% of Euro‐Americans and 95% of Asian‐Americans) correctly completed the first sentence with go and the second sentence with come.
26
COHEN, HOSHINO‐BROWNE, AND LEUNG
Percentage of “come” sentences
Participants' construction of social scenarios 100 95 90 85 80 75 Euro-Americans
Asian-Americans
Sentences with motion toward self Sentences with motion toward friend Fig. 6. Percentage of sentences where motion ‘‘comes’’ toward the self versus ‘‘comes’’ toward a friend for social scenarios that were spontaneously constructed by Euro‐American and Asian‐American participants. Note: Motion coming (rather than going) toward the self implies a first‐person perspective for the self. Motion coming (rather than going) toward a friend implies a mental model that tries to enter into the friend’s first‐person perspective.
Krauth‐Gruber, & Ric, 2005). Briefly, we found that Euro‐Americans were more likely to map out and ‘‘embody’’ time and space from their own perspective, whereas Asian‐Americans were more likely to map out and ‘‘embody’’ time and space from the perspective of another person. For our embodiment study, we adapted a clever paradigm developed by Boroditsky and Ramscar (2002). Their paradigm takes advantage of an ambiguity in language. Thus, for example, the sentence ‘‘Next Wednesday’s meeting has been moved forward two days’’ can either mean that the meeting has been moved to Monday (if we envision time moving toward us) or Friday (if we envision ourselves moving through time). On the basis of an ‘‘embodied cognition’’ hypothesis, Boroditsky and Ramscar (2002) demonstrated that our conception of time depends importantly on how we think about our bodies as moving (or not moving) through space. Thus, they found that people who are at the front of a cafeteria line (and hence have been moving through space) tend to believe that moving the meeting 2 days forward means the meeting is now on Friday, whereas people who are at the back of a cafeteria line (and hence have been stationary or moving very little) tend to believe the meeting is now on Monday. The motion (versus lack of
CULTURE AND PERSPECTIVE
27
motion) through space ‘‘embodies’’ our motion through time (versus time’s motion past us). In our study, we had Asian‐American and Euro‐American participants read vignettes in which either they or another person were described as moving through space (by, e.g., riding a roller coaster at an amusement park). (Again, the vignettes involved either a social scene or a nonsocial scene.) After reading the vignettes, we then asked participants questions such as Boroditsky and Ramscar’s ‘‘two days forward’’ question. If Euro‐ Americans are more likely to ‘‘embody’’ the self ’s perspective (rather than a friend’s), they should be more likely to believe the meeting has been ‘‘moved forward’’ to Friday when the previously imagined scenario involved the self moving through space rather than a friend moving through space. Conversely, if Asian‐Americans are more likely to ‘‘embody’’ the friend’s perspective (rather than the self’s), they should be more likely to believe the meeting has been ‘‘moved forward’’ to Friday when the previously imagined scenario involved the friend moving through space rather than the self. These predictions were confirmed with Euro‐Americans answering ‘‘Friday’’ 82% of the time after imagining the self moving through space compared to 67% of the time after imagining the friend moving through space, whereas Asian‐ Americans answered ‘‘Friday’’ only 43% of the time after imagining the self moving through space compared to 69% of the time after imagining the friend moving through space. And again, this tendency was only observed when the preceding ‘‘motion’’ vignettes involved social scenes rather than nonsocial scenes (ethnicity self/other motion social/nonsocial story interaction, p < .05). Thus, not only did Euro‐Americans and Asian‐Americans mentally model the world diVerently in the reaction time and sentence construction studies, they also embodied the world diVerently, with Euro‐Americans being relatively more likely to sense and embody the imagined motion of the self through time and space and Asian‐Americans being relatively more likely to sense and embody the imagined motion of their friend through time and space (Leung & Cohen, in press). 3. Summary Sections III through VI demonstrated that Asian‐Americans (compared to Euro‐Americans) were more likely to have an ‘‘outsider’’ perspective on the self. The data on mental models and embodiment and (for males) collective memories also provided some suggestion that Asian‐Americans were more likely to try to get into the ‘‘insider’’ perspective of other people. In the next section, we extend this issue and trace out some potential implications of this for social interactions and perceptions.
28
COHEN, HOSHINO‐BROWNE, AND LEUNG
B. THE OUTSIDER PERSPECTIVE AND INSIDER BIAS To believe your own thought, to believe that what is true for you in your private heart is true for all men,—that is genius. Ralph Waldo Emerson in ‘‘Self‐reliance’’ I celebrate myself, And what I assume you shall assume, For every atom belonging to me as good belongs to you. Walt Whitman in ‘‘Song of Myself’’
We have argued that Asian‐Americans are more likely to get outside of their own heads and experience themselves through the eyes of another. This ability to get outside their own heads may make them less vulnerable to the sort of egocentric biases that come from too much attention to one’s own internal thoughts and feelings (see also Wu & Keysar, in press). For Euro‐ Americans, however, the overwhelming salience of their own internal phenomenology may cause them to rely too much on their own ‘‘insider’s’’ perspective and lead them to insuYciently adjust away from it as they think about the external world. For Euro‐Americans, working from an insider’s perspective and too absorbed in their own thoughts, there becomes a fundamental confusion between what is in their own heads, what is in other people’s heads, and what is actually ‘‘out there’’ in objective reality. As described below, these insider biases seem to have a phenomenological origin. In Section VII, we examine how this might operate to cause a confusion between what is in one’s own head and what is out there in reality. In Sections VIII and IX, we extend this point by examining how a similar process might operate to cause a confusion between what is in one’s own head and what is in other people’s heads.
VII. Confusing What Is in One’s Own Head and What Is Out There One very vivid demonstration of insider bias comes from a study by Newton (1991, also see GriYn & Ross, 1991). In this study, participants were brought into the laboratory in pairs. One person (the ‘‘sender’’) was instructed to tap out the rhythm to a popular song (e.g., ‘‘The national anthem’’), whereas the other was instructed to guess what the song was. It turns out people are not really able to detect which song the sender is tapping out. They correctly guessed the song on about 3% of trials. This point, however, is lost on the senders, who believe that the song they have tapped
CULTURE AND PERSPECTIVE
29
out is relatively obvious. (They estimate a 50% chance of accuracy, rather than a 3% chance of accuracy.) The error comes because the sender is attending to something very diVerent than just the taps on the table. As GriYn and Ross (1991) note, the sender hears the taps, but in addition she is also hearing the song fully orchestrated in her head as she taps along to a mental representation of cymbals crashing, a string section, lyrics, and so on. This fully orchestrated song that is playing in her head is easily identifiable, even though the only thing ‘‘out there’’ in the world is the sparse and very ambiguous tapping noise. Attention to either the sparse objective stimulus in the world or the rich, idiosyncratic representation in one’s head profoundly aVects one’s likelihood of figuring out which songs are indeed identifiable. The prediction for our study was that the insider perspective would make Euro‐Americans perform worse than Asian‐Americans in this task. Stated in the converse, because Asian‐Americans are more likely to focus on the external world, they may be less susceptible to the error Euro‐Americans make when they confuse their internal representation of the song with the sparse tapping sounds ‘‘out there’’ in reality. By concentrating on the objective reality, the Asian‐Americans will be better able to discern which rhythms are distinctive and hence easy to guess versus which will be much harder to guess. (As an example of the latter, the ‘‘outsider’’ listening to the rhythm will have a relatively diYcult time telling the diVerence between the tapping that corresponds to Frank Sinatra’s ‘‘My Way’’ and that which corresponds to Britney Spears’ ‘‘Oops! . . . I did it again.’’ The ‘‘insider,’’ who is hearing in her mind either Frank versus Britney’s voice, orchestras versus synthesizers, and so on, will think the songs are perfectly identifiable.) Note that a cognitive load manipulation may have interesting consequences here, depending on what the person is attending to. If the attention is on the objective tapping, the cognitive load will distract the person from listening to the beat and hence will make them worse at discerning what songs will or will not be picked up by an observer. On the other hand, if the attention is focused on the misleading idiosyncratic representation in one’s own head, the load manipulation will disrupt one’s ability to concentrate on that; and as a consequence, a person may actually improve in discerning which songs will be picked up. Thus, in the no‐load conditions, Asian‐Americans may be more accurate than Euro‐Americans because the Asian‐Americans are concentrating on the beat ‘‘out there,’’ whereas the Euro‐Americans are concentrating on the idiosyncratic representation of the song in their heads. In the load conditions, Euro‐Americans should be more accurate than Asian‐Americans because the cognitive load interferes with the Asian‐Americans’ abilities to concentrate on the beat and Euro‐Americans’ abilities to concentrate on the misleading representation in their heads.
30
COHEN, HOSHINO‐BROWNE, AND LEUNG
For our experiment, we modified Newton’s original procedure (1991) in two ways (Hoshino‐Browne & Cohen, 2004). First, we wanted to make it possible for people who were listening only to the beat to guess the songs. Thus, we constrained the possibilities, informing both participants of the possible songs they would be listening to. Second, we wanted to make sure that there would not be a response bias due to either underconfidence or modesty on the part of Asian‐Americans in estimating their ability to tap out the songs in identifiable ways. Thus, instead of having the ‘‘sender’’ tap out the songs, we had a standardized recording of tapping sounds that indicated the rhythm of the various songs (see Newton, 1991, p. 44). Method. Participants were 80 Euro‐Americans and 84 Asian‐Americans, with pairs run one at a time. The experimenter explained that we were studying people’s ability to assess whether others can identify familiar songs when their rhythms are tapped. There was a brief get‐acquainted session for the pair and then the experimenter played music samples for the participants to ensure that all knew the names and beats of the songs in the study. The experimenter explained that in this study, he would play a tape in which the rhythm of the song was tapped out, without vocals, instrumentation, or any other clue to the song’s identity. At that point, the experimenter gave Partner A of the pair a list of five songs in the order that their rhythms would be played. For each song, partner B’s task was to guess which song was being played as well as his or her confidence in the guess. Partner A’s task was to indicate the likelihood that partner B would correctly identify the song. In the cognitive load condition, partner A was also given the task of rehearsing an eight‐digit number. In the no‐load condition, no such task was given to partner A. (Partner A and partner B were separated by a divider so they could not see each other.) Results. As shown in Fig. 7, the ethnicity cognitive load interaction emerged. In the no‐load condition, Asian‐Americans were relatively more likely than Euro‐Americans to discern which songs would be correctly detected by their partner (M ¼ .63 for Asian‐Americans versus M ¼ .52 for Euro‐Americans). In the load condition, this pattern reversed with Euro‐ Americans improving their performance and Asian‐Americans showing declines in performance (M ¼ .49 for Asian‐Americans versus M ¼ .60 for Euro‐Americans). Thus, the interaction was significant, F(1,78) ¼ 4.45, p < .04, eVect size f ¼ .24. Under no‐load conditions, Asian‐Americans were more accurate, presumably because they were attending to the beat ‘‘out there’’ in the world, whereas Euro‐Americans were attending to their own internal representations of the song in their heads. When load was introduced, the Asian‐Americans became less accurate whereas the Euro‐ Americans became more accurate, presumably because Asian‐Americans’ default attention to the objective stimuli was disturbed whereas it was
CULTURE AND PERSPECTIVE
31
Mean accuracy score
0.65
0.60
0.55
0.50
0.45 Euro-Americans
Asian-Americans No load
Load
Fig. 7. Euro‐Americans’ and Asian‐Americans’ accuracy under load and no load conditions in figuring out whether their partner can identify a song just from the tapping sounds.
Euro‐Americans’ default attention to their own idiosyncratic representations that were disturbed. VIII. Confusing What Is in One’s Own Head with What Is in Other People’s Heads: The Illusion of One’s Own Transparency and Empathy‐as‐Projection Section VII examined the focus of attention and how one can confuse what is in one’s own head with what exists in external reality. Sections VIII and IX extend this point by examining the way one may confuse what is in one’s own head with what is in other people’s heads. Two related phenomena are examined in this section. The first is what Gilovich et al. (1998) have called the ‘‘illusion of transparency’’ in which a person believes that others know and understand her feelings to a greater extent than they actually do. The second—which might be called the ‘‘illusion of empathy’’—has to do with the way a person may falsely believe she is empathizing with or understanding others. Thus, the first error involves falsely believing that others know us, whereas the second error involves falsely believing that we know others. Both of these errors are rooted in the phenomenology of the insider perspective. In the ‘‘illusion of transparency’’ or what Vorauer and Ross (1999) have called the ‘‘failure to suppress the self,’’ people overestimate the extent to which others can read their thoughts and feelings (see also Vorauer, Cameron, Holmes, & Pearce, 2003; Vorauer & Claude, 1998). In very clever
32
COHEN, HOSHINO‐BROWNE, AND LEUNG
analyses, both Gilovich et al. as well as Vorauer and Ross (1999) showed that these illusions derive from attending too much to one’s own internal experiences. Thus, in one study, Gilovich and colleagues had participants lie or tell the truth to a group of their peers and then guess whether these peers could catch their lies. As Gilovich et al. (1998) note, the person telling the lie in most cases will feel a certain discomfort with lying—that is, there is a certain subjective unease, nervousness, or physical arousal that a person feels when lying. Under the ‘‘illusion of transparency,’’ one believes that others can pick up on one’s own emotion and discomfort to a greater extent than they actually can; one’s own phenomenology of unease dominates. Importantly, both Gilovich and colleagues as well as Vorauer and Ross (1999) showed that the extent of this bias depended on participants’ own attentiveness to their internal states. The more conscious they were of their own thoughts and feelings, the more they thought that others knew those feelings (see also Laing, 1974, p. 338, on ‘‘ ‘plate glass’ feelings.’’) The process may work diVerently for Asian‐Americans, however, because of a habitual tendency to take the outsider perspective and norms that require overriding personal preferences in favor of what the group deems is important (Morling, Kitayama, & Miyamoto, 2002; see also Morling & Fiske, 1999; Rothbaum, Morelli, Pott, & Liu‐Constant, 2000; Weisz, Rothbaum, & Blackburn, 1984; Ybarra & Trafimow, 1998). That is, in Asian culture, part of being a good group member means going along with what other people want and suppressing one’s own personal desires and emotions for the sake of harmony. One may think or feel X but express Y. In this case, attentiveness to one’s own internal psychological or physiological states would make very salient the discrepancy between what is being privately felt and what is being publicly shown to others. And thus, the more attentive to internal states, the less Asian‐Americans should show this illusion of transparency. The ethnicity X self‐consciousness interaction in this study would follow from greater introspective self‐consciousness, leading to more illusion of transparency for Euro‐Americans and less illusion of transparency for Asian‐Americans. If the illusion of transparency involves a person overestimating how well others know him, the illusion of empathy relates to the opposite mistake of a person overestimating how well he knows others. The root of empathy, at least in theory, is understanding. Among other methods, it can be achieved through listening and watching other people [‘‘what is this person (verbally or nonverbally) expressing?’’]; it can be achieved by knowing something about the person’s background, preferences, and habits (‘‘what would this person think and feel in their situation, given what I know about them?’’); or it can be achieved through projecting oneself (‘‘what would I think and feel in their situation?’’). This typology of methods obviously makes the methods seem more separate than they actually are; however, on the basis of insider
CULTURE AND PERSPECTIVE
33
versus outsider perspective taking, we would expect that Euro‐Americans and Asian‐Americans would diVer in their general routes to understanding others. That is, generally speaking, Euro‐Americans with their insider perspective should implicitly follow the third method (‘‘how would I feel?’’). Asian‐Americans, with their greater attention to the external world, should more likely follow some combination of the first two methods. In the present study, sizing up another’s thoughts and feelings involves detecting when they are lying. The lie‐detecting procedure used below (in which people are randomly assigned to lie or tell the truth to strangers) is a bit artificial and does have its limits (more on that below) (Hoshino‐Browne & Cohen, 2004). However, there are some advantages to this procedure. The first is that there are known, reliable signals that liars produce (Mann, Vrij, & Bull, 2002). Most people are not very good at detecting those signals; but theoretically, if a person is paying close attention to the other person, she should be able to pick up these signals and be reasonably accurate in figuring out when the other person is lying. Thus, if Asian‐Americans use attention to the world as their primary way of empathizing or understanding, these reliable signals should mean that there will be a positive relationship between empathy and accuracy in lie detection. The second advantage of the procedure below is that it gives us greater clarity on how the Euro‐Americans are empathizing. The prediction is that from their insider perspective, Euro‐Americans will engage in a crude sort of projective ‘‘matching’’ as they try to read other people. Thus, if Euro‐ Americans are caught up in the phenomenology of their own internal emotions, they may engage in a very crude sort of projection where their own discomfort with a given topic gets projected onto others. All other things equal, a person will be more uncomfortable about a topic they have lied about than one they have not lied about. And so, from their insider perspective, this discomfort and unease with a given topic would be projected onto the other as the other’s discomfort and unease, and thus a person should be more likely to believe that the other is lying about what she herself has lied about. For Euro‐Americans then, we would expect empathy to be correlated with this crude form of projective matching. As Nickerson (1999) has pointed out, trying to figure out other people by thinking about ourselves is often a good strategy. It can be done deliberately as a person consciously tells himself to think about how he would feel in another’s situation, or it can be done unconsciously or nondeliberately as the person more or less projects himself onto the other without explicitly realizing that he is doing so. An advantage of the experimental setup below is that in this study, we can (almost surely) rule out the possibility that participants are using projection as a consciously chosen, deliberate strategy. [If participant A knows that participant B has been randomly assigned to lie to either
34
COHEN, HOSHINO‐BROWNE, AND LEUNG
the first or second statement and knows that he (A) has independently been randomly assigned, then he knows that his own feelings are obviously completely uninformative about what the other person might be thinking or feeling.] To restate, the insider and outsider perspectives should aVect how we perceive other people. First, the more Euro‐Americans dwell on their own internal phenomenology, the more likely they should show the illusion of transparency. For Asian‐Americans, who habitually take the outsider perspective, the opposite should be true: attention to the internal should highlight how what is privately felt and publicly shown can be quite diVerent. Thus, we would expect an ethnicity self‐consciousness interaction in predicting the illusion of transparency. Second, as a result of their insider perspective, Euro‐ Americans should try to understand and figure out others by more or less unconsciously projecting themselves onto others. For Asian‐Americans, empathy should be less likely to involve projecting oneself onto others. Further, because there are reliable signals people emit when they lie, attention to the outside world should mean that empathic Asian‐Americans will pick up these signals and be able to more accurately guess when the other person is lying. These last two eVects involve (1) an ethnicity empathy interaction in which empathy is positively related to this crude form of projective matching for Euro‐Americans and (2) an ethnicity empathy interaction in which empathy is positively related to actual accuracy in telling the diVerence between lies and truth for Asian‐Americans. Method. Participants were 49 Euro‐Americans and 53 Asian‐Americans. The experimenter explained that this study was about how well people could discern lies from truth, and that participants would take turns lying and telling the truth to each other. Each participant received the same six pairs of statements. For example, (1a) ‘‘Describe your favorite movie and why you like it.’’ and (1b) ‘‘Describe your favorite book and why you like it.’’ For each round, partner A and partner B were each randomly assigned which statement to tell the truth to and which to lie to. Partner A went first, telling a lie to the assigned statement and the truth to the other. Partner B’s task was to guess which of the two answers was the lie and which was the truth, and partner A’s task was to indicate her chance of getting caught. Then partner B took his turn, with partner A guessing which answer was the lie and partner B estimating his own chance of getting caught. The process continued for six rounds. Participants then filled out the Davis (1980) Empathy Scale and the Self‐consciousness Scale (Feningstein, Scheier, & Buss, 1975), which had our measure of private self‐consciousness. Illusory transparency was calculated as a person’s estimate for the chance that he or she would be caught, controlling for the chance that he or she actually was caught. Lie detection accuracy scores were computed as the
CULTURE AND PERSPECTIVE
35
Felt transparency, after correcting for accuracy
proportion of times a person was correct in guessing which of the partner’s answers were lies. Results. Regression analyses supported all three hypotheses. First, there was a significant ethnicity private self‐consciousness interaction predicting the illusion of transparency, with self‐consciousness correlating with relatively more illusion of transparency for Euro‐Americans and relatively less illusion of transparency for Asian‐Americans, (see Fig. 8), t(98) ¼ 2.05, p < .05, eVect size f ¼ .21.10 Second, there was also a significant interaction of ethnicity empathy in predicting accuracy in figuring out when the other person was lying, t(70) ¼ 3.23, p < .002, eVect size f ¼ .39. The more empathic Asian‐Americans were, the more likely they were to figure out when the other person was lying versus telling the truth (r ¼ .37). For the Euro‐Americans, the relation between empathy and accuracy was nonsignificant and, in fact, slightly negative. However, Euro‐American empathy scores were correlated with their tendency to unconsciously project their own discomfort of lying onto others. As shown in Fig. 9, those who were most empathic (as indicated by self‐report on the Davis scale) were relatively more likely to engage in a crude projective matching of their own discomfort with a given topic, whereas this was not
66.0 64.5 63.0 61.5 60.0 −1 SD on private self-consciousness Euro-Americans
+1 SD on private self-consciousness Asian-Americans
Fig. 8. Felt transparency for Euro‐Americans and Asian‐Americans as a function of their private self‐consciousness.
10
In Sections VIII and IX, individuals are nested within dyads and groups. Regression analyses were run with the statistical package SUDAAN that takes nesting into account. Analyses that ignore the nesting give similar results.
36
COHEN, HOSHINO‐BROWNE, AND LEUNG
Lie projection
0.7
0.6
0.5
0.4 −1 SD on empathy Euro-Americans
+1 SD on empathy Asian-Americans
Fig. 9. Euro‐Americans and Asian‐Americans of high or low empathy and their tendency to project their own lies or discomfort onto others.
at all true for Asian‐Americans [interaction t(70) ¼ 2.09, p < .04, eVect size f ¼ .25]. Overall then, the insider perspective on the self seemed to drive the Euro‐ Americans to both the illusion of transparency and the illusion of empathy. Their own phenomenology dominated their perceptions, and they were likely to be biased by attention to their own inner feelings when they were trying to figure out what others knew about them and also what they knew about others. On the other hand, being more likely to take the outsider perspective and get outside their own heads, Asian‐Americans did not show these eVects. Taking an outsider perspective on the self, the more attention one paid to the internal, the more one realized how internal feelings and public expressions can diVer. Further, getting outside of one’s own head allows one to take a diVerent route to empathy. Instead of inadvertently projecting one’s own thoughts and feelings onto the other to understand them, one is able to focus attention on the other and pick up reliable cues that provide for greater accuracy.
IX. Projection‐as‐Empathy in a Group Setting In the present study (done with Alison Luby), we wanted to extend the projection‐as‐empathy finding of the last study in two ways. First, we wanted to examine projection‐as‐empathy in a more natural setting.
CULTURE AND PERSPECTIVE
37
That is, instead of taking turns lying, participants would be involved in a three‐person group discussion. After the discussion, they would be asked how much they liked the other members of the group, and they would be asked to guess how much the other members of the group liked each of the other people. Three‐person groups (rather than dyads) allowed us to extend the findings in a second way. From participants’ ratings, we were able to look at two diVerent sorts of insider projections: in reciprocal projection, Person A might believe that others felt toward her the way she felt about others; and in third party projection, Person A might believe that others felt toward C the way she felt toward C (see also Newcomb, 1961). The prediction was that because Euro‐Americans are more likely to project their insider bias onto others as they empathize, both forms of projection would be positively correlated with Euro‐Americans’ empathy scores. In contrast, this projection‐as‐empathy eVect should not hold for Asian‐Americans. An ethnicity empathy interaction was expected to predict both reciprocal and third‐party projection. Method. Sixty Euro‐Americans and 60 Asian‐Americans participated in three‐person discussion groups and were assigned to come to a collective decision about which of two fictional drugs to market (Karau, 1994). After this collective decision was reached, the experimenter explained that she also wanted to learn people’s private opinions, and she gave a questionnaire to each participant individually. Relevant to the issue of projection, the questionnaire asked about (1) how much the participant liked the other group members and (2) how much the participant thought that each of the other group members liked the people in the group. Results. Regression analyses were consistent with the hypothesis that Euro‐Americans took their own insider perspective and projected it onto others as they tried to understand others’ thoughts and feelings. There was a significant ethnicity empathy interaction predicting reciprocal projection, as empathy was associated with greater reciprocal projection for Euro‐ Americans but not for Asian‐Americans (interaction t ¼ 2.17, p < .04, eVect size f ¼ .2). In addition, the same interaction was shown for third‐party projection, with empathy again being associated with more projection for Euro‐Americans but not for Asian‐Americans (t ¼ 2.2, p < .03, eVect size f ¼ .2).
A. DISCUSSION OF SECTIONS VIII AND IX In Section VIII, it was shown that Euro‐Americans were likely to engage in empathy‐as‐projection in the dyadic lie detection task. In this section, it was shown that the empathy‐as‐projection finding replicated in a triadic group
38
COHEN, HOSHINO‐BROWNE, AND LEUNG
discussion setting with both reciprocal and third‐party projection. In neither of these last two studies did Asian‐Americans show the same insider bias of empathy ¼ projection. (If anything, it was the opposite: projection was something unempathic Asian‐Americans did.) There are diVerent ways people can attempt to understand others; and this inherent ambiguity is part of the definition of the English word ‘‘empathy.’’ The American Heritage Dictionary online (2003) lists two definitions of empathy that reflect two diVerent routes to understanding: ‘‘(1) Identification with and understanding of another’s situation, feelings, and motives. See synonyms at pity. (2) The attribution of one’s own feelings to an object.’’ The Oxford English Dictionary online (2003) conflates these two meanings: ‘‘The power of projecting one’s personality into (and so fully comprehending) the object of contemplation,’’ with the Oxford English Dictionary definition assuming that projecting one’s own feelings leads to ‘‘fully comprehending.’’ There is one English word ‘‘empathy,’’ but it seems to cover, or at least make ambiguous, two quite distinct processes. In contrast, both Chinese and Japanese have at least two words that separate these processes; Yiqing versus Tongqing in Chinese and Kanjo‐inyu versus Kyokan or Kyomei in Japanese correspond to the processes of projection versus entering into the other’s perspective, respectively. A final note: That those with the highest empathy scores projected the most (among Euro‐Americans) has the potential to produce some very ironic outcomes. That is, if I do not consciously realize that I am using the empathy‐as‐ projection technique (as was almost surely the case in Section VIII and as may have been the case in Section IX), my perceptions and understanding of the world may become quite circular. Thus, the irony: I project my thoughts and feelings onto another person, and then I become surprised at how much kinship I feel with that other person, how well I can read his thoughts, and how well I feel I know him generally, because—amazingly!—he is just like me.
X. Characterizing the World How people characterize objects and actions has been seen as quite fundamental by psychologists and others studying culture (Chua, Leu, & Nisbett, 2005; Imai & Gentner, 1997; Menon, Morris, Chiu, & Hong, 1999; Morris & Peng, 1994; Nisbett et al., 2001; Norenzayan, Smith, Kim, & Nisbett, 2002). Our phenomenological experience of an event is fundamentally tied up both with our perceptions of causation (Heider, 1944; Nisbett, 2007; Peng & Knowles, 2003; Storms, 1973; Taylor & Fiske, 1975; Wegner, Fuller, &
CULTURE AND PERSPECTIVE
39
Sparrow, 2003; Wegner, Sparrow, & Winerman, 2004) and with our beliefs about the inherent properties of persons and things (MacLeod, 1947). In this section, we examine characterization as it applies to either seeing the world through one’s own eyes or attempting to enter into another person’s perspective on the world. Further, we examine how these tendencies may change as children develop across the elementary school years. In this study, if Euro‐Americans and Asian‐Americans experience the self and the world from an insider or outsider perspective, this should fundamentally aVect the way they come to characterize the reality around them. At the very basic level of description, do children learn to characterize the world through what they can see through their own eyes as they observe the actions of others? Or, do children learn to characterize the world through getting into the minds of others and examining what those others might be thinking and feeling (Lillard & Flavell, 1990)? The prediction in this study was that as elementary school children get older, there will be a movement from the first type of ‘‘objective,’’ observable characterization of behavior toward the second type of ‘‘subjective’’ characterization of others’ thoughts and feelings; and in particular, this developmental trend to enter into others’ internal experiences will be especially pronounced for Asian‐American children. In this study (done with Tanya Smith), we pitted observable, behavioral descriptions versus unobservable, internal descriptions as participants characterized scenes involving others. Both types of characterizations by themselves were incomplete. Thus, children might be shown a picture of a man playing with a baby as both appeared quite happy. A description of ‘‘The man is playing with the baby’’ is incomplete in that it does not tell us if the man is happy, bored, surprised, frustrated, and so on. A description of ‘‘The man is happy with the baby’’ is incomplete in that it does not tell us whether the man is playing with the baby, feeding the baby, simply staring at the baby, and so on. The child thus has to concentrate on the scene’s most salient features, even if that means leaving some other information out. Further, the pictures in this study involved both social/interactional scenes and those without such social/interactional content. The essential features of such scenes are predicted to diVer for Asian‐American children. It is social situations that cue perspective switching. And thus, we expect that as Asian‐ American children get older, it will be social or interactional situations (rather than nonsocial ones) that will show the greatest developmental shift away from behavioral descriptions and toward descriptions of others’ internal states. Method. Participants were 62 Euro‐American and 46 Asian‐American children, ages 5–12, drawn from two multiethnic schools. Participants were shown a series of 15 pictures and drawings. The eight drawings were from Lillard and Flavell (1990) and the pictures were added to present a greater
40
COHEN, HOSHINO‐BROWNE, AND LEUNG
range of stimuli. Some of the stimuli showed a person (child or adult) interacting with others in a social situation (e.g., a man playing with a baby, a boy hugging his mother, and so on), whereas other stimuli showed a person alone (e.g., a boy looking at some cupcakes, a girl opening a box, and so on). For each picture, the child was asked, ‘‘What’s the best way to describe this picture to a friend?’’ One of the options described the picture in terms of the internal, unobservable states of the actor (e.g., ‘‘She is feeling nervous,’’ ‘‘He is very happy with the baby’’), whereas the other described the picture in terms of the behavior of the actor that could be directly observed (e.g., ‘‘She is talking into the microphone,’’ ‘‘He is tickling the baby’’). Results. For social/interactional situations, Asian‐American children showed a developmental trend away from describing scenes by observable behaviors they could see with their own eyes and toward describing scenes by the internal, unobservable thoughts and feelings of others (r ¼ .31, p < .03). Euro‐American children showed no such developmental trend (r ¼ .06). (Thus, the ethnicity age interaction in regression, p < .05.) Importantly, this was only true for the social situations, not for the nonsocial ones. This helps rule out any response bias explanation and highlights the way that it is social situations that pull for this emphasis on attending to the internal, unobservable aspects of a scene. The three‐way interaction of ethnicity age social/nonsocial scene was significant in a regression at t ¼ 2.42, p < .02, eVect size f ¼ .23.
A. POSSIBLE FUTURE STUDIES For the simple pictures in this study, we forced the distinction between describing a person’s behavior versus describing her thoughts and emotions. However, as stimuli get more complex, so will people’s characterizations. Complex characterizations are likely to contain a mix of the observable behaviors and unobservable feelings of other people. However, some details will always have to be left out of the representation and some will simply fade more quickly over time. As time passes, memories of an event are likely to be simplified to their gist, and thus, even for complex scenes, it would not be surprising if cultural diVerences become even greater as temporal distance from a scene increases, with observable behaviors becoming relatively more prominent in the memories and representations of Euro‐Americans and unobservable thoughts and feelings becoming relatively more prominent in the memories and representations of Asian‐Americans (see also Libby & Eibach, 2002; Peterson, 1981). An interesting experimental test of this would involve showing pictures or film clips to participants and then coming back
CULTURE AND PERSPECTIVE
41
to them, say, a day, a week, a month, or a year later and asking them to recall what they were shown. Responses like ‘‘I remember that situation involved something where a boy felt really nervous about messing up in front of others’’ versus ‘‘I remember that situation involved a boy shooting free throws’’ would be informative about what participants recalled as particularly salient. Further, selective description and selective recall have implications for communication as stories get told and retold to others (Bartlett, 1932; Cohen & Nisbett, 1997; Lyons & Kashima, 2001). The ‘‘sharpening and leveling’’ (Allport & Postman, 1947) of the stories should reflect the prevailing biases of the culture, and thus as the stories are told and retold, we might expect that Euro‐American stories are relatively more likely to turn into stories of actions and observable events whereas Asian‐ American stories are relatively more likely to turn into stories about the thoughts and feelings other people have, the internal lives that occur below the surface veneer of events. Patterns of characterization, memory, and communication are likely to be mutually reinforcing. What we communicate is based on our internal representations of events and becomes further tailored to what we think our audience is interested in hearing (Chiu, Leung, & Kwan, 2007; Wang, 2004; Wang & Conway, 2004; Wang & Ross, in press; Wang et al., 2000). And conversely, what we selectively attend to and remember about scenes is based, to some degree, on what we know we will have to (or want to) communicate to others (Zajonc, 1960). XI. General Discussion The studies described in this chapter are consistent with the notion that the phenomenological gestalts are diVerent for Euro‐Americans and Asian‐ Americans in some fundamental ways. It is not ‘‘simply a metaphor’’ to say that Asian‐Americans are more likely to see themselves the way another person would (see also LakoV, 1987; LakoV & Johnson, 1982). In memories of situations where they are at the center of attention, Asian‐Americans really did see themselves the way an outsider would have. In our self‐control, ‘‘mental toughness’’ task, when others were made salient, Asian‐Americans reported that they really were more likely to have a third‐person, detached experience of looking at themselves as they tried to control their pain sensations. In reading stories and visualizing and embodying narratives, they were relatively more likely to make a mental model in which they imagined their friends’ experience from a first‐person perspective and their own experience from a third‐person perspective. And when they experienced a variety of emotions, the habitual representation of the generalized other was looking back at them. When they felt ashamed, they saw faces as
42
COHEN, HOSHINO‐BROWNE, AND LEUNG
looking at them with contempt; when they felt sad, they saw faces looking at them with sympathy; when they felt angry, they saw faces looking at them with fear; and so on. In all these ways, the self is experienced as the object rather than the subject. Because they saw themselves through other people’s eyes and got out of their own heads, Asian‐Americans were less likely to fall prey to egocentric errors. Euro‐Americans rather than Asian‐Americans were more likely to be overwhelmed by their own ‘‘insider’’ phenomenology and project their internal experiences onto the outside world. Euro‐Americans were likely to confuse what was in their own heads with what was objectively out there in the world. In the music tapping study, they did not realize how far ‘‘beyond the information given’’ their own idiosyncratic representation of a song went as they added orchestration, voices, and synthesizers to the sparse tapping noises that were ‘‘out there’’ in reality. Further, they were likely to confuse what was in their own heads with what was in other people’s heads. In the lie detection study, they were relatively more susceptible to the ‘‘illusion of transparency’’ as they dwelled more in their own phenomenological experience. In both dyadic groups (the lie detection study) and triadic groups (the group interaction study), they empathized by projecting their own internal thoughts and feelings onto other people. They engaged in empathy‐as‐ projection even when their feelings would be definitionally and obviously uncorrelated with others’ feelings (as in the lie detection study) or when their feelings were simply empirically uncorrelated with others’ feelings (as in the group interaction study). In the group study, they engaged in empathy‐as‐ projection both when they were judging other people’s reaction toward them and when they were judging other people’s reaction to a third party. Finally, in a developmental study, Euro‐Americans and Asian‐Americans showed starkly diVerent maturational patterns as they learned to characterize the world. When they chose between characterizing social scenes in terms of either observable behaviors that they could see with their own eyes versus trying to enter into another’s perspective by attempting to understand the other’s thoughts and feelings, Asian‐American children showed a developmental trend from the former to the latter whereas Euro‐American children did not. These results suggest some fundamental diVerences in the phenomenological experience of self and the world. Most research on cross‐cultural diVerences has focused on ideologies—values, beliefs, attitudes, and so on—rather than on phenomenologies. In research on culture and self, for example, more attention has been paid to the contents of self than to the way the self is experienced. It has focused on what we think about rather than how the raw sense data from the world is turned into our conscious experience. As we noted in the introduction, however, an emphasis on the how and on the
CULTURE AND PERSPECTIVE
43
construction of phenomenological experience is important for two (very diVerent) reasons.
A. UNDERSTANDING FELT EXPERIENCE First, to restate, the phenomenological experience is important in its own right. In a famous essay, Nagel (1974) asked, ‘‘What is it like to be a bat?’’ One of Nagel’s points was an antireductionist argument about human consciousness. That is, Nagel argued that an essential aspect of consciousness is subjective experience. Thus, even if we understood all the biology of the brain down to the last synapse, we could not truly claim to have understood consciousness unless we understood humans’ subjective experience. It feels a certain way to be a human, and it feels very diVerent than it would feel to be a bat. There are certain qualia to our subjective consciousness of the world, and without these qualia we cannot understand the essential nature of human experience. Now it is obviously not the case that Asian‐Americans and Euro‐Americans are like humans and bats. However, it is the case that their phenomenological experience of being a self in the world can be quite diVerent in certain situations. The qualia of these feelings cannot be reduced to diVerences in values, beliefs, attitudes, norms, and so on. Without understanding the subjective nature of how the Asian‐American or Euro‐American self in the world is experienced, it would be diYcult to claim that psychologists have understood what it is to be an encultured being. Our understanding is incomplete without the phenomenological data (see also Mesquita & Frijda, 1992; Rozin, 2001; Shweder, 1997). Thus, it is extremely important to know about the what of culture (In what ways does the content of the self diVer across cultures? What are the expectations and norms for various types of social behavior? What do people from diVerent cultures believe and value? and so on). However, it is also important to understand the how—how does our subjective experience of the self and of the social world diVer.11 11 Taken to an extreme, Nagel’s argument could imply that people from two cultures (or even two people from the same culture) could never understand each other. However, as Nagel (1974) notes: ‘‘It is often possible to take up a point of view other than one’s own, so the comprehension of such facts is not limited to one’s own case. There is a sense in which phenomenological facts are perfectly objective: one person can know or say of another what the quality of the other’s experience is. They are subjective, however, in the sense that even this objective ascription of experience is possible only for someone suYciently similar to the object of ascription to be able to adopt his point of view—to understand the ascription in the first person as well as in the third, so to speak. The more diVerent from oneself the other experiencer is, the less success one can expect with this enterprise (p. 441).’’ Given the substantial common ground between all
44
COHEN, HOSHINO‐BROWNE, AND LEUNG
B. PHENOMENOLOGY AND IDEOLOGY: THE LESSONS OF EXPERIENCE Second, understanding phenomenological experience is important because it is probably a key link in two of the most important problems having to do with psychology and culture today. That is, how do we connect culture and individual level psychology, and how do we reconcile long‐standing cultural traditions (that sometimes persist over millennia) with the notion that culture is dynamic, not static. Hong et al. (2000) have talked about ‘‘dynamic constructivism’’ and Kitayama and Markus (1999) about the ‘‘mutual constitution’’ of psyche and culture. In either case, what is happening is that cultures and individual‐level psyches are constantly recreating each other. Macro‐level cultural patterns and micro‐level individual experiences reinforce and recreate one another. In the context of this chapter, we suggest that it is important to consider how micro‐level phenomenological experience and macro‐level cultural ideologies are linked together in bidirectional causation. This bidirectional causation implies (1) that beliefs, attitudes, and values structure our basic phenomenological experience of the world in systematic ways and (2) that certain types of phenomenological experiences create certain sorts of beliefs, values, and attitudes about the world. 1. The Lessons of Experience from the Outsider Perspective Think, for example, of what lessons one draws from an ‘‘outsider’’ phenomenological experience of the self and social world (see top panel of Fig. 10). For example, the experience of seeing oneself in memory as an outsider would makes it evident that the individual is embedded in a social context where he or she is being watched and watched over by other people. In real‐time action, online third‐person experiencing would have the same eVect. Additionally, the experience of relational projection is literally the experience of having the generalized other look on. And finally, constructing mental models in which I take the first‐person perspective of close friends and the third‐person perspective on myself highlights the significance of the group, draws attention to the thoughts and feelings of close others, and strengthens the bonds of sympathy human beings, understanding is probably limited in most cases only by failures of imagination, analogy, or vocabulary. More colloquially, the absurdity of taking the argument about subjectivity and unbridgeable chasms too far is reminiscent of the quip attributed (perhaps apocryphally) to Walker Percy: The average deconstructionist is an academic who argues for the incomprehensibility of language just before calling his or her spouse to ask them to order a pizza. Yes, there is a sense in which language will involve ambiguity and referential diYculties, but it usually works well enough to get the large double sausage with pepperoni home more or less on time.
CULTURE AND PERSPECTIVE
Asian-American Phenomenological Experience • Experience of “generalized other” looking at self
Asian-American Ideology •
Individual within a tight community that watches and watches over him/her
•
Emphasis on propriety
•
Culture of “Face”
•
Ethical stance of “do not unto others,” reticence
•
Ideal of unself-conscious certitude, assertiveness, and self-confidence
•
Tocqueville’s habits of the heart. Characteristic types of American sociability. o Alliances and easy, fast friendships o Side-by-side relationships
•
Interventionist, ethical stance of “do unto others”
•
Ideal of the rational, agentic individual who knows what he/she wants and is able to go out and get it.
•
Importance of the “authentic” self.
• Other-centered empathy • Seeing self as outsider would in memory • Online experience of detached observer’s perspective for self-control
45
• Construal in terms of others’ internal states • Mental models using first-person perspective for others and thirdperson perspective on self • Cognitive embodiment of others’ perspective Euro-American Phenomenological Experience •
Lack of third-person imagery
•
Relative inattention to others’ internal states
•
Egocentric projection
•
Projection as empathy
•
Confusion of what is in the head with what is “out there” in the world
•
Mental models using first-person perspective for self and thirdperson perspective for others
•
Strong salience of one’s own internal states
Euro-American Ideology
Fig. 10. Phenomenology and ideology: The embedded (and embodied) assumptions and lessons of experience from the insider and outsider perspectives.
between group members. The lessons of experience are that the self is a social object, these lessons highlight the importance of propriety before others, and they enable the self‐scrutiny or self‐criticism that drives one to improve and meet others’ expectations. The metaphor of ‘‘face’’ is one of the central concepts in East Asian societies and this metaphor fits in quite nicely with the
46
COHEN, HOSHINO‐BROWNE, AND LEUNG
phenomenological third‐person experiencing of self we have been describing. Literally, one can never see one’s own face, except by taking the perspective of other people and looking at oneself the way others would. The face is something that can only be seen from the outside. The West, too, uses the metaphor of ‘‘face,’’ but it is probably not a coincidence that the metaphor was imported from the East. ‘‘Saving face’’ is an idiom derived from Chinese expressions and developed by English expatriates in China (Oxford English Dictionary online, 2003). And perhaps it is worth considering other parallels between the concept of ‘‘face’’ as a metaphor and the literal, physical instantiation of one’s face. In another stunning experiment, Miyamoto, Kitayama, and Schwarz (2004) have illustrated how emotions become most psychologically real to Japanese participants when they are experienced in the gaze of another. The details and background of the experiment bear some fleshing out for the purposes of examining the metaphor. In earlier studies done at the University of Illinois and Manheim University, Strack, Martin, and Stepper (1988) cleverly provided support for the facial feedback hypothesis of emotion. Briefly, respondents examined cartoons as they were instructed (under a cover story) to hold a pen between their teeth or between their lips. Holding the pen between the teeth produces a facial expression characteristic of a smile as the lips spread. In contrast, holding the pen between the lips produces a facial expression characteristic of negative aVect as the lips purse. Strack et al. (1988) found that those who held the pen between their teeth (simulating a smile) thought the cartoons were funnier than those who held it between their lips (simulating a pout). Miyamoto et al. (2004) repeated this study in Japan and found startling results. Individual Japanese participants showed no evidence of the facial feedback eVect when they saw the cartoons. The expressions on their face had no implications whatsoever for their own subjective emotional state. However, when Japanese participants were put in front of a poster with schematic drawings of faces, the facial feedback eVect emerged. Thus, only under the gaze of these schematic others did the expressions on their own face have any meaning. Only in this state where they imagined they were being looked at did the contortions of their face have any implications for their subjective state. We should not push the point too far. However, in this study, just as the concept of ‘‘face’’ is meaningless without an audience, so also are the physical expressions of one’s actual face meaningless without being gazed on by an imagined other. More generally, Miyamoto et al.’s findings (2004)—along with the dissonance findings of Kitayama et al. (2004)—imply that for the Japanese, experiences become most psychologically real when one is taking the outsider perspective on oneself and simulating the gaze of the generalized other. The phenomenological experience of emotion under these circumstances supports and is supported by the Japanese cultural idea that
CULTURE AND PERSPECTIVE
47
emotions are essentially interpersonal (rather than intrapersonal) experiences [Diener, Oishi, & Lucas, 2003; Kitayama, Markus, & Kurokawa, 2000; Mesquita, 2001; Mesquita & Leu, 2007; for a parallel notion of emotions as essentially interpersonal rather than internal see also, Levenson, Ekman, Heider, and Friesen’s work (1992) with the Minangkabau of West Sumatra. Unlike American participants who posed for researchers, Minangkabau men who posed facial expressions of anger, fear, disgust, sadness, and happiness did not report subjectively experiencing these emotions. Among other possibilities, this could be because the ‘‘critical element’’ of emotion in Minangkabau culture is ‘‘the meaningful involvement of another person’’ (Levenson et al., 1992, p. 985). Additionally, for experiments on ‘‘face’’ and the incorporation of public representations and the generalized other’s perspective, see experiments in Kim and Cohen (2007)]. 2. The Lessons of Experience from the Insider Perspective On the other hand, think of the lessons one draws from insider experiences (see bottom panel of Fig. 10). An ideal type. The unconscious confusion of what is in my head, what is out there in reality, and what is in your head will lead to the confidence that my perception of reality is correct, that you as well perceive the same things I do, and that we understand each other. In these circumstances, there is no reason to be reticent, to doubt my perceptions, or to make obsessive eVorts to consider other perspectives. These experiences create the person with the self‐assurance, the confidence, the certitude in one’s convictions, and the action orientation that is one of the ideal American types. A person of this ideal type (1) knows what he or she wants and (2) is able to go out and get it. The certitude, conviction, and action orientation that derive from the insider experience obviously relate to the latter point (going out and getting what one wants). And in a loose culture where harmony is valued less than it is in a tight culture, even the obliviousness or insensitivity to others (which is sometimes produced by the insider perspective) may be eYcacious because it allows for the single‐minded pursuit of one’s goals. (Or, if it is not eYcacious, the insensitivity to others will be at least less severely punished in loose rather than tight cultures.) With respect to the former point, knowing what one wants often demands an attentiveness to and a focus on one’s internal thoughts and desires (see also Wang, 2004; Wang & Conway, 2004; Wang et al., 2000). Such attentiveness to the internal can lead to egocentric thinking, as in the studies here. However, in a culture that views ‘‘true’’ preferences and beliefs as those that are internal to the person and are uncontaminated by the influence of others (Kitayama et al., 2004), such attention to the internal may be necessary for
48
COHEN, HOSHINO‐BROWNE, AND LEUNG
being an ‘‘authentic’’ person, with clear preferences, desires, and beliefs that make up one’s individuality. To crudely simplify, we might summarize by saying that Euro‐Americans (compared to Asian‐Americans) are more likely to take the insider perspective because: (1) they can. (Obliviousness to others will simply be punished less in a loose culture than in a tight one.) And (2) they have to. (It is one of the cultural imperatives in Euro‐American culture to figure out what you want and to go get it. To know what you want takes a lot of attention to your internal world. And to get what you want, it sometimes helps to have self‐assurance, certitude, and occasionally, obliviousness to other people.) Alliances and side‐by‐side relationships. The insider experience further ties into the patterning of relationships between one person and another as well as between a person and the group. That is, it has been remarked that Americans prefer their relationships ‘‘side‐by‐side’’ rather than ‘‘face‐to‐face,’’ and related to the conception of the ‘‘side‐by‐side’’ orientation are Americans’ treasured egalitarianism, the varied cooperative but weak ties with peers described by Tocqueville, and more generally, an individualism that makes it easy for Americans to form relationships but at the same time keeps these relationships relatively shallow [see observations by Mead (1965), following the lead of Kurt Lewin]. We suggest that the insider perspective that Euro‐Americans take on the world facilitates this. Thus, the Euro‐American insider perspective that creates empathy‐as‐projection facilitates a fast sort of knowing that gives Americans great skill at forming quick friendships and a great many genial acquaintances. The insider perspective also makes Euro‐Americans quite ready to assume that others perceive reality similarly and thus makes Euro‐Americans ready to assume others’ agreement (see also Thorne, 1987). This (real or illusory) agreement creates a readiness for action and a willingness to join with peers to accomplish something that all are presumably side‐by‐side in working toward (thus, the creation of the many goal‐directed associations and organizations that Tocqueville described). And we think it is significant here that Tocqueville described these associations as essentially directed toward a goal. The aim of these associations was not the building of relationships per se, but rather these associations were often ad hoc alliances for the purposeful accomplishment of some goal that Americans were working side‐by‐side, shoulder‐to‐shoulder, and arm‐in‐arm to achieve.12 If Tocqueville’s Democracy in America (2000, xvii) is indeed both ‘‘the best book ever written on democracy and the best book 12
C. S. Lewis in his book, The Four Loves, proposes a similar notion in characterizing Friendship and Companionship (versus Eros). Lovers are ‘‘face‐to‐face,’’ concerned with each other. Friends and Companions are ‘‘side‐by‐side’’ in pursuit of some common goal. Friendship (and Companionship), he notes, ‘‘must be about something’’ (Lewis, 1960, p. 98, emphasis added).
CULTURE AND PERSPECTIVE
49
ever written on America,’’ it is so because it got something essentially right about a nation of egalitarian individualists ready to quickly form and quickly dissolve side‐by‐side alliances in pursuit of goals and ‘‘self‐interest rightly understood’’ as common interest (book 2, chapter 8). The insider phenomenology—and the characteristic patterns of sociality to which it leads—works quite well for individuals and their ‘‘habits of the heart’’ in such a system. At the same time, whereas the insider perspective allows for fast friends and alliances, it can also lead to a false sense of knowing; thus, these relationships can lack the intimacy that is found in ‘‘face‐to‐face’’ relationships where people must take time to listen and where the relationship itself (rather than some common goal) is the focus of attention. In essence, by allowing people to presume agreement and preventing them from probing for a deeper knowing, the insider perspective facilitates the fast, genial, but relatively superficial side‐by‐side relationships that Euro‐Americans are skilled at building. Again, the insider perspective and the ‘‘habits of the heart’’ Tocqueville described are congruent. (Note: this is not to say Euro‐ Americans lack truly intimate relationships—only that they accumulate a great many of the side‐by‐side kind).13 [See also work by Yuki and colleagues (Yuki, 2003; Yuki, Maddux, Brewer, & Takemura, 2005) arguing that Japanese tend to pay close attention to the habits and preferences of in‐ group members, whereas Americans tend to simply assume that their in‐group members are more or less similar to them.] Interventionist, ‘‘Do unto others’’ morality. Finally, there are also implications of the ‘‘empathy‐as‐projection’’ phenomenology and the readiness for action that relate to some very basic ethical creeds of Western and Eastern philosophies. In Christianity, the Golden Rule is expressed as ‘‘Do unto others as you would have them do unto you.’’ In the Analects, Confucius states the rule in the negative as ‘‘What you do not want done to yourself, do not do to others.’’ Philosophers and theologians have grappled with the question of the diVerences between the positive and negative phrasing of the rule. Some have dismissed the issue merely as a grammatical quibble. Others have argued that the positive injunction encourages us to project our
13
Some of Margaret Mead’s observations tend to trivialize this aspect of American sociability by focusing on how superficial the connections can sometimes be. Mead noted how Americans can bond over similarities that Europeans find trivial. Mead snickeringly describes the ‘‘feverish grabs at a common theme’’—‘‘enthusiastic preferences for the same movie actor, the same brand of peaches, the same way of mixing a drink’’—that Americans use to create rapport (Mead, 1965, pp. 34–35). These are questions that ‘‘say diagrammatically, ‘Are you the same kind of person I am? Good, how about a Coke?’ ’’ (Mead, 1965, p. 29). In Mead’s example, a trivial, tenuous connection is established, and this calls forth egocentric projections that make the two Americans think they are brothers under the skin.
50
COHEN, HOSHINO‐BROWNE, AND LEUNG
own desires onto others as we assume that others are essentially similar to us, whereas the negative injunction encourages more reticence and limits the scope of our projections and actions. Philosopher Antonio Cua (1995), for example, argued that ‘‘the negative formulation of the Golden Rule may thus be construed as a counsel of modesty, which stresses the importance of being aware of one’s limited perspective’’ (Weiss, 1995, p. 506). Mitchell (1993) views the diVerence in phrasing as essentially superficial, but acknowledges the ambiguity of interpretation: ‘‘The point of the Golden Rule is empathy. Jesus certainly didn’t mean it as a projection of egoism, but it has often been taken that way. So the ‘negative’ is perhaps the more helpful one’’ ( p. 189). The negative restrains the impulse toward universalizing (for better or for worse). (More colloquially, it echoes Shaw’s quip (1903) in Man and Superman: ‘‘Do not do unto others as you would that they should do unto you. Their tastes may not be the same.’’) In terms of the paths described in Fig. 10, the phenomenological experience of empathy‐as‐projection; the confusion between what we think, feel, and prefer and what other people think, feel, and prefer; and the naı¨ve realism of mistaking what is in our heads with what is really out there in the world is quite consonant with a system of ethics that encourages one to ‘‘do unto others’’ on the basis of perceived similarity. And conversely, experiences that highlight the limitations of one’s own perspective are likely to be associated with a much less interventionist system of ethics. Attempting to think in terms of another person’s thoughts, feelings, and imagery will not always be successful (especially given the diYculties that arise when one lives in a culture where harmony is valued and true preferences cannot always be revealed.) However, the attempt to see the world through another’s perspective is probably enough to prevent at least some actions that might have arisen from an egocentric projection of one’s own desires, beliefs, and values. We close this discussion on the lessons of experience by oVering the ‘‘reverse’’ thought experiment from the introduction. That is, imagine two societies where a group of individuals begin ‘‘tabula rasa’’ as rational but unencultured beings. In society A, imagine the individuals such that each person (1) has the phenomenological experience of seeing herself as an outsider would in memory, in real‐time experience, and in her mental models; (2) has the habitual feeling of being watched by a generalized other; (3) has the imagery of imagining and embodying another’s sensations and perceptions; (4) construes situations in terms of other people’s internal states; and (5) empathizes through paying careful attention to other people. What kind of culture would emerge from the interactions of people in society A? On the other hand, imagine society B, where the average individual (1) surveys the world only through her own eyes in memories, real‐time experience, and mental models; (2) is overpowered by her own internal thoughts and
CULTURE AND PERSPECTIVE
51
sensations such that she unknowingly projects her own experience onto other people and also confidently confuses what is in her own head with objective reality; and (3) learns to characterize the world in terms of what is immediately visible to her rather than in terms of other people’s internal thoughts and feelings. What sort of culture would emerge from the interactions of people in society B? It would be our contention that one possible result is that the people in society A would develop a tight, collectivistic culture, where the individual is seen as embedded in a community that watches and watches over her, where a concept such as ‘‘face’’ would be important, where sympathy would be prized, and where behavior would be guided by norms of propriety and circumspection. On the other hand, the people in society B would develop a culture that would prize individualism, agency, and assertiveness; hold freedom as a prominent ideal; create side‐by‐ side relationships based on common interests; foster a sociability where an easy, fast sense of knowing allows friendships and alliances to both quickly develop and dissolve; would view ‘‘genius’’ as deriving from the belief that ‘‘what is true for you in your private heart is true for all men;’’ and would develop an interventionist ethical code in which benevolence is to ‘‘do unto others as you would have them do unto you’’ (Fig. 10).14,15 14
Of course, there are multiple possible cultures that could emerge in society A and society B, because people’s mutual interdependence leads to multiple possible equilibriums (see also Cohen, 2001). We simply present one plausible outcome. We also note that the reverse thought experiment begs the question of why people in diVerent societies have diVerent structures of phenomenological experience in the first place. This issue underscores our belief in bidirectional causality and feedback loops between micro‐level experience and macro‐level culture. Finally, we recognize that with their evolutionary history, humans do not enter the world tabula rasa. Nor (past early infancy) can they be unencultured (Konner, 2002). 15 On the surface, there seems to be a paradox between Euro‐Americans’ belief that (1) they are unique but that (2) others are just like them. Potentially, this may result from the diVerent degrees to which the self and others are cognitively elaborated in the minds of Euro‐Americans, which then allows for a classic framing eVect in similarity judgments. (All other things equal, if object A has a richly elaborated schema and object B has a poorly elaborated schema, then B will seem more similar to A than A is to B. And thus, Markus and Kitayama (1991) showed that North Americans were relatively more likely than South Asians to see others as similar to the self, but they were relatively less likely than South Asians to see the self as similar to others.) Potentially also, the false consensus and false uniqueness eVects are examples of Euro‐Americans’ motivated, self‐enhancing cognitions (Heine, 2007; Heine & Lehman, 1995, 1997a,b; Heine & Renshaw, 2002; Heine, Takata, & Lehman, 2000; Heine et al., 2001; Oishi, 2002; Oishi & Diener, 2003; Snibbe, Kitayama, Markus, & Suzuki, 2003; Tweed, White, & Lehman, 2004; Wang, 2004; White & Lehman, 2005). That is, Euro‐Americans generally tend to show false consensus on matters of opinions and on undesirable behaviors, whereas they tend to show false uniqueness on matters of ability or on desirable behaviors (Myers, 1990; but see Kruger, 1999; Miller & Prentice, 1994). For the present purposes, however, both false uniqueness and false consensus can also be seen as evidence of the more general point that Euro‐Americans can sometimes be drastically out of touch with what other people are thinking, feeling, and doing.
52
COHEN, HOSHINO‐BROWNE, AND LEUNG
C. OTHER EXAMPLES OF MICRO‐LEVEL EXPERIENCE AND MACRO‐LEVEL IDEOLOGY IN HOMEOSTASIS That the individual‐level phenomenology of self is tied up with our ideological beliefs about the individual and the group is related to the more general point that micro‐level experience and macro‐level cultural ideology can be mutually reinforcing as cultures and individual psyche ‘‘make each other up’’ (see Fig. 11). Thus, for instance, discussing how social interdependence and general cognitive styles are related, Nisbett et al. (2001) have argued that a hierarchical, collectivistic social order that stressed social interdependence in Asia led to a style of thinking where attention was directed toward the larger social field. Taking the metaphor of Asian holistic philosophy (versus Greek atomistic philosophy) literally, Nisbett and colleagues argued that habitual attention to the social field led to a perceptual and cognitive style that literally makes salient ‘‘backgrounds’’ rather than central objects, relations between entities rather than entities themselves, contexts rather than properties in isolation, and so on (Ji, Peng, & Nisbett, 2000; Ji, Zhang, & Nisbett, 2004; Kitayama, DuVy, Kawamura, & Larsen, 2003; Norenzayan, Choi, & Peng, 2007). Thus, ‘‘holistic’’ cultural and social structural arrangements led to a holistic style of perception and cognition, and this holistic thinking style in turn reinforced cultural assumptions about the importance of the group, the power of situations, and the inherent interdependence of people (Choi & Nisbett, 1998; Choi, Nisbett, & Norenzayan, 1999; Hong et al., 2000; Miller, 1984; Morris & Peng, 1994). If ‘‘behavior engulfs the field’’ and leads to a focus on the individual actor for American perceivers, perhaps one should say that ‘‘relationships engulf the field’’ (or are more likely to do so) for Asian perceivers (Heider, 1944). Quoting Resnick (1994, pp. 476–477), Nisbett et al. (2004) argue that ‘‘the tools of thought . . . embody a culture’s intellectual history. . .. Tools have theories built into them, and users accept these tools— albeit unknowingly—when they use these tools.’’ Eastern holistic (or Western analytic) cognitive and perceptual styles ‘‘exist in homeostasis with the social practices that surround them’’ (Nisbett et al., 2004, p. 56). Bidirectional causality. That micro‐level phenomenology and macro‐level cultural ideology can aVect each other causally implies that (1) temporarily changing phenomenology should lead to a (temporary) change in the endorsement of various ideological beliefs and (2) conversely, temporarily making certain ideological beliefs salient should lead to (temporary) changes in phenomenological experience. Regarding these points, three sets of studies are quite interesting. First, one of the manipulations from objective self‐awareness research and from self‐objectification theory is to put a study participant in front of a mirror. In a study with Japanese and North American college students, Heine (2005) had participants evaluate themselves either in
53
CULTURE AND PERSPECTIVE Phenomenological Experience
Ideas/Ideology
Separation of objects from their environment, perceptual focus on a central object or actor, whose “behavior engulfs the field.”
Analytical tradition. Emphasis on individualism (Nisbett et al., 2001; Heider,1944)
Integration of objects with their environment, binding central object to the background. Relationships engulf the field.
Holistic tradition. Emphasis on interdependence (Nisbett et al., 2001)
Embodiment in physical structures, for example, the comportment of a person’s body and its associated subjective experience
Cultural ideas and beliefs [e.g., about the nature of and proper roles for males and females (Bourdieu, 1977)].
Embodiment of emotion. For example, phenomenology of shame−feelings of shrinking, head down and gaze avoidant (literally, a refusal to present one’s face), displays of appeasement
Other people or higher being as judge, primacy of one’s relationship to them, reinforcement of one’s subordinance to hierarchy or community
Mimicry, fellow feeling
Insider perspective on self versus outsider perspective on self
Subjective experience of emotion in isolation versus subjective experience of emotion when under the gaze of “generalized other”
Beliefs about dignity of the other, shared fate (Smith,1759)
Ideal of high self-esteem versus ideal of self-criticism and self-improvement (Heine et al., 1999; Oishi & Diener, 2003)
Emotions as intrapersonal phenomenon versus interpersonal phenomenon (Markus & Kitayama, 1991; Mesquita & Karasawa, 2002)
Fig. 11. Further examples of phenomenological experience and cultural ideas in homeostasis.
front of a mirror or not in front of a mirror (the control condition). He found that for Japanese, their self‐assessments were unaVected by the presence or absence of a mirror. On the other hand, for Americans in the control condition, their self‐assessments were very positive. However, when the Americans
54
COHEN, HOSHINO‐BROWNE, AND LEUNG
were put in front of a mirror, they became more self‐critical such that Americans in front of a mirror now produced self‐assessments similar to those of the Japanese. It seems that for Americans, giving them the phenomenology of the self‐as‐object produces the same sorts of self‐critical practices that Japanese routinely show (Heine et al., 1999; Oishi & Diener, 2003; see Fig. 11). Further, the absence of a mirror eVect for the Japanese suggests that the Japanese, even in the control condition, may experience the same sort of self‐as‐object phenomenology that the mirror induces. Obviously, the data are not conclusive because similar eVects can derive from diVerent causal processes, but they are suggestive. Second, another intriguing study was done by van Baaren, Horgan, Chartrand, and Dijkmans (2004) on the bidirectional causation between behavioral mimicry and a holistic (context‐dependent) thinking style. Behavioral mimicry is the simple (probably unconscious) tendency to imitate the gestures and actions of people with whom we are interacting (Chartrand & Bargh, 1999). van Baaren and colleagues showed not only that holistic thinkers were more likely to mimic than analytic thinkers were. They also showed that: (1) a confederate who mimicked participants caused participants’ thinking styles to become more holistic and (2) a prime that induced participants to think holistically caused them to show increased mimicry of a confederate. In this latter case, participants who were induced to attend to stimuli holistically [either by focusing on global (versus local) processing or by focusing on the relationship of objects within a picture] were subsequently more likely to mimic the behavior of an experimental confederate. Thus, there seemed to be a reciprocal loop: an imitative action tendency seemed to both cause and be caused by a certain attentional focus/cognitive style. The phenomenology and the imitative tendency reinforced each other. Such findings are perfectly consonant with the argument of scientists who contend that mimicry may be the root of empathy. ‘‘Mirror neurons’’ fire in humans (and monkeys) when we perform an action X or when we watch another perform an action X (Carr, Iacoboni, Dubeau, Mazziotta, & Lenzi, 2003; Ramachandran, 2003). Neural activation when we simply observe an emotion is similar to that shown when we imitate the emotion. And similar brain activation occurs when a person receives pain and when he or she sees a loved one about to experience the same pain (Singer et al., 2004). Mimicry and a feeling of empathy go together quite naturally (Bower, 2003; Smith, 1759). Again, the phenomenological experience here may be crucial: the feelings that make it feel like one has just undergone the same experience as another probably derive from and also give rise to some sense of solidarity and kinship with the other. One feels another’s emotions because there is a common bond with the other, but one also probably feels a common bond because one feels the other’s emotion (Fig. 11; see also Chartrand & Bargh,
CULTURE AND PERSPECTIVE
55
1999; Lakin & Chartrand, 2003; Lakin, JeVeris, Cheng, & Chartrand, 2003; van Baaren, Maddux, Chartrand, de Bouter, & van Knippenberg, 2003). Third, fascinating studies have also shown the eVects that occur when an independent or interdependent social orientation is primed in participants. More specifically, raising the salience of independence or interdependence, or in the case of Chinese biculturals, raising the salience of one’s Chinese heritage or American heritage, produces a number of quite subtle eVects. Thus, making one’s independence or interdependence salient changes the way one processes stimuli in terms of local versus global focus as well as influences the extent to which people attend to the relations between stimuli (Kuhnen, Hannover, & Schubert, 2001; Kuhnen & Oyserman, 2002). For bicultural individuals, having participants look at Chinese (versus American) icons makes them more likely to characterize a scene in terms of group (versus individual) behavior (Hong et al., 2000). Our ideological beliefs and the phenomenological structure of our personal experience go hand‐in‐hand because perceiving the world a certain way leads to certain beliefs and holding certain beliefs leads to perceiving the world a certain way.
D. PHENOMENOLOGY, METHODOLOGY, AND INTROSPECTION: THE INVISIBILITY AND THE STRENGTH OF THE SOFT EMBODIMENT OF CULTURE Methodologically, discerning the structure of personal experience can sometimes be more diYcult than discerning beliefs. There are pitfalls and perils to asking people about their attitudes, but overall, researchers can generally get a decent accounting of people’s beliefs (Heine, Lehman, Peng, & Greenholtz, 2002; Oishi et al., 2005; Peng, Nisbett, & Wong, 1997; Schimmack, Oishi, & Diener, 2005). Other people’s phenomenological experience, however, is not something that will ‘‘pop‐out’’ to researchers. We can observe other people’s behaviors and listen to what they say, but the structure of their phenomenological experience is not obviously and immediately accessible to us. We have found, at least in discussing the matter with participants, that whereas some recognized that Asian‐Americans tend to focus more on the group and Euro‐Americans tend to focus more on the individual, they were not aware that people from the two cultures might have such phenomenologically diVerent experiences. The nature of these experiences is not so readily articulated to researchers for two reasons. First, sometimes participants do not have the language to describe their experiences. A good example here is first‐person versus third‐person memory. Many people whom we discussed memory with had had both first‐ and third‐person memories, but
56
COHEN, HOSHINO‐BROWNE, AND LEUNG
they had never really thought about the distinction, had not reflected on the experience, and did not have the ready terminology to describe it. Second, many people generally assumed that the way their phenomenological experience was structured was basically ‘‘the way everybody does it’’ (similar to the way a person generally assumes her experience of the color ‘‘blue’’ and other people’s experience of the color blue are pretty much the same). Perhaps this belief in direct perception or this sort of primitive realism is the ultimate egocentrism error. However, what makes phenomenological experience so hard to get at is also what makes it so powerful. Because it is rarely reflected on and because it is the felt experience we are conscious of rather than the raw data of our senses, the powerful influence we exert on the shape and structure of our thought can go unnoticed by ourselves and by others. Bourdieu (1977) has made similar points—both about (1) the link between ideology and phenomenology and (2) the strength of an ideology that has become embodied and perhaps hidden in people’s subjective experience. Regarding the first point and also arguing that much cultural learning does not take place in the form of explicit declarations of rules, he argues that: ‘‘In all societies, children are particularly attentive to the gestures and postures which, in their eyes, express everything that goes to make an accomplished adult—a way of walking, a tilt of the head, facial expressions, ways of sitting and of using implements, always associated with a tone of voice, a style of speech, and (how could it be otherwise?) a certain subjective experience’’ (Bourdieu, 1977, p. 87; see also Bem, 1972; Cacioppo, Priester, & Bernston, 1993; Chen & Bargh, 1999; James, 1890; Laird, 1974).16 A culture’s ideas become ‘‘em‐bodied’’ in individuals: In Bourdieu’s example, this embodiment is ‘‘hard’’—it is in the way we literally comport our physical bodies. In the studies we have run, this embodiment is ‘‘soft’’—it is in the way we cognitively represent and psychologically structure our bodily experience, from either an insider or an outsider perspective (Boroditsky & Ramscar, 2002; Niedenthal et al., 2005; Zajonc & Markus, 1984). Regarding the second point, this embodiment makes an ideology particularly powerful and resistant to change because its expression has become so implicit and so natural. Bourdieu makes the strong argument that ‘‘the principles embodied in this way are placed beyond the grasp of consciousness, and hence cannot be touched by voluntary, deliberate transformation, cannot 16
The same point about children learning adult values through mimicry and embodiment can also be made in reverse. Adults can learn about the views, values, and experience of a child through mimicry of the child. Imitate a child for an hour or a day and one can come to appreciate the wonder of small things (such as how much can be found between the cracks of a sidewalk), the peace and simplicity of repetition, the arbitrariness of many social norms, the urgency of the moment (why do kids run everywhere?), and an implicit theory of learning that lets you get up and try again after you have literally fallen down 50 times already that day.
CULTURE AND PERSPECTIVE
57
even be made explicit; nothing seems more ineVable, more incommunicable, more inimitable, and, therefore, more precious, than the values given body, made body by the transubstantiation achieved by the hidden persuasion of an implicit pedagogy, capable of instilling a whole cosmology, an ethic, a metaphysic, a political philosophy, through injunctions as insignificant as ‘stand up straight’ or ‘don’t hold your knife in your left hand’ ’’ (Bourdieu, 1977, p. 94). Bourdieu perhaps overstates the case, but we agree with the general point that ideas implicit or embodied in a phenomenology (and sometimes ‘‘hidden’’ in a phenomenology) can be particularly powerful and resistant to challenge.
E. GENERALIZATION In this chapter, we have compared Asian‐Americans and Euro‐Americans. However, the outsider perspective is, of course, not unique to Asian‐ Americans. For example, in the strong honor cultures of North Africa, the self‐as‐seen‐by‐others is of paramount importance. Bourdieu (1965, p. 212) wrote: ‘‘Perhaps the conclusion is that the important position accorded to the sentiment of honour is characteristic of ‘primary’ societies in which the relationship with others, through its intensity, intimacy, and continuity, takes precedence over the relationship with oneself; in which the individual learns the truth about himself through the intermediary of others; and in which the being and the truth about a person are identical with the being and truth that others acknowledge in him.’’ Though there are important diVerences between face and honor (Leung & Cohen, 2004), this description of honor cultures resonates well with descriptions of the face cultures of East Asia. The Japanese salaryman (businessman) in the modern corporate hierarchy and Bourdieu’s Kabyle villager in Algeria both may share a structure of experiences where one takes an outsider’s perspective on oneself [see experiments on face in Kim and Cohen (2007), showing how what is publicly known about the individual becomes represented in the private self for Hong Kong respondents; see also Fischer, Manstead, & Mosquera, 1999; Mosquera, Manstead, & Fischer, 2002]. Additionally, in cultures where God is always watching, one might expect to find people also taking the outsider perspective on themselves. And here, the eVect might not be limited to social situations. An omniscient God will know how one behaves no matter where one is, and further, an omniscient God will know what one is thinking—regardless of how one outwardly behaves. After a wrongdoing then, one does not ‘‘simply’’ feel guilt; one feels shame before witnesses and before God. In fact, this shame versus guilt distinction—and the relative decline of shame as compared to guilt in the
58
COHEN, HOSHINO‐BROWNE, AND LEUNG
West—can be seen in the various English translations of the Bible in the past four centuries. In the King James Version of the Bible (published in 1611), shame is mentioned five times as often as guilt. By the turn of the twentieth century, the ratio was down to 3:1. And in the relatively secularized society of North America today, that ratio is now down to 1:1 (see Cohen, 2003 for further discussion). Shame indicates one’s submission to a community or a deity (see also Fig. 11, row 4). Shame focuses on how the self would be seen by others, and in deeply religious cultures where shame is acutely felt, the same outsider perspective may be more common as people think about how other people or God are looking on. In the studies above, it is probably not the Asian‐Americans that are unusual. Instead, modern North Americans and modern Europeans may be the proverbial ‘‘anthropological veto’’ to the universal rule: It is they who may have the peculiar ‘‘default’’ phenomenology (Konner, 2002; Sanchez‐ Burks, 2002). The question then becomes not, do other groups share the same eVects as Asian‐Americans, but rather, what groups might show the same phenomenology as the Euro‐Americans? To do a cultural census on this point is an undertaking worthy of Sisyphus. For now, we simply speculate that around the world, the tendency to experience the world from the ‘‘inside out’’ even in social situations is probably less common than the tendency to take the outsider perspective in such situations. Again, Euro‐ Americans may not realize the peculiarity of their ‘‘default’’ phenomenology (and the peculiar ideology it is tied up with) because such is the nature of the phenomenological eVect—starting with oneself and experiencing the world from the ‘‘inside out’’ just feels so natural to Euro‐Americans that it may seem ‘‘strange’’ to have it another way. For groups where the self is predominately realized through other people, such an outsider phenomenology may not seem so very strange or awkward at all.
F. SUMMARY In sum, we have tried to argue that Euro‐Americans and Asian‐Americans have phenomenological gestalts that can be diVerent in some very basic and fundamental ways. Euro‐Americans are likely to experience the world from the ‘‘inside out,’’ starting with their own experience and projecting it onto others. Asian‐Americans are likely to experience the self from the ‘‘outside in,’’ taking an observer’s perspective on the self. This ability to get outside one’s own head may work against the egocentric biases that Euro‐Americans are susceptible to when they allow their own internal phenomenology to dominate. These diVerences are far from absolute—people from both cultures are able to take both insider and outsider perspectives (as seen in Sections IV, V, VI,
CULTURE AND PERSPECTIVE
59
and X), but there are diVerences in the ‘‘default’’ perspective one takes in social situations (Rozin, 2003). The insider–outsider diVerences are not ‘‘simply’’ metaphorical but reflect the real, felt experience of people. And these diVerences in the phenomenology of experience are intimately related to diVerences in cultural ideologies about the individual, the group, and tight versus loose social norms. The ideologies and the phenomenologies probably mutually reinforce each other. Thus, seeing the self from a third‐person or generalized other’s perspective reinforces (and is reinforced by) the idea that the individual is part of a group and is watched and watched over by other people. Similarly, experiencing the world from the inside out helps create (and is created by) the self‐assured, confident, genial, and ready‐for‐action individual that is one of the ideal American types. The diVerences outlined here are not peculiar to the Asian‐American versus Euro‐American distinction, though when it comes to various situations, it may be that the ‘‘default’’ Euro‐American phenomenology is relatively unusual across the globe. People may not spontaneously articulate that they are taking either an insider versus an outsider perspective on the self, in part because it feels so natural that to do something else would seem strange. That is, we may see the self and the social world through either an insider or outsider perspective, but we often may not spontaneously think about how that perspective was created and how that perspective shapes our experience of and beliefs about reality. Nevertheless, understanding the structure of personal experience is crucially important for at least two reasons. It is important both in and of itself if psychologists are to capture the lived experience of what it means to be in one culture or another, and it is important if we are to understand the way that micro‐level psychological processes (such as individual‐level phenomenology) and macro‐level culture recreate each other.
Acknowledgments Work in this paper was supported by funds from the University of Illinois, Swarthmore College, Social Sciences and Humanities Research Council of Canada, and the University of Waterloo. The authors are grateful to Mark Zanna for his helpful suggestions.
References Allport, G., & Postman, L. (1947). The psychology of rumor. New York: Henry Holt. American Heritage Dictionary online (2003). Empathy. Retrieved October 24, 2003. Baldwin, M. W., & Holmes, J. G. (1987). Salient private audiences and awareness of the self. Journal of Personality and Social Psychology, 52, 1087–1098.
60
COHEN, HOSHINO‐BROWNE, AND LEUNG
Bartlett, F. C. (1932). Remembering. Cambridge: Cambridge University Press. Bem, D. (1972). Self‐perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 1–62). New York: Academic Press. Black, J. B., Turner, T. J., & Bower, G. H. (1979). Point of view in narrative comprehension, memory and production. Journal of Verbal Learning and Verbal Behavior, 18, 187–198. Blanke, O., Ortigue, S., Landis, T., & Seeck, M. (2002). Stimulating illusory own‐body perceptions. Nature, 419, 269–270. Boroditsky, L. (2000). Metaphoric structuring: Understanding time through spatial metaphors. Cognition, 75, 1–28. Boroditsky, L. (2001). Does language shape thought? English and Mandarin speakers’ conceptions of time. Cognitive Psychology, 43, 1–22. Boroditsky, L. (2003). Linguistic relativity. In L. Nadel (Ed.), Encyclopedia of cognitive science (pp. 917–921). London: MacMillan Press. Boroditsky, L., & Ramscar, M. (2002). The roles of body and mind in abstract thought. Psychological Science, 13, 185–188. Bourdieu, P. (1965). The sentiment of honour in Kabyle society. In J. Peristiany (Ed.), Honour and shame (pp. 191–242). London: Weidenfeld and Nicholson. Bourdieu, P. (1977). Outline of a theory of practice. Cambridge: Cambridge University Press. Bower, B. (2003). Repeat after me. Science News, 163, 330. Bower, G. H., & Morrow, D. G. (1990). Mental models in narrative comprehension. Science, 247, 44–48. Butler, R., Damarin, F. L., Beaulieu, C., Schwebel, A. I., & Thorn, B. E. (1989). Assessing cognitive coping strategies for acute postsurgical pain. Psychological Assessment, 1, 41–45. Cacioppo, J., Priester, J., & Bernston, G. (1993). Rudimentary determinants of attitudes II. Journal of Personality and Social Psychology, 65, 5–17. Campbell, J., Trapnell, P., Heine, S., Katz, I., Lavallee, L., & Lehman, R. (1996). Self‐concept clarity. Journal of Personality and Social Psychology, 70, 141–156. Carr, L., Iacoboni, M., Dubeau, M.‐C., Mazziotta, J., & Lenzi, G.‐L. (2003). Neural mechanisms of empathy in humans. Proceedings of the National Academy of Sciences USA, 100, 5497–5502. Chan, D., Gelfand, M., Triandis, H., & Tzeng, O. (1996). Tightness‐looseness revisited. International Journal of Psychology, 31, 1–12. Chartrand, T. L, & Bargh, J. A. (1999). The chameleon eVect. Journal of Personality and Social Psychology, 76, 893–910. Chen, M., & Bargh, J. A. (1999). Consequences of automatic evaluation. Personality and Social Psychology Bulletin, 25, 215–224. Chiu, C. Y., Leung, K. Y., & Kwan, L. (2007). Language, cognition, and culture. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 668–688). NY: Guilford. Choi, I., & Nisbett, R. E. (1998). Situational salience and cultural diVerences in the correspondence bias and actor‐observer bias. Personality and Social Psychology Bulletin, 24, 949–960. Choi, I., Nisbett, R. E., & Norenzayan, A. (1999). Causal attribution across cultures: Variation and universality. Psychological Bulletin, 125, 47–63. Chua, H., Leu, J., & Nisbett, R. (2005). Culture and diverging views of social events. Personality and Social Psychology Bulletin, 31, 925–934. Cohen, D. (2001). Cultural variation. Psychological Bulletin, 127, 451–471. Cohen, D. (2003). The American national conversation about (everything but) shame. Social Research, 70, 1075–1108. Cohen, D. (2007). Methods in cultural psychology. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 196–236). NY: Guilford. Cohen, D., & Gunz, A. (2002). As seen by the other. Psychological Science, 13, 55–59.
CULTURE AND PERSPECTIVE
61
Cohen, D., & Nisbett, R. E. (1997). Field experiments examining the culture of honor. Personality and Social Psychology Bulletin, 23, 1188–1199. Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York: Harper. Cua, A. (1995). Reasonable persons and the good. In L. Hahn (Ed.), The philosophy of Paul Weiss (pp. 495–514). Peru, Illinois: Open Court. Davis, M. H. (1980). A multidimensional approach to individual diVerences in empathy. Catalog of Selected Documents in Psychology, 10, 85. Diener, E., Oishi, S., & Lucas, R. (2003). Personality, culture and subjective well‐being. Annual Review of Psychology, 54, 403–425. Diener, E., & Wallbom, M. (1977). EVects of self‐awareness on antinormative behavior. Journal of Research in Personality, 10, 107–111. Doi, T. (2002). The anatomy of dependence. NY: Kodansha America. Dubois, W. E. B. (1903). The souls of black folk. Chicago: A. C. McClurg & Co. Duval, S., & Wicklund, R. A. (1972). A theory of objective self awareness. Oxford: Academic Press. Eastwood, J. D., Gaskovski, P., & Bowers, K. S. (1998). The folly of eVort. International Journal of Clinical and Experimental Hypnosis, 46, 77–91. Edgerton, R. (1992). Sick societies. NY: Free Press. Eibach, R., Libby, L., & Gilovich, T. (2003). When change in the self is mistaken for change in the world. Journal of Personality and Social Psychology, 84, 917–931. Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective taking as egocentric anchoring and adjustment. Journal of Personality and Social Psychology, 87, 327–339. Epley, N., Morewedge, C., & Keysar, B. (2004). Perspective taking in children and adults. Journal of Experimental Social Psychology, 40, 760–768. Feningstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private self‐consciousness. Journal of Consulting and Clinical Psychology, 43, 522–527. Fischer, A. H., Manstead, A. S. R., & Mosquera, P. M. (1999). The role of honour‐related vs. individualistic values in conceptualising pride, shame, and anger: Spanish and Dutch cultural prototypes. Cognition & Emotion, 13, 149–179. Fredrickson, B., Roberts, T., Noll, S., Quinn, D., & Twenge, J. (1998). That swimsuit becomes you. Journal of Personality and Social Psychology, 75, 269–284. Fujita, K., Trope, Y., Liberman, N., & Levi‐Sangi, M. (2006). Construal levels and self control. Journal of Personality and Social Psychology, 90, 351–367. Gallagher, S. (2004). Neurocognitive models of schizophrenia: A neurophenomenological critique. Psychopathology, 37, 8–19. Gelfand, M., Higgins, M., Nishii, L., Raver, J., Dominguez, A., Murakami, F., et al. (2002). Culture and egocentric perceptions of fairness in conflict and negotiation. Journal of Applied Psychology, 87, 833–845. Gibson, J. J. (1979). The ecological approach to visual perception. Boston, MA: Houghton MiZin. Gilovich, T., Medvec, V., & Savitsky, K. (1998). The illusion of transparency. Journal of Personality and Social Psychology, 75, 332–346. GriYn, D., & Ross, L. (1991). Subjective construal, social inference, and human misunderstanding. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 24, pp. 319–359). San Diego: Academic Press. Heidegger, M. (1962). Being and time. New York: Harper. Heider, F. (1944). Social perception and phenomenal causality. Psychological Review, 51, 358–374. Heine, S. J. (2005). Constructing good selves in Japan and North America. In R. Sorrentino, D. Cohen, J. M. Olson, & M. P. Zanna (Eds.), Culture and social behavior: The Ontario Symposium (Vol. 10, pp. 95–116). Mahwah, NJ: Erlbaum.
62
COHEN, HOSHINO‐BROWNE, AND LEUNG
Heine, S. (2007). Culture and motivation. In S. Kitayama & D. Cohen (Eds.), Handbook of Cultural Psychology (pp. 714–733). NY: Guilford. Heine, S. J., Kitayama, S., Lehman, D. R., Takata, T., Ide, E., Leung, C., et al. (2001). Divergent consequences of success and failure in Japan and North America. Journal of Personality and Social Psychology, 81, 599–615. Heine, S. J., & Lehman, D. R. (1995). Cultural variation in unrealistic optimism: Does the West feel more invulnerable than the East? Journal of Personality and Social Psychology, 68, 595–607. Heine, S. J., & Lehman, D. R. (1997a). Culture, dissonance, and self‐aYrmation. Personality and Social Psychology Bulletin, 23, 389–400. Heine, S. J., & Lehman, D. R. (1997b). The cultural construction of self‐enhancement: An examination of group‐serving biases. Journal of Personality and Social Psychology, 72, 1268–1283. Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal need for positive self‐regard? Psychological Review, 106, 766–794. Heine, S. J., Lehman, D. R., Peng, K., & Greenholtz, J. (2002). What’s wrong with cross‐ cultural comparisons of subjective Likert scales? Journal of Personality and Social Psychology, 82, 903–918. Heine, S. J., & Renshaw, K. (2002). Interjudge agreement, self‐enhancement, and liking: Cross‐ cultural divergences. Personality and Social Psychology Bulletin, 28, 578–587. Heine, S. J., Takata, T., & Lehman, D. R. (2000). Beyond self‐presentation: Evidence for self‐ criticism among Japanese. Personality and Social Psychology Bulletin, 26, 71–78. Hong, Y., Morris, M., Chiu, C., & Benet‐Martinez, V. (2000). Multicultural minds: A dynamic constructivist approach to culture and cognition. American Psychologist, 55, 709–720. Hoshino‐Browne, E., & Cohen, D. (2004, January). Inside out and outside in: Cross‐cultural variations in social perspective‐taking among Asian‐Canadians and European‐Canadians. Poster presented at the Fifth Annual Meeting of the Society for Personality and Social Psychology, Austin, Texas. Hoshino‐Browne, E., Zanna, A. S., Spencer, S. J., Zanna, M. P., Kitayama, S., & Lackenbauer, S. (2005). On the cultural guises of cognitive dissonance: The case of Easterners and Westerners. Journal of Personality and Social Psychology, 89, 294–310. Imai, M., & Gentner, D. (1997). A cross‐linguistic study of early word meaning. Cognition, 62, 169–200. Iyengar, S. S., & Lepper, M. R. (1999). Rethinking the role of choice: A cultural perspective on intrinsic motivation. Journal of Personality and Social Psychology, 76, 349–366. James, W. (1890). The principles of psychology. New York: Dover. Ji, L., Peng, K., & Nisbett, R. E. (2000). Culture, control, and perception of relationships in the environment. Journal of Personality and Social Psychology, 78, 943–955. Ji, L., Schwarz, N., & Nisbett, R. E. (2000). Culture, autobiographical memory, and behavioral frequency reports. Personality and Social Psychology Bulletin, 26, 585–593. Ji, L. J., Zhang, Z., & Nisbett, R. E. (2004). Is it culture, or is it language? Examination of language eVects in cross‐cultural research on categorization. Journal of Personality and Social Psychology, 87, 57–65. Karau, S. (1994). Group interaction and group decision making. Unpublished dissertation, Purdue University. Kim, Y., & Cohen, D. (2007). Public representations and the private self in face and dignity cultures. Champaign: University of Illinois, Unpublished manuscript. Kitayama, S., DuVy, S., Kawamura, T., & Larsen, J. (2003). Perceiving an object and its context in diVerent cultures. Psychological Science, 14, 201–206. Kitayama, S., DuVy, S., & Uchida, Y. (2007). Self as cultural mode of being. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 136–174). NY: Guilford.
CULTURE AND PERSPECTIVE
63
Kitayama, S., & Markus, H. R. (1999). Yin and yang of the Japanese self. In D. Cervone & Y. Shoda (Eds.), The coherence of personality (pp. 242–302). New York: Guilford. Kitayama, S., Markus, H. R., & Kurokawa, M. (2000). Culture, emotion, and well‐being: Good feelings in Japan and the United States. Cognition and Emotion, 14, 93–124. Kitayama, S., Markus, H., Matsumoto, H., & Norasakkunkit, V. (1997). Individual and collective processes in the construction of the self. Journal of Personality and Social Psychology, 72, 1245–1267. Kitayama, S., Snibbe, A. C., Markus, H.R, & Suzuki, T. (2004). Is there any ‘‘free’’ choice?: Self and dissonance in two cultures. Psychological Science, 15, 527–533. Konner, M. (2002). The tangled wing: Biological constraints on the human spirit. NY: Henry Holt. Kruger, J. (1999). Lake Wobegon be gone! The ‘‘below‐average eVect’’ and the egocentric nature of comparative ability judgments. Journal of Personality and Social Psychology, 77, 221–232. Kuhnen, U., Hannover, B., & Schubert, B. (2001). The semantic–procedural interface model of the self: The role of self‐knowledge for context‐dependent versus context‐independent modes of thinking. Journal of Personality and Social Psychology, 80, 397–409. Kuhnen, U., & Oyserman, D. (2002). Thinking about the self influences thinking in general. Journal of Experimental Social Psychology, 38, 492–499. Kwan, V., Bond, M. H., & Singelis, T. M. (1997). Pancultural explanations for life satisfaction. Journal of Personality and Social Psychology, 73, 1038–1051. Laing, R. D. (1974). Self‐consciousness. In R. Solomon (Ed.), Existentialism (pp. 337–342). NY: Random House. Laird, J. (1974). Self‐attribution of emotion. Journal of Personality and Social Psychology, 29, 475–486. Lakin, J., & Chartrand, T. (2003). Using nonconscious behavioral mimicry to create aYliation and rapport. Psychological Science, 14, 334–339. Lakin, J., JeVeris, V., Cheng, M., & Chartrand, T. (2003). The chameleon eVect as social glue. Journal of Nonverbal Behavior, 27, 145–162. LakoV, G. (1987). Women, fire, and dangerous things. Chicago: University of Chicago. LakoV, G., & Johnson, M. (1982). Metaphors we live by. Chicago: University of Chicago. Leung, A. K.‐y., & Cohen, D. (2004, August). Honor and reciprocity. Poster presented at the International Conference of Psychology. Beijing, China. Leung, A. K.‐y., & Cohen, D. (in press). The soft embodiment of culture: Camera angles and motion through time and space. Psychological Science. Levenson, R. W., Ekman, P., Heider, K., & Friesen, W. V. (1992). Emotion and autonomic nervous system activity in the Minangkabau of West Sumatra. Journal of Personality and Social Psychology, 62, 972–988. Levine, R., Martinez, T., Brase, G., & Sorenson, K. (1994). Helping in 36 U.S. cities. Journal of Personality and Social Psychology, 67, 69–82. Lewis, C. S. (1960). The four loves. New York: Harcourt, Brace. Libby, L. K., & Eibach, R. P. (2002). Looking back in time: Self‐concept change aVects visual perspective in autobiographical memory. Journal of Personality and Social Psychology, 82, 167–179. Libby, L. K., & Eibach, R. P. (2003, February). Seeing meaning: Visual perspective and action identification in mental imagery. Poster presented at the Fourth Annual Meeting of the Society for Personality and Social Psychology, Los Angeles, CA. Lillard, A. S., & Flavell, J. H. (1990). Young children’s preference for mental state versus behavioral descriptions of human action. Child Development, 61, 731–741. Lyons, A., & Kashima, Y. (2001). The reproduction of culture. Social Cognition, 19, 372–394.
64
COHEN, HOSHINO‐BROWNE, AND LEUNG
MacLeod, R. B. (1947). The phenomenological approach to social psychology. Psychological Review, 54, 193–210. Mann, S., Vrij, A., & Bull, R. (2002). Suspects, lies and videotape. Law and Human Behavior, 26, 365–376. Markus, H., & Kitayama, S. (1991). Culture and the self. Psychological Review, 98, 224–253. Mead, G. (1934). Mind, self, and society. Chicago: University of Chicago Press. Mead, M. (1965). And keep your powder dry. Oxford: Morrow. Menon, T., Morris, M. W., Chiu, C., & Hong, Y. (1999). Culture and the construal of agency: Attribution to individual versus group dispositions. Journal of Personality and Social Psychology, 76, 701–717. Mesquita, B. (2001). Emotions in individualist and collectivist contexts. Journal of Personality and Social Psychology, 80, 68–74. Mesquita, B., & Frijda, N. (1992). Cultural variations in emotions. Psychological Bulletin, 112, 179–204. Mesquita, B., & Karasawa, M. (2002). DiVerent emotional lives. Cognition & Emotion, 16, 127–141. Mesquita, B., & Leu, J. (2007). The cultural psychology of emotion. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 734–759). NY: Guilford. Miller, D., & Prentice, D. (1994). Collective errors and errors about the collective. Personality and Social Psychology Bulletin, 20, 541–550. Miller, J. G. (1984). Culture and the development of everyday social explanation. Journal of Personality & Social Psychology, 46, 961–978. Miller, P. J., & Goodnow, J. J. (1995). Cultural practices. In J. J. Goodnow, P. J. Miller, & F. Kessel (Eds.), Cultural practices as contexts for development. San Francisco: Jossey‐ Bass. Mitchell, S. (1993). The Gospel according to Jesus. NY: Harper Collins. Miyamoto, Y., Kitayama, S., & Schwarz, N. (2004). Emotion and facial feedback. Unpublished manuscript, University of Michigan. Morelli, G., & Rothbaum, F. (2007). Situating the child in context: Attachment relationships and self‐regulation in diVerent cultures. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 500–527). NY: Guilford. Morling, B., & Fiske, S. T. (1999). Defining and measuring harmony control. Journal of Research in Personality, 33, 379–414. Morling, B., Kitayama, S., & Miyamoto, Y. (2002). Cultural practices emphasize influence in the U.S. and adjustment in Japan. Personality and Social Psychology Bulletin, 28, 311–323. Morris, M., & Peng, K. (1994). Culture and cause. Journal of Personality and Social Psychology, 67, 949–971. Mosquera, P. M., Manstead, A. S. R., & Fischer, A. H. (2002). The role of honor concerns in emotional reactions to oVences. Cognition & Emotion, 16, 143–163. Myers, D. (1990). Social psychology. New York: McGraw‐Hill. Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83, 435–450. NemeroV, C., & Rozin, P. (1989). ‘‘You are what you eat.’’ Ethos, 17, 50–69. Newcomb, T. (1961). The acquaintance process. NY: Holt, Rinehart, & Winston. Newton, E. L. (1991). The rocky road from actions to intentions. Unpublished doctoral dissertation, Stanford University. Nickerson, R. (1999). How we know—and sometimes misjudge—what others know: Imputing one’s knowledge to others. Psychological Bulletin, 125, 737–759. Niedenthal, P., Barsalou, L., Winkielman, P., Krauth‐Gruber, S., & Ric, F. (2005). Embodiment in attitudes, social perception, and emotion. Personality and Social Psychology Review, 9, 184–211.
CULTURE AND PERSPECTIVE
65
Nigro, G., & Neisser, U. (1983). Point of view in personal memories. Cognitive Psychology, 15, 467–482. Nisbett, R. (2007). Cultural psychology past, present, and future. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology. NY: Guilford. Nisbett, R., Peng, K., Choi, I., & Norenzayan, A. (2001). Culture and systems of thought: Holistic versus analytic cognition. Psychological Review, 108, 291–310. Nisbett, R., Peng, K., Choi, I., & Norenzayan, A. (2004). Culture and systems of thought. Retrieved Nov., 15, 2004 from http://www.personal.umich.edu/nisbett/ Norenzayan, A., Choi, I., & Peng, K. (2007). Perception and cognition. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 569–594). NY: Guilford. Norenzayan, A., Smith, E. E., Kim, B. J., & Nisbett, R. E. (2002). Cultural preferences for formal versus intuitive reasoning. Cognitive Science, 26, 653–684. Oishi, S. (2002). The experiencing and remembering of well‐being: A cross‐cultural analysis. Personality and Social Psychology Bulletin, 28, 1398–1406. Oishi, S., & Diener, E. (2003). Culture and well‐being. Personality and Social Psychology Bulletin, 29, 939–949. Oishi, S., Hahn, J., Schimmack, U., Radhakrishan, P., Dzokoto, V., & Ahadi, S. (2005). The measurement of values across cultures: A pairwise comparison approach. Journal of Research in Personality, 39, 299–305. Oxford English Dictionary online (2003). Empathy, face. Retrieved October 24, 2003. Oyserman, D., & Lee, W. S. (2007). Priming ‘culture’: Culture as situated cognition. In S. Kitayama & D. Cohen (Eds.), Handbook of cultural psychology (pp. 255–279). NY: Guilford. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism. Psychological Bulletin, 128, 3–72. Peng, K., & Knowles, E. D. (2003). Culture, education, and the attribution of physical causality. Personality and Social Psychology Bulletin, 29, 1272–1284. Peng, K., Nisbett, R., & Wong, N. (1997). Validity problems comparing value across cultures and possible solutions. Psychological Methods, 2, 329–344. Percy, W. (1983). Lost in the cosmos. New York: Washington Square Press. Peterson, C. (1981). Memory and the ‘‘dispositional shift.’’ Social Psychology Quarterly, 43, 372–380. Piaget, J., & Inhelder, B. (1956). The child’s conception of space. London: Routledge & Kegan Paul. Plaut, V., Markus, H., & Lachman, M. (2002). Place matters. Journal of Personality and Social Psychology, 83, 160–184. Ramachandran, V. S. (2003). Mirror neurons and imitation learning as the driving force behind ‘‘the great leap forward’’ in human evolution. Retrieved October 24, 2003, from http:// www.edge.org Reischauer, E., & Jansen, M. (1995). The Japanese today. Cambridge, MA: Belknap. Resnick, L. (1994). Situated rationalism. In L. Hirschfeld & S. Gelman (Eds.), Mapping the mind (pp. 474–494). Cambridge: Cambridge University Press. Rosenthal, R., & Rosnow, R. L. (1991). Essentials of behavioral research: Methods and data analysis (2nd ed.). New York: McGraw‐Hill. Ross, M., & Sicoly, F. (1979). Egocentric biases in availability and attribution. Journal of Personality and Social Psychology, 37, 322–336. Rothbaum, F., Morelli, G., Pott, M., & Liu‐Constant, Y. (2000). Immigant‐Chinese and Euro‐ American parents’ physical closeness with young children. Journal of Family Psychology, 14, 334–348. Rothbaum, F., Pott, M., Azuma, H., Miyake, K., & Weisz, J. (2000). The development of close relationships in Japan and the United States. Child Development, 71, 1121–1142.
66
COHEN, HOSHINO‐BROWNE, AND LEUNG
Rozin, P. (2001). Social psychology and science: Some lessons from Solomon Asch. Personality and Social Psychology Review, 5, 2–14. Rozin, P. (2003). Five potential principles for understanding cultural diVerences in relation to individual diVerences. Journal of Research in Personality, 37, 273–283. Rozin, P., Markwith, M., & NemeroV, C. (1992). Magical contagion beliefs and fear of AIDS. Journal of Applied Social Psychology, 22, 1081–1092. Rozin, P., Markwith, M., & Ross, B. (1990). The sympathetic magical law of similarity, nominal realism and neglect of negatives in response to negative labels. Psychological Science, 1, 383–384. Rozin, P., Millman, L., & NemeroV, C. (1986). Operation of the laws of sympathetic magic in disgust and other domains. Journal of Personality and Social Psychology, 50, 703–712. Sanchez‐Burks, J. (2002). Protestant relational ideology and (in)attention to relational cues in work settings. Journal of Personality and Social Psychology, 83, 919–929. Sartre, J. P. (1948). The emotions. Oxford: Philosophical Library. Sartre, J. P. (1965). Being and nothingness. Oxford: Philosophical library. Scheier, M. F., & Carver, C. S. (1977). Self‐focused attention and the experience of emotion. Journal of Personality and Social Psychology, 35, 625–636. Schimmack, U., Oishi, S., & Diener, E. (2005). Individidualism. Personality and Social Psychology Review, 9, 17–31. Schwartz, B. (2000). Self‐determination: The tyranny of freedom. American Psychologist, 55, 79–88. Schwartz, B., Ward, A., Monterosso, J., Lyubomirsky, S., White, K., & Lehman, D. R. (2002). Maximizing versus satisficing: Happiness is a matter of choice. Journal of Personality and Social Psychology, 83, 1178–1197. Shaw, G. B. (1903). Man and superman. Cambridge, MA.: The University Press. Shweder, R. (1997). The surprise of ethnography. Ethos, 25, 152–163. Singer, T., Seymour, B., O’Doherty, J., Kaube, H., Dolan, R. J., & Frith, C. D. (2004). Empathy for pain involves the aVective but not sensory components of pain. Science, 303, 1157–1162. Smith, A. (1759). The theory of moral sentiments. Oxford: Clarendon Press. Snibbe, A. C., Kitayama, S., Markus, H., & Suzuki, T. (2003). They saw a game. Journal of Cross‐Cultural Psychology, 34, 581–595. Snibbe, A. C., & Markus, H. R. (2005). You can’t always get what you want: Educational attainment, agency, and choice. Journal of Personality and Social Psychology, 88, 703–720. Storms, M. (1973). Videotape and the attribution process. Journal of Personality and Social Psychology, 27, 165–175. Strack, F., Martin, L., & Stepper, S. (1988). Inhibiting and facilitating conditions of the human smile. Journal of Personality and Social Psychology, 54, 768–777. Taylor, S., & Fiske, S. T. (1975). Point of view and perceptions of causality. Journal of Personality and Social Psychology, 32, 439–445. Thorne, A. (1987). The press of personality: A study of the conversations between intraverts and extraverts. Journal of Personality and Social Psychology, 53, 718–726. Tocqueville, A. (2000). In Harvey C. Mansfield & D. Winthrop (Eds.), Democracy in America. Chicago: University of Chicago. Trafimow, D., Triandis, H., & Goto, S. (1991). Some tests of the distinction between the private and the collective self. Journal of Personality and Social Psychology, 60, 649–655. Triandis, H. (1994). Culture and social behavior. New York: McGraw‐Hill. Triandis, H. (2004). The many dimensions of culture. Academy of Management Executive, 18, 88–93. Tweed, R., White, K., & Lehman, D. (2004). Culture, stress, and coping. Journal of Cross‐ Cultural Psychology, 35, 652–668.
CULTURE AND PERSPECTIVE
67
van Baaren, R., Horgan, T., Chartrand, T. L., & Dijkmans, M. (2004). The forest, the trees, and the chameleon: Context dependency and nonconscious mimicry. Journal of Personality and Social Psychology, 86, 453–459. van Baaren, R., Maddux, W., Chartrand, T., de Bouter, C., & van Knippenberg, A. (2003). It takes two to mimic. Journal of Personality and Social Psychology, 84, 1093–1102. Vandello, J., & Cohen, D. (1999). Patterns of individualism and collectivism across the United States. Journal of Personality and Social Psychology, 77, 279–292. Vandello, J. A., & Cohen, D. (2003). Male honor and female fidelity. Journal of Personality and Social Psychology, 84, 997–1010. Vorauer, J., Cameron, J., Holmes, J., & Pearce, D. (2003). Invisible overtures: Fears of rejection and signal amplification bias. Journal of Personality and Social Psychology, 84, 793–812. Vorauer, J., & Claude, S. (1998). Perceived versus actual transparency of goals in negotiation. Personality and Social Psychology Bulletin, 24, 371–385. Vorauer, J., & Ross, M. (1999). Self‐awareness and feeling transparent: Failing to suppress one’s self. Journal of Experimental Social Psychology, 35, 415–440. Wang, Q. (2004). The emergence of cultural self‐construct: Autobiographical memory and self‐ description in American and Chinese children. Developmental Psychology, 40, 13–15. Wang, Q., & Conway, M. A. (2004). The stories we keep: Autobiographical memory in American and Chinese middle‐aged adults. Journal of Personality, 72, 911–938. Wang, Q., Leichtman, M. D., & Davies, K. I. (2000). Sharing memories and telling stories: American and Chinese mothers and their 3‐year‐olds. Memory, 8, 159–177. Wang, Q., & Ross, M. (2007). Culture and memory. In S. Kitayama and D. Cohen (Eds.), Handbook of cultural psychology (pp. 645–667). New York: Guilford. Warren, R. P. (1971). All the King’s Men. New York: Bantam. Weber, M. (1951). The religion of China. Glencoe, IL: Free Press. Wegner, D. M., Fuller, V., & Sparrow, B. (2003). Clever hands. Journal of Personality and Social Psychology, 85, 5–19. Wegner, D. M., Sparrow, B., & Winerman, L. (2004). Vicarious agency. Journal of Personality and Social Psychology, 86, 838–848. Weiss, P. (1995). The philosophy of Paul Weiss. Chicago: Open Court. Weisz, J., Rothbaum, F., & Blackburn, T. (1984). Standing out and standing in. American Psychologist, 39, 955–969. White, K., & Lehman, D. (2005). Culture and social comparison seeking. Personality and Social Psychology Bulletin, 31, 232–242. Wu, S. & Keysar, B. (in press). The eVect of culture on perspective taking. Psychological Science. Ybarra, O., & Trafimow, D. (1998). How priming the private self or collective self aVects the relative weights of attitudes and subjective norms. Personality and Social Psychology Bulletin, 24, 362–370. Yuki, M. (2003). Intergroup comparison versus intragroup relationships: A cross‐cultural examination of social identity theory in North American and East Asian cultural contexts. Social Psychology Quarterly, 66, 166–183. Yuki, M., Maddux, W. W., Brewer, M. B., & Takemura, K. (2005). Cross‐cultural diVerences in relationship‐ and group‐based trust. Personality and Social Psychology Bulletin, 31, 48–62. Zajonc, R. B. (1960). The process of cognitive tuning in communication. Journal of Abnormal and Social Psychology, 61, 159–167. Zajonc, R. B., & Markus, H. (1984). AVect and cognition: The hard interfrace. In C. Izard, J. Kagan, & R. Zajonc (Eds.), Emotions, cognitions, and behavior (pp. 73–102). Cambridge: Cambridge University Press.
UNCERTAINTY–IDENTITY THEORY
Michael A. Hogg
While I write this chapter, millions of people in the Darfur province of Sudan have been terrorized oV their land; the entire population of Iraq has little idea what the future of their country will be; survivors of hurricane Katrina are dispersed across the United States; people in Britain are anxious about immigration and are toying with the idea of supporting the British National Party; people in a small town in Tasmania wait to hear if members of their community have been found alive in a mine collapse; air travelers the world over have no idea what new security arrangements await them when they get to the airport; and we all wonder about the consequences of further escalation in the price of oil and of the standoV over Iran’s uranium enrichment program. The world is an uncertain place, it always has been, and these uncertainties can make it very diYcult to predict or plan our lives and to feel sure about the type of people we are. In this chapter, I describe how feelings of uncertainty, particularly about or related to self, motivate people to identify with social groups and to choose new groups with, or configure existing groups to have, certain properties that best reduce, control, or protect from feelings of uncertainty. I consider this uncertainty–identity theory to be a development of the motivational component of social identity theory. It addresses why, when, and how strongly people identify with groups, and why groups may have particular generic properties in certain contexts. Of particular relevance to contemporary postmodern society, uncertainty reduction theory provides an account of zealotry and the cult of the ‘‘true believer’’ in the thrall of ideology and powerful leadership—an account of conditions that may spawn extremism, a silo mentality, and a loss of moral or ethical perspective. In this chapter, I describe uncertainty–identity theory and some conceptual elaborations and applications, review direct and indirect empirical 69 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39002-8
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
70
MICHAEL A. HOGG
tests, and locate the theory in the context of related ideas and theories in social psychology. I start with a historical sketch of why, when, and how uncertainty–identity theory was developed, then go on to discuss uncertainty reduction as a motivation for human behavior. I then detail the process by which group identification reduces uncertainty and describe a program of studies showing that people who feel uncertain are more likely to identify and identify more strongly with groups. High‐entitativity groups are best equipped to reduce uncertainty through identification—entitativity moderates the uncertainty–identification relation. I discuss this idea and describe research that supports it, and then extend the analysis to deal with extremism and totalistic groups—describing how extreme uncertainty may encourage strong identification (zealotry, fanaticism, being a true believer) with groups that are structured in a totalistic fashion. Again I describe some research supporting this idea. The next section deals with extensions, applications, and implications of uncertainty–identity theory. I discuss the relation between depersonalization and self‐projection processes in uncertainty‐motivated group identification, and then, in a subsection entitled central members, marginal members, leaders, and deviants, I focus on the role of group prototypicality in uncertainty reduction processes. The role of trust, the relation between uncertainty, identity, and ideology, and the role of uncertainty in social mobilization are also discussed. The final section, before concluding comments, discusses other theories, approaches, and topics that deal with constructs related to those discussed by uncertainty–identity theory. Specifically, I discuss uncertainty as a state versus a trait, with a focus on the constructs of need for cognitive closure and uncertainty orientation; the role played by culture in uncertainty; and the relevance of terror management, compensatory conviction, self‐verification, and system justification. I. Historical Background Social identity theory has its origins in Tajfel’s early research on social categorization and his desire to provide a cognitive explanation of prejudice and discrimination (Tajfel, 1969)—an explanation that intentionally avoided attributing such behaviors to aberrant personality or interpersonal processes. Rather, Tajfel felt that prejudice and discrimination was a reflection of intergroup behavior in a particular social context on the part of people who identified with one of the groups. He famously defined social identity as ‘‘. . . the individual’s knowledge that he belongs to certain social groups together with some emotional and value significance to him of this group
UNCERTAINTY–IDENTITY THEORY
71
membership’’ (Tajfel, 1972, p. 292), and then, in collaboration with Turner focused on the conditions that produced more or less intergroup conflict and particular forms of intergroup behavior (Tajfel, 1974; Tajfel & Turner, 1979). This social identity theory of intergroup relations placed theoretical importance on the fact that groups compete over evaluatively positive distinctiveness because positive distinctiveness would be reflected in social identity and thus individual group members’ self‐concept. Ultimately, self‐ enhancement was considered a key motivation for social identity processes (Turner, 1975)—a motivation that, according to Billig (1985), social identity theory needed in order to be able to account for social change. As the principal motivation for social identity processes, self‐enhancement and self‐esteem became a key focus of social identity research—prompting Abrams and Hogg (1988) to postulate the self‐esteem hypothesis as an attempt to formalize the idea. From the outset, Abrams and Hogg warned against too closely mapping the group level construct of positive social identity onto the individual level construct of self‐esteem—a warning endorsed and supported by many others subsequently (Aberson, Healy, & Romero, 2000; Crocker & Luhtanen, 1990; Long & Spears, 1997; Luhtanen & Crocker, 1992; Rubin & Hewstone, 1998). As the waters surrounding the role of self‐esteem in social identity processes became increasingly muddy and crowded, Hogg and Abrams (1993) wondered whether other motivations might play a key role in social identity processes. Given that the process of social categorization was fundamental to social identity processes—a point fully elaborated by self‐categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), the social identity theory of the group—we felt that whatever motivated people to categorize was probably also a central motivation for social identity processes. This prompted us to suggest that because categorization reduces uncertainty by engaging schematic knowledge structures, uncertainty reduction might be that motivation (Hogg & Abrams, 1993). An epistemic motivation related to uncertainty was implicit in Tajfel’s early discussion of social categorization (Tajfel, 1972, 1974). Tajfel (1969, p. 92) believed that people engage in a ‘‘search for coherence’’ to preserve the integrity of the self‐image and that ‘‘This need to preserve the integrity of the self‐image is the only motivational assumption that we need to make in order to understand the direction that the search for coherence will take.’’ Tajfel and Billig (1974) suggested that one reason why people identify with minimal groups might be to impose structure on intrinsically uncertain circumstances. This idea was not pursued further—the motivational focus shifted to positive distinctiveness.
72
MICHAEL A. HOGG
A motivational role for uncertainty is also implicit in Turner and colleagues’ later self‐categorization theory (Turner, 1985; Turner et al., 1987)—but again not elaborated. More recently, social identity researchers have noted that disagreement with fellow group members would raise uncertainty and that conformity to group norms would reduce uncertainty by increasing consensus (Abrams, Wetherell, Cochrane, Hogg, & Turner, 1990; McGarty, Turner, Oakes, & Haslam, 1993; Turner, 1991). This analysis was restricted to social influence in groups and has not been elaborated to theorize uncertainty reduction as a basic motive for social identification itself and for forms of group and intergroup behavior as a whole (Smith, Hogg, Martin, & Terry, in press). Research and concepts relating to the basic motivational role of uncertainty reduction in social identity processes were formally integrated and finally published by Hogg (2000a; also see Hogg, 2001a; Hogg & Mullin, 1999). Two subsequent chapters touched on cultural dimensions (Hogg, 2006a) and implications of uncertainty reduction theory for organizations and corporate leadership (Hogg, 2007). More substantial developments focused on the types of groups and identities best suited to uncertainty reduction through group identification—discussing the role of entitativity and some implications for zealotry, ideology, and group extremism (Hogg, 2004, 2005a). Integrative statements and reviews of contemporary social identity theory incorporate uncertainty reduction as a motivational component, along with positive distinctiveness and self‐enhancement (Abrams, Hogg, Hinkle, & Otten, 2005; Hogg, 2003, 2005b, 2006b; Hogg & Abrams, 2003). It is important to note that uncertainty–identity theory was originally formulated with the circumscribed goal of developing a better understanding of the motivational underpinnings of social identity processes, in particular the fundamental process of identifying with a group in the first place. Conceived as a development of social identity theory, the focus was narrowly on the relation between feelings of uncertainty and group identification. Only subsequently were the wider implications of uncertainty–identity theory for the structure of groups and the nature of people’s membership in and attachment to groups explored. These implications locate uncertainty–identity theory in a wider literature on the consequences of uncertainty and on the causes of social extremism. Although in earlier publications I have referred to the ‘‘uncertainty reduction hypothesis’’ or to ‘‘uncertainty reduction theory,’’ uncertainty–identity theory is, in retrospect, a more accurate label as it correctly specifies the scope of the theory—a focus on the link between subjective uncertainty and group identification. It also avoids confusion with Berger and Calabrese’s (1975; see Bradac, 2001) interpersonal communication‐focused uncertainty reduction theory.
UNCERTAINTY–IDENTITY THEORY
73
II. Uncertainty Uncertainty–identity theory rests on the motivational tenet that feeling uncertain about ones perceptions, attitudes, values, or feelings is uncomfortable. At best it is an exhilarating challenge to be confronted and resolved—uncertainty is exciting and makes us feel edgy and alive, and delivers us a sense of satisfaction and mastery when we resolve such uncertainties. Meeting new people, going to parties, backpacking in exotic lands all raise uncertainty, but in a ‘‘good’’ way. At worst, uncertainty is highly anxiety provoking and stressful—it makes us feel impotent and unable to predict or control our world and what will happen to us in it. Being lost in a dangerous place, socially isolated, unclear about who we are, or how we fit in all raise uncertainty, in a ‘‘bad’’ way. Although we strive to resolve, manage, or avoid feeling uncertain, we do not do this all the time—some uncertainties we simply do not care about. For example, you may be profoundly uncertain about the rules of cricket but simply not care because none of your friends know any better or place much value on the game—in which case there is no motivation to resolve this uncertainty. However, if you moved to live in India or Australia where people are obsessed with cricket, your uncertainty about cricket would suddenly become a dramatically more serious aVair—you would now be highly motivated to resolve your uncertainty. Consistent with the cognitive miser or motivated tactician models of social cognition (Gollwitzer & Bargh, 1996; Nisbett & Ross, 1980), we only expend cognitive energy resolving those uncertainties that are important or matter to us in a particular context. The cricket example suggests one factor that imparts motivational impetus to feeling uncertain, and that is self‐relevance. We are particularly motivated to reduce uncertainty if, in a particular context, we feel uncertain about things that reflect on or are relevant to self, or if we are uncertain about self per se; about our identity, who we are, how we relate to others, and how we are socially located. Ultimately, people like to know who they are and how to behave and what to think, and who others are and how they might behave and what they might think. An important caveat about uncertainty reduction is that it is certainly more appropriate to talk about reducing uncertainty than achieving certainty. There is no such thing as absolute certainty—you cannot feel completely certain but only less uncertain (Pollock, 2003). This is reflected in the way that formal science rests on probability judgments, confidence intervals, and the quantification of uncertainty. In social psychology, we only feel ‘‘certain’’ that something is true if the probability of it occurring by chance is less than 1 in 20. For ordinary people very much the same thing holds. We act to some extent like naive scientists (Heider, 1958) and cognitive misers (Nisbett & Ross, 1980)
74
MICHAEL A. HOGG
in making a judgment that we are ‘‘suYciently’’ certain about something to desist from dedicating further cognitive eVort to uncertainty reduction. That uncertainty reduction requires work is consistent with the way that uncertainty is woven into the fabric of natural systems—natural systems tend toward entropy and chaos (i.e., uncertainty and disorganization), and it requires ‘‘work’’ to reduce entropy (Lorenz, 1993). Hence, uncertainty–identity theory is about reducing uncertainty rather than achieving certainty. A second important caveat is that the pursuit of uncertainty reduction does not rule out the possibility that individuals or groups sometimes embark on courses of action that in the short term increase uncertainty. One example is when the individual or group is confident that the experience of short‐term uncertainty is necessary to resolve more enduring contradictions and uncertainties that have arisen. This idea has parallels with the way that formal science progresses—periods of ‘‘normal science’’ where uncertainty is low and small contradictions accumulate but are concealed, punctuated by ‘‘scientific revolutions’’ where contradictions and uncertainties burst to the fore to sponsor a ‘‘paradigm shift’’ and subsequent reduction of uncertainty (Kuhn, 1962; Popper, 1959). Another example is when a current state of aVairs in one’s life or the society in which one lives is unbearable and a measured risk must be taken to improve things—change is risky and uncertain and therefore not undertaken lightly (Jost & Hunyady, 2002). A third caveat is that it is useful to distinguish between epistemic and aVective dimensions of uncertainty—knowing that you are uncertain about something, and feeling uncertain. The epistemic dimension is relatively focused, whereas the aVective dimension is more diVuse—thus we can feel generally uncertain, but not be sure exactly what about. The implication is that measures of feelings of uncertainty may remain relatively unchanged despite the fact that specific uncertainties are being epistemically resolved.
A. UNCERTAINTY AS A HUMAN MOTIVATION The idea that uncertainty plays a key role in motivating human behavior is certainly not new (Fromm, 1947), and there is evidence that uncertainty about one’s attitudes, beliefs, feelings, and perceptions, as well as uncertainty about oneself and other people, is aversive (Lopes, 1987; Sorrentino & Roney, 1986) and may be associated with physiological arousal in the hypothalamic‐ pituitary‐adrenal axis of the brain (Greco & Roger, 2003). Uncertainty about one’s relationships with others in a group may even provoke mistrust and paranoia (Kramer & Wei, 1999)—a process with profound consequences when transposed to intergroup relations.
UNCERTAINTY–IDENTITY THEORY
75
More than 50 years ago, Festinger (1954a,b) maintained that there is a ‘‘motivation to know that one’s opinions are correct and to know precisely what one is and is not capable of doing’’ (Festinger, 1954b, p. 217). Knowing one is correct is a critical human motivation that drives people to reduce their uncertainty by checking the validity of their perceptions against physical reality, or if physical reality checks are not possible by comparing their perceptions, beliefs, and attitudes with those of similar others or by calibrating their abilities through comparison with slightly dissimilar others (Suls & Wheeler, 2000). Uncertainty motivates a search for information to reduce uncertainty, though this search may be biased in the service of self‐enhancement (Wills, 1991). Although uncertainty can be reduced by making physical reality checks, for example touching the stove to confirm it is hot, our subjective certainties and thus knowledge of the world are overwhelmingly based on social consensus. Almost all nontrivial knowledge rests on social comparisons (Moscovici, 1976) that reflect agreement from fellow in‐group members and, depending on context, disagreement from out‐group members (Turner, 1975; also see Hogg, 2000b). The implication is that uncertainty reduction is an overwhelmingly social motivation that may be related to group membership. Epistemic motives related to uncertainty also have a high profile in contemporary social psychology. For example, they make an appearance in the recent explosion of research on the self and self‐motives (see Baumeister, 1998; Leary & Tangney, 2003; Sedikides & Gregg, 2003), which considers the quest for self‐definition to be a persistent and central feature of human existence (Gaertner, Sedikides, & Graetz, 1999; Sedikides & Strube, 1997). Uncertainty‐related motives are particularly central to research on uncertainty orientation (Sorrentino & Roney, 1999), compensatory conviction (McGregor & Marigold, 2003; McGregor, Zanna, Holmes, & Spencer, 2001), and cognitive closure and closed mindedness (Kruglanski, 1989, 2004). There is also a possible role for uncertainty‐related motives in terror management theory (Pyszczynski, Greenberg, & Solomon, 1999, 2004). Many of these theories and constructs are discussed later in this chapter.
B. UNCERTAINTY REDUCTION AND THE PURSUIT OF MEANING What is the relation between uncertainty reduction and the pursuit of meaning? Many scholars believe that the primary human motive is the search after meaning (Bartlett, 1932). For example, existentialist philosophers
76
MICHAEL A. HOGG
(e.g., Kierkegaard, Heidegger) and other existentialists (Camus, Sartre) believe that the key feature of the human condition is a search for meaning in life. People are meaning‐makers who seek to construct a coherent worldview. In his overarching account of human motivation, Maslow (1987) places meaning‐making at the apex of the hierarchy of needs. However, others do not aVord such status to meaning. For example, Sedikides and Strube (1997) feel that self‐esteem has greater impact as a human motive, and Vignoles and associates (Vignoles, Regalia, Manzi, Golledge, & Scabini, 2006) aVord equal status, specifically in the area of identity construction, to five other motives—self‐esteem, continuity, distinctiveness, belonging, and eYcacy. In their meaning maintenance model, Heine, Proulx, and Vohs (2006) argue that people are meaning‐makers driven to establish associative frameworks that (1) tie together elements of the world, (2) tie together elements of themselves, and (3) most importantly bind self to the world. This view of meaning seems, however, to focus less on meaning than on associative links that one has confidence in and one is certain about, suggesting that reduced uncertainty is critical and may motivationally underpin meaning. Uncertainty and meaningfulness are closely associated and are generally inversely related. To the extent that one feels less uncertain about something, its meaning is clearer, and vice versa. This relation begs the questions of which is motivationally primary. Do we pursue meaning and if things are meaningful feel certain, or do we pursue certainty and if things are certain feel a sense of meaningfulness? Instead of uncertainty reduction should we actually be talking about the pursuit of meaning? Uncertainty reduction is probably more basic, in an evolutionary and comparative sense—animals are more likely to experience and act on certainty than meaningfulness. Only humans engage in a discourse of meaning. However, general models of human motivation, such as Maslow’s (1987) hierarchy of needs, place meaning at the apex—most people do not have the luxury of pursuing meaning much of the time, and are more involved on a day‐to‐day basis in feeling certain about themselves and their place in the world. Reduction of uncertainty may also be more proximally associated with human social cognition. To the extent that we approach life as naive scientists, cognitive misers, or motivated tacticians (Taylor, 1998; also see above), context‐specific feelings of reduced uncertainty about aspects of our life require less cognitive eVort than the construction of a meaningful subjective world. In addition, the causal attribution processes that we can use to make the world more meaningful rest ultimately on judgments, albeit often inaccurate, of certainty about a causal association (Trope & Gaunt, 2003). Certainty and meaning are associated, but it is the feeling of uncertainty that motivates us, rather than the knowledge that something has little meaning.
UNCERTAINTY–IDENTITY THEORY
77
C. TYPES OF UNCERTAINTY Uncertainty is a feeling that can be very wide ranging and diVuse, for example feeling uncertain about one’s future, or very specific and focused, for example feeling uncertain about what to wear to a party. Uncertainty can also vary in the degree to which it reflects on or relates to self‐conception in a particular context. As mentioned earlier, uncertainty about or related to self is likely to have the greatest motivational force because the self is the critical organizing principle, referent point, or integrative framework for perceptions, feelings, and behaviors. It is this self‐uncertainty that is most directly implicated in social identity processes. Uncertainty is triggered by the social context in which one finds oneself— we all feel uncertain at diVerent times. However, there is no doubt that some of us tend to feel more uncertain about more things more often and that some of us simply feel more uncomfortable with uncertainty than others. This can be an enduring disposition (e.g., authoritarian personalities—Adorno, Frenkel‐Brunswik, Levinson, & Sanford, 1950), a cultural form (e.g., uncertainty avoidance—Hofstede, 1980), or a passing phase in our lives (e.g., adolescence, bereavement). Uncertainty–identity theory focuses on uncertainty as a context‐induced state. It is produced by contextual factors that challenge people’s certainty about their cognitions, perceptions, feelings, and behaviors, and ultimately, certainty about and confidence in their sense of self. Although the locus of uncertainty is to be found overwhelmingly in the social context, and therefore anyone is prone to uncertainty, biographical factors may have some influence on people’s general orientation toward uncertainty and the reduction of uncertainty. It is this latter personality and individual diVerences perspective that has tended to dominate the social psychology literature on uncertainty. The idea that some people are less tolerant of uncertainty than others is an old one, for example Adorno et al.’s (1950) description of the authoritarian personality and Rokeach’s (1960) notion of a dogmatic or closed‐minded personality. There is also evidence that people vary in their need for structure or closure and their fear of invalidity. People who need structure or closure are more concerned to reduce uncertainty quickly than to be correct, whereas people who fear invalidity are able to tolerate uncertainty while they engage in a prolonged search for validity (Kruglanski, 1989; Kruglanski & Webster, 1996; Neuberg & Newson, 1993; Webster & Kruglanski, 1998; also see Kruglanski, 2004). Relatedly, Sorrentino and associates have explored individual diVerences in uncertainty orientation (Sorrentino, Hodson, & Huber, 2001; Sorrentino & Roney, 1999; Sorrentino & Short, 1986). Uncertainty‐ oriented people seek out information that may raise uncertainty, and work
78
MICHAEL A. HOGG
on the resolution of uncertainty to satisfy a self‐assessment motive. These people are ‘‘need to know,’’ scientific, or investigative types. Certainty‐ oriented people are concerned with self‐verification and the maintenance of existing beliefs; they avoid situations of uncertainty and if confronted by uncertainty, they defer to others or use heuristics to resolve uncertainty quickly. Finally, there are more broadly related constructs that describe individual diVerences in the complexity and number of explanations people have of other people (attributional complexity; Fletcher, Danilovics, Fernandez, Peterson, & Reeder, 1986), in how much people like to think deeply about things (need for cognition; Cacioppo & Petty, 1982), and in the complexity of people’s cognitive processes and representations (cognitive complexity; Crockett, 1965). People also diVer in self‐concept clarity, the extent to which self‐beliefs are clearly and confidently defined, internally consistent, and stable (Campbell, 1990; Campbell et al., 1996); self‐complexity, the number of diVerent or independent dimensions that underlie self‐conception (Linville, 1987); and compartmentalization of the self (Showers, 1992). In contrast to these individual diVerence perspectives on uncertainty, perspectives in communication and organization sciences treat uncertainty as context contingent. From a communication perspective, people communicate to reduce uncertainty, and eVective communication requires a degree of interpersonal certainty (Berger, 1987; Berger & Calabrese, 1975; Ford, Babrow, & Stohl, 1996; see Bradac, 2001). Uncertainty in interpersonal (and cross‐cultural—Gudykunst, 1985) communicative contexts is maladaptive and can produce adverse reactions. Uncertainty reduction is also an important motivational element in theories of organizational socialization (Lester, 1987; Saks & Ashforth, 1997). Organizational newcomers experience uncertainty that motivates them to seek information, through interaction and communication with superiors and peers, to reduce their uncertainty. Many organizations implement formal induction procedures to achieve this socialization goal. One situation in which uncertainty may be particularly elevated is when there is an organizational merger or acquisition; employees are uncertain about their employment future, their role in the new organization, and their organizational identity. One way in which employees may reduce this uncertainty is by identifying strongly, often more strongly than before, with their premerger organization. This reduces self‐related uncertainty, but it may also inhibit identification with and commitment to the overarching new organization. Mergers and acquisitions are notoriously unsuccessful from an organizational identity point of view (Terry, Carey, & Callan, 2001). Uncertainty may play a key role in this. More generally, the contemporary world of work
UNCERTAINTY–IDENTITY THEORY
79
may be a particularly potent and enduring source of uncertainty (Hogg, 2007). III. Social Identity Feelings of uncertainty have diVerent causes and diVerent foci. Uncertainty– identity theory focuses on context‐induced feelings of uncertainty that are about self or things that relate to, reflect on, or matter to self. To the extent that a particular context that induces uncertainty endures, for example a long‐lasting economic crisis, uncertainty and attempts to reduce or fend oV uncertainty may endure. There may be individual diVerences in how much uncertainty people feel in a given context and in how people respond to uncertainty; however this is treated, to use a statistical metaphor, as error variance—it is not the focus of uncertainty–identity theory. This orientation toward personality and individual diVerences is consistent with the group‐ focused metatheory that informs social identity theory (Abrams & Hogg, 2004; Turner, 1999; Turner & Onorato, 1999; see Hogg, in press). Feelings of uncertainty about or reflecting on self can be resolved in many diVerent ways. However, the crux of uncertainty–identity theory is that group identification is one of the most potent and eVective ways to do this. From a social identity perspective (for overviews see Hogg, 2003, 2005b, 2006b), group identification is produced by self‐categorization (Turner et al., 1987). Human groups are social categories that we cognitively represent as prototypes—prototypes that embody all and any attributes that define the category and distinguish it from other categories in a specific context. One’s prototype of a group can describe members’ perceptions, beliefs, attitudes, values, feelings, and behaviors. The prototype of a group we belong to has prescriptive properties in describing how we ought to behave as a group member. Prototypes obey the metacontrast principle—they maximize the ratio of intergroup diVerences to intragroup diVerences, and thus perceptually accentuate similarities within groups and diVerences between groups (cf. Tajfel, 1959). This principle ensures that the prototype we have of a specific group is influenced, more or less dramatically, by what group it is being compared to and for what purpose. When we categorize someone as a member of a specific group, we assign the group’s attributes to varying degree to that person. We view them through the lens of the prototype of that group; seeing them not as unique individuals but as more or less prototypical group members—a process called depersonalization. When we categorize others, in‐group or out‐group members, we stereotype them and have expectations of what they think
80
MICHAEL A. HOGG
and feel and how they will behave. When we categorize ourselves, self‐ categorization, exactly the same process occurs—we assign prescriptive in‐ group attributes to ourselves, we autostereotype, conform to group norms, and transform our self‐conception. In this way, group identification very eVectively reduces self‐related uncertainty. It provides us with a sense of who we are that prescribes what we should think, feel, and do. Because self‐categorization is inextricably linked to categorization of others, it also reduces uncertainty about how others will behave and what course social interaction will take. It also provides consensual validation of our worldview and sense of self, which further reduces uncertainty. Because people in a group tend to have a shared prototype of ‘‘us’’ and a shared prototype of ‘‘them,’’ our expectations about the prototype‐based behavior of others often tend to be confirmed, and our fellow group members agree with our perceptions, beliefs, attitudes, and values and approve of how we behave. Clearly, identification can eVectively reduce uncertainty, and can protect one from uncertainty. The implication is that uncertainty reduction motivates group identification—we identify with groups in order to reduce or protect ourselves from uncertainty. When people feel uncertain about themselves or things reflecting on self, they ‘‘join’’ new groups (e.g., sign up as a member of an environmental group), identify with or identify more strongly with existing self‐inclusive categories (e.g., one’s nation), or identify with or identify more strongly with groups that they already ‘‘belong’’ to (e.g., one’s work team). Uncertainty reduction provides a motivational context and impetus for making specific social categorizations contextually salient as the basis of social identification. Uncertainty reduction directly frames the way in which we draw on chronically and situationally accessible categorizations and investigate their comparative and normative fit (Oakes, 1987) in order to render a particular social categorization psychologically salient. The very notion that an accessible categorization needs to fit implies that it reduces feelings of uncertainty about the social context and our place within it. The uncertainty–identity theory conception of the relation between uncertainty and group identification represents a relatively hydraulic model of group motivation. Uncertainty, however induced, mobilizes one to psychologically identify and is reduced by identification. However, feelings of uncertainty are multiply determined and can be addressed in many diVerent ways. Identification is only one way to address uncertainty, but one that is particularly eVective in the case of self‐related uncertainties. Feelings of uncertainty can also be fleeting. As soon as one uncertainty is reduced, one’s mind is assailed by new uncertainties or we seek out new ones to
UNCERTAINTY–IDENTITY THEORY
81
resolve. The epistemic/aVective disjunction described above is relevant here. Epistemic resolution of uncertainty through identification may not map directly onto aVective change—you may still feel uncertain, but now about other things.
A. BASIC HYPOTHESIS TESTS The most basic prediction that can be made from uncertainty–identity theory is that the more uncertain people are the more likely they are to identify, and to identify more strongly, with a self‐inclusive social category. The first tests of this hypothesis used the minimal group paradigm. The minimal group paradigm (Bourhis, Sachdev, & Gagnon, 1994; Diehl, 1990) is particularly appropriate for at least three reasons: (1) it occupies a pivotal role in the development of social identity theory; (2) it is implausible that people discriminate on the basis of a minimal categorization because of a self‐enhancement motivation (cf. Cadinu & Rothbart, 1996); and (3) minimal group studies are situations of high uncertainty due to the novelty of the situation and the strangeness of the resource distribution task, and therefore participants readily use the minimal categorization to reduce uncertainty about themselves and how they should behave. The first study (Hogg & Grieve, 1999) was a 2 (categorization) 2 (uncertainty) experiment in which student participants (N ¼ 151) were not categorized or were explicitly randomly categorized into X‐ and Y‐groups (the categorization variable). This was done under normal minimal group conditions, which embodied relatively high subjective uncertainty, or under conditions where uncertainty was lowered by giving participants three practice trials on the standard minimal group allocation matrices (the uncertainty variable). Uncertainty was checked after the manipulation (three items measuring how uncertain they felt about the allocation task), and measured again in the final questionnaire—the manipulation was only marginally eVective. On a composite measure of in‐group bias (three items, ¼ .87), we found, as predicted, that only those participants who had been categorized under high uncertainty expressed significant in‐group bias and significantly greater bias than participants in the other three conditions. By reducing uncertainty, we had, as predicted, eliminated the minimal categorization aVect. The rationale was that by reducing uncertainty, we had reduced the motivation to identify and thus reduced in‐group bias. Participants categorized under uncertainty also showed significant reduction in uncertainty over time, whereas uncategorized participants did not. We were able to conclude that in minimal group studies, social categorization per se does not
82
MICHAEL A. HOGG
produce intergroup discrimination. It is social categorization under conditions of subjective uncertainty that motivates participants to define themselves in terms of the minimal categorization, which in turn generates diVerential intergroup perceptions, feelings, and behavior (i.e., discrimination). We repeated this experiment, with methodological refinements (Grieve & Hogg, 1999, Experiment 1, N ¼ 119). The key diVerences were that low‐ uncertainty participants completed 12 rather than only 3 practice matrices (the uncertainty checks revealed this manipulation to be eVective), and group identification was measured by 10 items ( ¼ .87), adapted from Hains, Hogg, and Duck (1997) and Hogg and Hains (1996).1 As in Hogg and Grieve (1999), there was significant bias (three items, ¼ .84) only among those participants who were categorized under uncertainty, but we also found that these participants reported identifying significantly more strongly than other participants. Our third study (Grieve & Hogg, 1999, Experiment 2, N ¼ 105) was a replication of Grieve and Hogg (1999, Experiment 1), involving a diVerent manipulation of uncertainty. We wanted to manipulate uncertainty in a way that was separate from the subsequent allocation task used to express discrimination, and we wanted to elevate uncertainty as well as to reduce it. In a 2 (categorization) 2 (uncertainty) design, participants had their uncertainty elevated by having them write down what they thought was happening in each of five ambiguous pictures (from the Thematic Apperception Test), or lowered by writing down what they thought was happening in each of five unambiguous pictures (e.g., photos of everyday life). To check on this manipulation, they indicated how uncertain they felt about each of their descriptions—the manipulation was highly eVective. Participants were then explicitly categorized or not categorized, as in Grieve and Hogg, Experiment 1, and completed minimal group allocation matrices and measures of group identification. There was significant in‐group bias (three items, ¼ .63) only among participants who were categorized under elevated uncertainty
1 Group identification can be measured in many ways. In my own research, including the studies of uncertainty reported in this chapter, I have used between 3 and 10 items which generally ask participants to indicate their desire to get to know the group’s members, to join the group, and to stand up for the group; and to indicate the extent to which they identify with the group, like its members and the group as a whole, perceive personal similarity to the group and its members, feel ties to other members, and feel they fit in the group (1 not very much, 9 very much). Across literally dozens of studies over a period of 20 years, these kinds of measures form highly reliable scales (mostly with reliabilities greater than .90) (e.g., see Grieve & Hogg, 1999; Hains et al., 1997; Hogg & Grieve, 1999; Hogg & Hains, 1996; Hogg & Svensson, 2006; Hogg et al., 2006, 2007; Mullin & Hogg, 1998, 1999; Reid & Hogg, 2005).
UNCERTAINTY–IDENTITY THEORY
4
83
3.43b*
In-group bias
3 2 Not categorized Categorized
1.22a 1 0 −0.32a −1
−0.26a
Low
High Uncertainty
Fig. 1. Grieve and Hogg (1999, Experiment 2): EVect of categorization and uncertainty on in‐group bias, F(1, 104) ¼ 6.21, p < .05. Note: In‐group bias is a three‐item scale ( ¼ .63) that can take values between 12 (completely favoring the out‐group) and þ12 (completely favoring the in‐group). Means not sharing a subscript are significantly diVerent by simple main eVects. Asterisk indicates significantly diVerent from zero.
(Fig. 1), and these people also identified (10 items, ¼ .91) significantly more strongly than all other participants (Fig. 2). Continuing to explore diVerent aspects of uncertainty, a fourth study (Mullin & Hogg, 1998) modified the Grieve and Hogg methodology. In addition to the categorization manipulation, there were two uncertainty variables, task and situational, producing a 2 2 2 design (N ¼ 96). Task uncertainty was manipulated in a similar way to previous studies. Low‐ uncertainty participants were given six practice matrices, and told to do as many as they needed to feel completely certain about the task; high‐ uncertainty participants were not given practice matrices. Situational uncertainty was a dichotomous subject variable. High‐uncertainty participants had not yet taken part in an experiment; low‐uncertainty participants had already been in at least five experiments. We felt that situational uncertainty might be a compelling motivation for identification because it related more directly to the relation between self and others in the social setting. Checks confirmed the eVectiveness of both forms of uncertainty. As predicted, there was a significant interaction between categorization and task uncertainty, and between categorization and situational uncertainty on both in‐group bias (three items, ¼ .93) and identification (five items, ¼ .82). Participants categorized under either task or situational uncertainty identified more strongly than other participants, and were the only ones to express
84
MICHAEL A. HOGG
9 8
Identification
7 5.93c 6 4.67b
5 4
3.68a
3.60a
Not categorized Categorized
3 2 1 Low
High Uncertainty
Fig. 2. Grieve and Hogg (1999, Experiment 2): EVect of categorization and uncertainty on group identification, F(1, 104) ¼ 7.61, p > .01. Note: Identification is a 10‐item scale ( ¼ .91) that can take values between 1 and 9. Means not sharing a subscript are significantly diVerent by simple main eVects.
significant bias. There was some support for reduced uncertainty associated with in‐group identification and bias. The four studies just described provide robust support, across diVerent operationalizations of uncertainty, for the key uncertainty–identity hypothesis that people are more likely to identify with a self‐inclusive category, and identify more strongly, when they are uncertain. However, there is less robust evidence that identification reduces reported uncertainty. Hogg and Grieve (1999) and Mullin and Hogg (1998) did find evidence that uncertainty‐ induced identification reduced uncertainty, but Grieve and Hogg (1999, Experiment 1) did not. Grieve and Hogg (1999, Experiment 2) did not have postidentification measures of uncertainty. This is one aspect of uncertainty– identity theory that requires further investigation. The evidence for identification reducing uncertainty may be less robust for methodological reasons (perhaps the use of repeated measures inhibited participants from reporting a change in uncertainty—they may not have wanted to seem uncertain about how uncertain they were), or for conceptual reasons (I discussed earlier how epistemic resolution of uncertainty through identification may not map directly onto aVective change—you may still feel uncertain, but now about other things). It is also worth noting that other less direct studies of uncertainty and identification have found evidence for identification reducing uncertainty (McGregor, Nail, Marigold, & Kang, 2005).
UNCERTAINTY–IDENTITY THEORY
85
B. RELATION OF UNCERTAINTY TO SELF‐ENHANCEMENT AND SELF‐ESTEEM Although the studies described in the previous section suggest that uncertainty motivates identification, there is a possible alternative explanation. Perhaps the manipulations of uncertainty also aVected self‐esteem (it is plausible to speculate that being made to feel uncertain may also depress self‐esteem) and people identified to elevate self‐esteem rather than reduce uncertainty—in which case uncertainty per se was not motivating identification. There are at least two reasons to be circumspect about this possibility: (1) social identity research on the self‐esteem hypothesis shows that although identification can elevate self‐esteem, depressed self‐esteem typically does not motivate identification (Aberson et al., 2000; Rubin & Hewstone, 1998) and (2) manipulations of self‐uncertainty do not necessarily aVect self‐esteem (McGregor et al., 2001). Nevertheless it seemed prudent to investigate. We conducted two studies (Hogg & Svensson, 2006). The first was a computer‐mediated minimal group study in which the three variables of uncertainty, group relevance, and opportunity to self‐aYrm were manipulated in a 2 2 2 design. Uncertainty was manipulated by having participants (N ¼ 168) perform an easy/unambiguous or diYcult/ambiguous eyewitness task, followed by a manipulation check and measures of state self‐esteem. Participants were then categorized as members of a group that was more, or less, relevant to them in terms of their career interests. The key‐ dependent measure was group identification (eight items, ¼ .86). However, in order to manipulate self‐aYrmation, half the participants were given the opportunity to self‐aYrm prior to indicating their identification. Drawing on self‐aYrmation theory (Sherman & Cohen, 2006; Steele, 1988), we argued that if uncertainty operated via self‐esteem, then the eVect of uncertainty on identification would disappear in the self‐aYrmation condition. This did not happen. Participants identified significantly more strongly when uncertain, particularly when the group was self‐relevant, irrespective of whether they aYrmed. Furthermore, this eVect remained significant when state self‐esteem, measured immediately after the uncertainty manipulation, was statistically partialed out of the analysis. For the second study, participants (N ¼ 101) were in real face‐to‐face small groups—group relevance was high throughout. They individually performed the eyewitness task from Study 1 (uncertainty manipulation), answered uncertainty checks (self‐esteem was not measured), and then indicated how much they identified with their group (eight items, ¼ .86)—as in Study 1 half the participants self‐aYrmed first (self‐aYrmation manipulation). Once again participants identified more strongly under high than low uncertainty, which was not aVected by self‐aYrmation.
86
MICHAEL A. HOGG
9 8 Identification
7 6 5
5.49b
5.67b
5.19ab
4.56a
Low status High status
4 3 2 1 Low
High Uncertainty
Fig. 3. Reid and Hogg (2005, Experiment 1): EVect of status and uncertainty on group identification, F(1, 59) ¼ 7.30, p ¼ .009. Note: Identification is a nine‐item scale ( ¼ .87) that can take values between one and nine. Means not sharing a subscript are significantly diVerent by simple main eVects.
The Hogg and Svensson (2006) studies show that uncertainty on its own motivates identification, independent of self‐esteem or self‐enhancement considerations. However, what happens when one feels uncertain but the available group has low status that mediates adverse self‐evaluation—does the need to reduce uncertainty win over the pursuit of positive identity, or vice versa? Reid and Hogg (2005) conducted two minimal group studies to investigate this. In Study 1 participants (N ¼ 64) were categorized, ostensibly as over‐ or underestimators (a frequently used minimal criterion), and given feedback that their group had relatively high or low status (did better or worse than other groups in the experiment on a perceptual task)—the group status manipulation. Uncertainty was manipulated by having participants perform an object‐counting task that was easy (very few objects) or diYcult (too many to count, they could only guess). In anticipation of a group activity, participants then indicated how strongly they identified with their group (nine items, ¼ .87). The key finding (Fig. 3) was that even when the group had low status, uncertainty significantly increased identification—thus, uncertainty reduction concerns prevailed over self‐enhancement considerations. Study 2 was a replication of Study 1 but with an additional variable manipulating the extent to which participants (N ¼ 210) felt they were a good fit with, and prototypical of, the group [cf. the notion of group relevance in Hogg and Svensson’s (2006) first study]. The prediction was that the eVects obtained in Study 1 would only emerge in the high prototypicality condition.
UNCERTAINTY–IDENTITY THEORY
87
This is precisely what happened. Participants identified (nine items, ¼ .91) significantly more strongly with the low‐status group when uncertainty was high than low, but only when they felt they fit the group well. Once again, uncertainty reduction prevailed over self‐enhancement. The Reid and Hogg studies show that uncertainty is a suYciently strong motive for identification that it can override group status and self‐enhancement concerns. This has obvious implications for an understanding of why minority or stigmatized groups may not rise up to challenge the status quo. The prospect of change raises uncertainty and thus encourages members to maintain or strengthen their identification, even though the group has low status. Perhaps they go on to justify their behavior by engaging in a discourse of system justification (cf. Jost & Hunyady, 2002). C. UNCERTAINTY RELEVANCE AND GROUP RELEVANCE Not all uncertainty motivates. As described earlier, if you feel uncertain but simply do not care, then the uncertainty does not motivate behavior. For uncertainty to motivate, it must matter to you that you are uncertain. Typically uncertainty about something important to you that reflects on or is focused on self matters and therefore motivates. To test this idea, Mullin and Hogg (1999) had students participate in a 2 (categorization) 2 (task uncertainty) 2 (task importance) minimal group experiment (N ¼ 128). Participants were randomly categorized as group members (‐ versus ‐group) or identified as individuals, after they had been given feedback to raise or lower feelings of uncertainty about the validity of their attitudes toward low‐importance or high‐importance issues (e.g., trivial commodity preferences versus important lifestyle and health preferences). Checks confirmed the eVectiveness of the importance and uncertainty manipulations. As predicted, participants who were categorized under high uncertainty about important issues identified (five items, ¼ .85) significantly more strongly than did participants in all seven other conditions. These people also expressed significantly greater desire than others to interact with and learn more about members of their group. This study confirms that uncertainty must matter in order for it to sponsor identification.
IV. Entitativity Two studies were described above (Hogg & Svensson, 2006, Experiment 1; Reid & Hogg, 2005, Experiment 2) showing that uncertainty is more likely to lead to identification, or leads to stronger identification, if the group is
88
MICHAEL A. HOGG
relevant to self‐definition. This begs the broader question of what kinds of groups, or what properties of groups, are best equipped to reduce uncertainty through identification. The answer proposed by uncertainty–identity theory is high‐entitativity groups (Hogg, 2004, 2005a). Entitativity is that property of a group, resting on clear boundaries, internal homogeneity, social interaction, clear internal structure, common goals, and common fate, which makes a group ‘‘groupy’’ (Campbell, 1958; Hamilton & Sherman, 1996). Groups can vary quite widely in entitativity from a loose aggregate to a highly distinctive and cohesive unit (Hamilton, Sherman, & Rodgers, 2004; Lickel et al., 2000). Generally, entitativity is more a matter of perceived interdependence and mutual social influence than mere similarity or homogeneity (Lickel, Rutchick, Hamilton, & Sherman, 2006). Group identification reduces uncertainty because it provides a clear sense of self that prescribes behavior and renders social interaction predictable. An unclearly structured low‐entitativity group that has indistinct boundaries, ambiguous membership criteria, limited shared goals, and little agreement on group attributes will do a poor job of reducing or fending oV self or self‐related uncertainty. In contrast, a clearly structured high‐entitativity group with sharp boundaries, unambiguous membership criteria, highly shared goals, and consensus on group attributes will do an excellent job. Identification via self‐categorization reduces uncertainty because self is governed by a prototype that prescribes cognition, aVect, and behavior. Prototypes that are simple, clear, unambiguous, prescriptive, focused, and consensual are more eVective than those that are vague, ambiguous, unfocused, and dissensual. Clear prototypes, such as the former, are more likely to be grounded in high‐than low‐entitativity groups. In addition, essentialism, the tendency to attribute properties of individuals or groups to invariant underlying qualities or essences (Haslam, Rothschild, & Ernst, 1998; Medin & Ortony, 1989; Yzerbyt, Judd, & Corneille, 2004), may be a positive function of perceived entitativity (Rothbart & Taylor, 1992). Thus, the more ‘‘groupy’’ a group is perceived to be, the more people view its attributes to be immutable and fixed essences that, for example, reflect biology and genetics. Because this process imparts the illusion of immutability to groups, it further renders entitative groups highly eVective in reducing and fending of uncertainty. From uncertainty–identity theory, the clear prediction is that although under uncertainty, especially self‐uncertainty, people will identify with groups, they will show a strong preference for high‐entitativity groups. People will seek out highly entitative groups with which to identify or they will work to elevate, subjectively or in reality, the entitativity of groups to which they already belong.
UNCERTAINTY–IDENTITY THEORY
89
A. STUDIES OF THE ROLE OF ENTITATIVITY IN UNCERTAINTY‐INDUCED IDENTIFICATION As a first test of this idea, Jetten, Hogg, and Mullin (2000) conducted two minimal group studies where judgmental uncertainty and group homogeneity (a proxy for entitativity) were orthogonally manipulated, and group identification and intergroup perceptions were measured. Uncertainty predicted identification, as in previous studies, but this was not moderated by homogeneity. Uncertainty and homogeneity did interact on a measure of in‐ group–out‐group stereotype diVerentiation—participants who were uncertain showed greater in‐group–out‐group stereotype diVerentiation when the in‐group was more homogeneous. However, stereotypic diVerentiation is a perceptual eVect that is not the same as identification and belonging, and in‐group homogeneity, a similarity perception, is not the same as in‐ group entitativity, and may actually be the least important component of entitativity (Hamilton et al., 2004; Lickel et al., 2006). The role of entitativity was better investigated in a series of four studies by Castano, Yzerbyt, and Bourguignon (2003), who manipulated diVerent aspects of entitativity and measured identification with the European Union (EU). Entitativity increased identification but only, or more strongly, among those with less extreme attitudes toward the EU. Uncertainty was not measured or manipulated—it can only be speculated that those with less extreme attitudes may have been less self‐conceptually uncertain. The main message of this research is that people identify more strongly with high‐entitativity groups (Castano, 2004; Yzerbyt, Castano, Leyens, & Paladino, 2000), which is consistent with Lickel et al.’s (2000) finding that in‐group entitativity perceptions were significantly correlated with reported importance of the group to self. To address some of the limitations of these earlier, indirect, and incomplete tests of the role of entitativity in uncertainty‐induced group identification, we (Hogg, Sherman, Dierselhuis, Maitner, & MoYtt, 2007) conducted two studies in which entitativity was measured or manipulated and uncertainty, specifically self‐uncertainty, was primed. Study 1 was a field experiment (N ¼ 114) in which we measured how entitative student participants felt the political party they supported was, and then primed uncertainty (high versus low) by asking them to focus on things that made them feel uncertain/certain about themselves and to write down those things that made them feel most uncertain/certain. The dependent measure was a nine‐item measure of identification with their political party ( ¼ .93). As predicted, under high uncertainty identification increased with increasing perceived entitativity (Fig. 4). Under low uncertainty, there was no significant relation between identification and entitativity. Another way to look at this is that when the group was high in
90
MICHAEL A. HOGG
9 8 Identification
7 6 High uncertainty Low uncertainty
5 4 3 2 1 Low
High Entitativity
Fig. 4. Hogg et al. (2007, Study 1): Group identification as a function of perceived entitativity for high‐ and low‐self‐uncertainty participants, ¼ .21, t ¼ 2.42, p ¼ .017. Note: Identification is a nine‐item scale ( ¼ .93) that can take values between one and nine.
entitativity, uncertain participants identified more than less uncertain participants, and vice versa for low‐entitativity groups. Study 2 was a laboratory experiment in which participants (N ¼ 89) took part in a computer‐mediated decision‐making group. On the basis of an initial test, they were given detailed controlled feedback that described their group as being high or low in entitativity in terms of its attributes and the way the group would approach the decision‐making task. Following an entitativity check, participants completed the uncertainty prime as in Study 1, and then filled out an eight‐item measure of group identification ( ¼ .94). Participants who were self‐uncertain and in a high‐entitativity group identified significantly more strongly than participants in all other conditions (Fig. 5). These direct tests confirm the predicted moderating role of entitativity in uncertainty‐induced identification. A particular strength is the direct manipulation of self‐uncertainty—previous uncertainty–identity research had manipulated judgmental or perceptual uncertainty that reflects on self, rather than self‐uncertainty directly. Another feature of this manipulation is that it may not influence self‐esteem—a similar priming procedure used by McGregor et al. (2001) did not significantly aVect self‐esteem. These studies also extend Pickett’s research. Pickett and colleagues (Pickett & Brewer, 2001; Pickett, Silver, & Brewer, 2002) argue that the relation between identification and group entitativity/inclusiveness is moderated by motivational states, and show that one motivational moderator is
UNCERTAINTY–IDENTITY THEORY
91
9 8 6.98b Identification
7 6 4.61a
5
4.72a Low uncertainty
3.93a 4
High uncertainty
3 2 1 Low
High Entitativity
Fig. 5. Hogg et al. (2007, Study 2): EVect of entitativity and uncertainty on group identification, F(1, 84) ¼ 8.05, p ¼ .006. Note: Identification is an eight‐item scale ( ¼ .94) that can take values between one and nine. Means not sharing a subscript are significantly diVerent by simple main eVects.
the need for assimilation/diVerentiation as specified by optimal distinctiveness theory (Brewer, 1991). Our studies show that self‐conceptual uncertainty, as specified by uncertainty–identity theory, is another powerful moderator of the entitativity–identification relation.
V. Social Extremism and Totalistic Groups The story so far is that subjectively important feelings of uncertainty about self or about matters that relate to or reflect on self motivate people to identify with groups, particularly groups that are relevant to self and are high in entitativity. This process may go one step further (Hogg, 2004, 2005a). When self‐uncertainty is acute or enduring, people may identify very strongly with groups that are not merely entitative but extreme—totalistic groups (Baron, Crawley, & Paulina, 2003). These groups have closed and carefully policed boundaries, uniform attitudes, values and membership, and inflexible customs. They are rigidly and hierarchically structured with a clearly delineated chain of legitimate influence and command, and substantial intolerance of internal dissent and
92
MICHAEL A. HOGG
criticism. Such groups are often ethnocentric—they are inward looking, and suspicious and disparaging of outsiders. They engage in relatively asocial and overly assertive actions. Their attributes resemble narcissism (cf. Baumeister, Smart, & Boden, 1996) expressed at the collective level (Montoya, Rosenthal, & Pittinsky, 2006; Pittinsky, Montoya, & Rosenthal, 2006)—characterized by grandiosity, self‐importance, envy, arrogance, haughtiness, entitlement, exploitativeness, excessive admiration, lack of empathy, fantasies of unlimited success, and feelings of special/unique/high status. These are ‘‘extreme’’ groups that furnish members with an all‐embracing, rigidly defined, exclusive, and highly prescriptive social identity and sense of self. In this way, uncertainty‐induced identification may underpin zealotry, extremism, and totalistic groups (Hogg, 2004, 2005a). (It should be noted that not all the attributes mentioned above will necessary be found in all extreme groups.) There is a well‐documented association between societal uncertainty and various forms of extremism or totalism, such as genocide (Staub, 1989), cults (Curtis & Curtis, 1993; Galanter, 1989; Singer, 1995), ultranationalism (Billig, 1978, 1982; Kosterman & Feshbach, 1989), blind patriotism (Staub, 1997), religious fundamentalism (Altemeyer, 2003; Batson, Schoenrade, & Ventis, 1993; Rowatt & Franklin, 2004), terrorism (Moghaddam & Marsella, 2004), ideological thinking (Doty, Peterson, & Winter, 1991; Jost, Glaser, Kruglanski, & Sulloway, 2003; Jost & Hunyady, 2002; Lambert, Burroughs, & Nguyen, 1999), and fanaticism and being a ‘‘true believer’’ (HoVer, 1951). Uncertainty–identity theory describes a psychological mechanism that converts uncertainty into totalism. The process may be initiated by extreme and enduring uncertainty—for example, widespread societal uncertainty caused by economic collapse, cultural disintegration, civil war, terrorism, and large‐scale natural disasters, or more personal uncertainty caused by unemployment, bereavement, divorce, relocation, adolescence, and so forth. Under these circumstances, extreme groups may do a better job than merely high‐entitativity groups at reducing or fending oV uncertainty. People seek out extreme groups to identify with, or they make existing groups more extreme. Furthermore, people will identify strongly with such groups. They will have a strong sense of belonging and a strong feeling of attachment to the group, and their sense of self will be comprehensively defined by the group—they could be described as zealots, fanatics, or true believers. Describing the true believer, HoVer writes: ‘‘To live without an ardent dedication is to be adrift and abandoned. He sees in tolerance a sign of weakness, frivolity, and ignorance . . .. He hungers for the deep assurance which comes with total surrender—with the wholehearted clinging to a
UNCERTAINTY–IDENTITY THEORY
93
creed and cause. What matters is not the contents of the cause but the total dedication and communion with a congregation’’ (HoVer, 1951, p. 90).
Totalistic groups have properties that are ideally suited to uncertainty reduction. Although these properties may seem unattractive to most of us because they rigidly constrain our freedom, they are precisely what is sought by people who are very self‐uncertain. Totalistic groups have rigid boundaries that unequivocally define who is in and who is out—there is no ambiguity or fuzziness about membership, and thus membership can be diYcult to achieve and it can also be diYcult to ‘‘leave.’’ Group membership may be actively policed by the group—once you are in, you really feel you belong. The group’s identity is clearly, unambiguously, and relatively simply defined, and often sharply polarized away from other groups (Isenberg, 1986; Moscovici & Zavalloni, 1969)—as a member you know exactly who you are and how you should behave and how others will behave. There is a strong expectation of homogeneity and consensus—the group is suspicious of deviants and marginal members and will typically vilify and persecute such individuals (Hogg, 2005c; Hogg, Fielding, & Darley, 2005; Marques, Abrams, Pa´ez, & Hogg, 2001; Marques, Abrams, & Serodio, 2001). A conformist silo mentality prevails—this preserves homogeneity and inhibits change by suppressing criticism (Hornsey, 2005), and may produce symptoms similar to groupthink (Janis, 1972). There is reliable and widespread social validation of one’s worldview. Orthodoxy prevails—there is a single standard of right and wrong attitudes and behaviors (Deconchy, 1984). It is quite likely that value similarity comes to the fore as a defining feature of the group. Values run deeper than attitudes (Rokeach, 1973) and provide a stronger and more enduring behavioral imperative couched in the absolutist language of moral superiority. In totalistic groups, values, attitudes, and beliefs are tightly woven together into ideological belief systems that are self‐contained and explanatory (Billig, 1982; Larrain, 1979; Thompson, 1990), providing a firm and unassailable platform of certitude. Together orthodoxy and ideology may help resolve the postmodern paradox. Dunn (1998) has argued that there is a postmodern paradox that makes the certainties and absolutes oVered by ideologies particularly attractive in a postmodern world of moral and behavioral relativities. Baumeister (1987) paints a picture of medieval society in which social relations were fixed and stable, and legitimized in religious terms. People’s lives and identity were clearly mapped out according to position in the social order; by visible ascribed attributes such as family membership, social rank, birth order, and place of birth. People were tightly locked into a matrix of prescribed group memberships and social relations. By the 1950s, these stable identities
94
MICHAEL A. HOGG
had been almost entirely replaced by a more atomistic individual‐oriented status society that ‘‘did not promote feeling of identification or collective involvement’’ (Nisbet, 1959, p. 17), producing the postmodern paradox in which people with today’s less structured self yearn for community and the collective aYliations of times past (Barber, 1995; Bashevkin, 1991; Dunn, 1998; Gergen, 1991). This general idea is invoked by Lewis (2004, p. 19) in his analysis of contemporary Islamic fundamentalism—he argues that ‘‘in a time of intensifying strains, of faltering ideologies, jaded loyalties, and crumbling institutions, an ideology expressed in Islamic terms is particularly appealing.’’ Totalistic groups tend to be relatively insular. Insularity provides a comfortably circumscribed world for members, but is also associated with marked ethnocentrism (Brewer & Campbell, 1976) and accentuated mistrust and fear of outsiders (Stephan & Stephan, 1985). Because totalistic groups are highly entitative, processes of essentialism (Haslam et al., 1998) are also likely to be enhanced, further rendering self and social context invariant and immutable. Finally, because high entitativity is associated with clear internal structure, totalistic groups are likely to be relatively rigidly structured in terms of role relations and communication networks (Bavelas, 1968), and leadership is likely to be relatively hierarchical and autocratic (cf. Lippitt & White, 1943).
A. STUDIES OF UNCERTAINTY, IDENTITY, AND EXTREMISM This account of the general structure and form of extremist groups makes it easy to see how they provide succor for the uncertain. They provide the sort of group identity that clearly defines self and protects self from uncertainty. The basic prediction is that under uncertainty people should identify strongly with such groups, or make the structure and form of existing groups that they belong to more extreme and totalistic. We conducted a triad of very similar studies to investigate this hypothesis (Hogg, Meehan, Parsons, Farquharson, & Svensson, 2006). The participants were Australian university students, who could be expected to be individualistic (Triandis, 1989), and to have an independent self‐ construal (Markus & Kitayama, 1991; Oyserman, Coon, & Kemmelmeier, 2002) and low ‘‘power distance’’ (Hofstede, 1980), inclining them generally to prefer more moderate groups (cf. Fiske, Kitayama, Markus, & Nisbett, 1998; Smith, Bond, & Kag˘itc¸ibasi, 2006). This, taken in conjunction with the fact that the level of self‐uncertainty in the studies could not be genuinely extreme, led us to make the context‐specific prediction that the participants’
95
UNCERTAINTY–IDENTITY THEORY
9 Group identification
8 7 6 5
High uncertainty Low uncertainty
4 3 2 1 Low (moderate)
High (extreme)
Extremism Fig. 6. Hogg et al. (2006, Study 1): Group identification as a function of group extremism for high‐ and low‐self‐uncertainty participants, ¼ .16, t ¼ 2.16, p ¼ .033. Note: Identification is a three‐item scale ( ¼ .89) that can take values between one and nine.
general tendency to identify more strongly with moderate than extreme groups would weaken or disappear with increasing self‐uncertainty. Study 1 was a field experiment (N ¼ 168) in which we measured participants’ self‐uncertainty on 20 diVerent dimensions ( ¼ .88) and then provided them with a carefully constructed manifesto, ostensibly provided by members of a student action group that, like them, opposed various government threats to Australian universities. The manifesto described either a relatively moderate group or a quite extreme group in terms of its general structure and form—operationalizing extremism by capturing the attributes described above. There was then a check on perceived extremism, after which participants indicated how much they identified with the group (three items, ¼ .89) and how favorably they evaluated the group (four items, ¼ .90). As anticipated, participants evaluated the moderate group more favorably than the extreme group, and also identified more strongly with it. However, as predicted, the tendency to identify more with the moderate than extreme group disappeared among more highly uncertain participants (Fig. 6). In Study 2 (N ¼ 84), we measured self‐uncertainty with a subset of 15 items from Study 1 ( ¼ .91). Again we manipulated extremism by presenting information on a student action group—this time focusing on a diVerent student issue, and presenting the information via a carefully constructed audio interview with key members of the group. We checked on the manipulation, and measured identification (six items, ¼ .97), but then also measured participants’ intentions to engage in protest activities on behalf of the group (six items, ¼ .93). Identification with an extreme group is
96
MICHAEL A. HOGG
9 Group identification
8 7 6 5
High uncertainty Low uncertainty
4 3 2 1 Lo w (moderate)
High (extreme)
Extremism Fig. 7. Hogg et al. (2006, Study 2): Group identification as a function of group extremism for high‐ and low‐self‐uncertainty participants, ¼ .24, t ¼ 2.70, p ¼ .009. Note: Identification is a six‐item scale ( ¼ .97) that can take values between one and nine.
more problematic if people also intend to engage in extreme behavior—even more so if people actually engage in those behaviors. Attitude–behavior research concludes that behavioral intentions are a good predictor of actual behavior (Ajzen & Fishbein, 1977; Kraus, 1995), and social identity research theorizes and shows that the strongest attitude–behavior correspondence occurs if attitudes are normative of an in‐group that people identify with (Terry & Hogg, 1996; Terry, Hogg, & White, 2000). In Study 2, we predicted that the eVects of uncertainty and extremism on behavioral intentions would reflect the eVects on identification and be mediated by identification. As predicted and replicating Study 1, the tendency to identify more with a moderate than an extreme group disappeared under high uncertainty (Fig. 7) and this was exactly reflected in behavioral intentions—identification mediated the eVect of uncertainty and extremism on behavioral intentions. Study 3 was a laboratory experiment (N ¼ 82) where we first manipulated extremism via a description of a student action group, focusing on a third diVerent student issue—but this time the description was in the form of a carefully staged (using actors) video interview with key members of the action group. We then manipulated, rather than measured self‐uncertainty— using the priming procedure from Hogg et al. (2007) described above. The dependent measures were a four‐item group evaluation scale adapted from Study 1 ( ¼ .83), a nine‐item identification scale ( ¼ .94), and the six‐item behavioral intentions scale from Study 2 ( ¼ .92). Unlike Study 1, participants did not evaluate the moderate group more favorably than the extreme group, but instead evaluated both groups more favorably when they were uncertain. On identification we replicated both previous studies—the
97
UNCERTAINTY–IDENTITY THEORY
9 Group identification
8 7 6
5.04b
5.42b
5.34b
5 3.69a
4
High uncertainty Low uncertainty
3 2 1 Low (moderate)
High (extreme)
Extremism Fig. 8. Hogg et al. (2006, Study 3): EVect of uncertainty and group extremism on group identification, F(1, 78) ¼ 10.70, p ¼ .002. Note: Identification is a nine‐item scale ( ¼ .94) that can take values between one and nine. Means not sharing a subscript are significantly diVerent by simple main eVects.
tendency to identify more with a moderate than an extreme group disappeared under high uncertainty, but here we also found that where the group was extreme participants identified significantly more if they were uncertain than if they were not (Fig. 8). As in Study 2, this eVect was mirrored in behavioral intentions and mediated by identification. Together, these three studies provide consistent support, across varying methods of manipulating extremism and both measured and manipulated self‐uncertainty, for uncertainty–identity theory’s predictions about uncertainty‐induced identification with extreme groups. Student participants’ preference to identify with moderate groups disappeared when they were more uncertain, and identification with extreme groups was stronger among participants who were high than low in uncertainty. This eVect was exactly mirrored in intentions to engage in behavior on behalf of the group. It remains to be shown empirically that when uncertainty is genuinely extreme (in the present studies uncertainty was high, but not extreme) participants identify significantly more with extreme than moderate groups.
VI. Extensions, Applications, and Implications of Uncertainty–Identity Theory In this section, I describe some extensions and implications of uncertainty– identity theory—to some extent consolidating ideas mentioned earlier.
98
MICHAEL A. HOGG
A. DEPERSONALIZATION AND PROJECTION The reason why group identification reduces uncertainty is that self‐ categorization transforms self so that perceptions, feelings, attitudes, and behaviors are prescribed by the group prototype—a process of depersonalization in which the group’s prototypical attributes are assigned to self. This process assumes that there is some content to the prototype—that there really are some attributes to assign to self. Indeed, the more richly, concretely, clearly, and distinctively defined a group is the better it is at reducing self‐uncertainty. As we have seen above, under uncertainty people prefer to identify, and identify more strongly with high‐entitativity groups that have these sorts of prototypes. What happens, then, if the group prototype is fuzzy and unclear, lacks consensus, or is impoverished and information‐poor? The simple answer, from above, is that under uncertainty we would be disinclined to identify because the group would do little to reduce uncertainty. Instead, we would identify with a diVerent more entitative group that had more clearly defined attributes. However, sometimes this is not possible; for example, when you join a new group and know little about it, when there are few if any other groups that are self‐relevant in that situation, or when you are externally constrained to define yourself in terms of the group. Under these circumstances, people use whatever information they can to construct a clear and substantial prototype for the group. For example, information can be gathered slowly from other people, the media, the Internet, and so on. However, one source of information that is immediately available, extraordinarily rich, and virtually instantaneous to apply is the self. People know a great deal about themselves, and knowing they share a category membership with others triggers an automatic assumption of shared attributes. If you do not know much about the other people or the group as a whole, you assume that the group and its members have attributes like yours—you project your attributes onto the group and, eVectively, make the group in your own image. This idea has been proposed by Crisp and colleagues (Crisp, Farr, Beck, Hogg, & Cortez, 2006; Crisp & Hogg, 2006). Drawing on self‐anchoring theory (Cadinu & Rothbart, 1996), the concept of social projection (Clement & Krueger, 2000; Krueger & Clement, 1996), and research by Otten (2002; Otten & Moskowitz, 2000; Otten & Wentura, 1999, 2001), Crisp and colleagues argue that a full analysis of the process of uncertainty reduction through group identification needs to consider not only uncertainty about or related to self directly, but also uncertainty about the attributes of particular social identities and groups that one could identify with. Feelings of self‐uncertainty typically hinge on who we are and how we should behave and relate to others in a given context, and we assign group
UNCERTAINTY–IDENTITY THEORY
99
attributes to ourselves through a process of self‐categorization‐generated depersonalization. Here, the main focus of our uncertainty is self not the group—we are relatively certain about the group’s attributes. If, however, self‐uncertainty is associated with lack of information about the group’s attributes, the main focus of uncertainty may actually be the group—we are more uncertain about the group than about self. Self‐categorization is here associated with a process of projection in which we assign our own attributes to the group. Projection reduces uncertainty because it is associated with an identity and (cognitively constructed) self‐definition that is ostensibly grounded in consensus. However, there is no doubt that projection is a less immediate route to uncertainty reduction. It is possible that it is a first step toward constructing a concrete prototype that we assign to self through depersonalization. Crisp et al. (2006) conducted two similar experiments (N ¼ 57 and 159) as a first step in investigating depersonalization and projection processes in uncertainty‐induced identification. In both studies, the participants were either undergraduate students who had just arrived at university (first year, freshmen) or students who had already been there for a year (second year, sophomores). It was reasoned that freshmen would be more uncertain than sophomores about the group (the university). The second predictor variable was a manipulation in which participants were primed to either focus on self or focus on the group. The dependent variable was a measure of self‐in‐group definitional overlap—overlap between participants’ description of themselves and their description of the group. As predicted, among sophomores (relatively certain about the in‐group) definitional overlap was greater in the group prime condition (a process of depersonalization), whereas among freshmen (uncertain about the in‐group) definitional overlap was greater in the self prime condition (a process of projection). The idea that the locus of uncertainty may aVect the process of group identification has many implications. For example, it may help explain the way in which group leaders initiate actions to define the group whereas followers do not. Perhaps leaders are more certain than followers about themselves as group members and therefore, in association with identification, they engage in more pronounced projection behaviors that seek to mold the group in their own image.
B. CENTRAL MEMBERS, MARGINAL MEMBERS, LEADERS, AND DEVIANTS This idea can be pursued further by drawing on the social identity theory of leadership (Hogg, 2001b; Hogg & van Knippenberg, 2003; van Knippenberg & Hogg, 2003; also see Ellemers, de Gilder, & Haslam, 2004) and on social
100
MICHAEL A. HOGG
identity analyses of deviance (Marques et al., 2001; Marques, Abrams et al., 2001) and of the prototypicality of group members (Hogg, 2005c). A key point in these analyses is that group members vary in terms of how prototypical of the group they are, how prototypical they consider themselves to be, and how prototypical others consider them to be. In groups, people pay very close attention to the prototype and to prototype‐relevant information, and therefore prototypicality impacts social processes within and between groups very significantly. Some implications of this are that prototypically central members find it relatively easy to be innovative and to show leadership, whereas prototypically marginal members do not, and are often perceived negatively and treated as deviants. Uncertainty may be involved in a variety of diVerent ways. Most obviously, marginal members will be more uncertain than central members of their membership status in the group and of the prototypical properties of the group. Research shows that prototypical members are considered to be more strongly identified with the group than marginal members, and are therefore trusted more than marginal members—their membership credentials are less often called into question (see Platow, Reid, & Andrew, 1998; Platow & van Knippenberg, 2001; Tyler, 1997; Tyler & Lind, 1992). Using the language of Higgins’s (1998) regulatory focus theory, marginal members may be more oriented toward promoting their identification in order to reduce their membership uncertainty, whereas prototypical members may be more oriented toward protecting their identification in order to stave oV the specter of membership uncertainty. Thus, prototypicality and centrality of membership may influence people’s orientation toward uncertainty reduction through identification, and the particular social behaviors they engage in to manage self‐uncertainty. Taken in conjunction with the earlier discussion of depersonalization and projection, we could speculate that central members (leaders?) tend to engage in behaviors that involve identity protection and self‐projection onto the group, whereas more marginal members tend to engage in behaviors involving identity acquisition and depersonalization. Marginal members may also have a strong need to feel that they can trust more prototypical members or leaders to treat them honestly, in matters of identity and membership and in matters relating to the nature of the group prototype (see below). A further twist to this story is that certainty is power. If you can call into question someone’s sense of self, their identity, then you have exercised enormous power over that person. Within groups, leaders can use their position of influence and power to keep subordinates teetering on the brink of uncertainty about their membership and wider sense of self. They can do this by continually redefining the group and the meaning of membership and
UNCERTAINTY–IDENTITY THEORY
101
by targeting marginal members as deviants to be pilloried, persecuted, and rejected. At the same time, the leader can oVer complete commitment and subjugation to and identification with the group as the ultimate resolution of uncertainty (see Hogg, 2007). Marris (1996) has elaborated this idea in the context of intergroup relations—talking about the ‘‘politics of uncertainty.’’ He argues that intergroup relations are a matter of groups struggling over certainty. Dominant groups are those whose existence is characterized by relative certainty (e.g., secure jobs, reliable income, media validation of one’s worldview) and subordinate groups by uncertainty (e.g., job insecurity, unpredictable income, unreliable housing). Groups actively strive to reduce uncertainty for their own group and increase uncertainty for relevant out‐groups—an intergroup struggle for certainty. Marginal members of a group or marginal groups in wider society may not always be powerless in the uncertainty game. They too can use uncertainty as a strategic tool to influence and control others. A key idea in the literature on minority influence (Martin & Hewstone, 2003; Moscovici, 1980; Mugny & Pe´rez, 1991) is that active minorities present a viewpoint that is consistent across time and widely shared by members of the minority. A viewpoint with these qualities causes the majority to feel uncertain about its own viewpoint— uncertainty that runs deep and ultimately leads to conversion to the minority’s viewpoint. For marginal status to have this eVect there probably needs to be a group of marginal members—a lone marginal would find it diYcult to create the perception of a shared viewpoint. From an uncertainty–identity theory perspective, conversion has its eVect not merely because the minority’s behavior style raises uncertainty in the majority, but because the uncertainty is ultimately about or focused on self and identity. Conversion involves the majority identifying with a redefined representation of majority group membership. The final way in which uncertainty may relate to leadership was foreshadowed in the earlier discussion of totalistic groups. It was argued that all things being equal, extreme uncertainty leads to strong identification with totalistic groups that have relatively hierarchical leadership structures. Hierarchical leadership provides unambiguous group structure and a clear definition of consensual prototypicality. This is probably true most of the time, but there may be exceptions. Some extremist groups may be ideologically opposed to hierarchical leadership. For example, some relatively extreme 1970s feminist groups actively eschewed hierarchical leadership (Freeman, 1973; Hackman, 2002, p.93). For other extremist groups, hierarchical leadership may put the group at risk due to elevated visibility (e.g., some 1950s communist cells), or may be diYcult to maintain due to other factors. For example, some contemporary terrorist groups are highly distributed as a network of insulated autonomous cells and no one strong leader.
102
MICHAEL A. HOGG
In this way, the communication structure of a group (Bavelas, 1968) may influence the extent to which strong hierarchical leadership is feasible, even in extremist groups. Highly centralized groups with a single communication node are conducive to the development of hierarchical leadership under uncertainty. Less centralized groups which do not have a single communication hub are less conducive to hierarchical leadership. In the latter case, conditions of uncertainty may encourage the group and its members to change the group’s communication structure. Sometimes this may be diYcult because the communication structure is tied to the very nature of the group—for example, many groups that rely on computer‐mediated communication (Hackman, 2002). Nevertheless, in the case of extreme groups with no apparent single strong leader, it may simply be that group life prevents the embodiment of leadership in a single person, but ideological or symbolic leadership prevails in the form of a strong prescriptive representation of group identity. C. UNCERTAINTY, IDENTITY, AND TRUST Trust is closely associated with predictability and uncertainty. The more we are able to trust someone, the more predictable they are and the less uncertain we feel. Trust plays a central role in group, specifically in‐group, life. In‐groups are ‘‘bounded communities of mutual trust and obligation’’ (Brewer, 1999, p. 433) in which members expect to be able to trust fellow members to do them no harm and to be acting in the best interest of the group as a whole (Levine & Moreland, 2002). This expectation serves a clear evolutionary function in allowing people to have confidence in those with whom they choose to cooperate (Kenrick, Li, & Butner, 2003). In‐group members who betray our group‐based trust by leaving the group to pursue their personal interest or by acting in ways that only benefit themselves and are to the detriment of the group as a whole reduce trust and raise uncertainty and thus invite harsh reactions. This dynamic can be particularly pronounced for central members who are prototypical of the group or act as group leaders. Disloyalty and violation of trust on the part of prototypical members is most disruptive of group equilibrium (Arrow, McGrath, & Berdahl, 2000; Van Vugt & Hart, 2004), and is a particularly potent source of uncertainty about what the group stands for, about one’s membership in the group, and ultimately about self. We do not have such expectations for marginal members. On the contrary, we expect them to be less reliable and less loyal, and are less surprised when they behave in these ways. In addition, we can react more harshly toward them, precisely because of their marginal position in the group. Thus, they can readily be marginalized as deviants.
UNCERTAINTY–IDENTITY THEORY
103
D. UNCERTAINTY, IDENTITY, AND IDEOLOGY The relation between uncertainty, identity, and ideology is discussed fully elsewhere (Hogg, 2005a) and is mentioned a number of times above. Ideology is a highly contested sociopolitical construct (Larrain, 1979; Thompson, 1990), but for our social psychological purposes an ideology can be defined as ‘‘an integrated system of apparently congruent beliefs and values that explain and justify the world, our place within it, our relationship with others, and our own and others’ actions. Furthermore, it is a belief system that is shared within a group or community’’ (Hogg, 2005a, p. 204). To this can be added the fact that ideologies are largely self‐contained universes. They define the problematic and provide answers to all and only those questions generated by the problematic. From this, it is quite clear that ideological belief systems are perfectly suited to uncertainty reduction. They do exactly what is needed to reduce self‐uncertainty. Ideologies arise under uncertainty and prevail to ward oV uncertainty. Ideologies are also almost universally attached to specific group membership—group memberships that people zealously cherish with extraordinary passion. It is no accident that monolithic ideological systems have arisen around the most fundamental human uncertainties—the origins of life and the universe, the organization of human society, and death and the afterlife. From uncertainty–identity theory, it can be seen why these scientific, religious, and political ideologies, particularly the latter two, are a perennial source of much of humanity’s greatest inhumanity. Social psychologists have begun to focus more closely on ideology and ideology‐related constructs (Crocker & Major, 1994; Duckitt, 2001; Greenberg, Koole, & Pyszczynski, 2004; Jost & Major, 2001; Jost et al., 2003; Major, Quinton, & McCoy, 2002). Particular ideological forms have been a specific focus: for example, social dominance orientation (Sidanius & Pratto, 1999), right wing authoritarianism (Altemeyer, 1998), system justification beliefs (Jost, Banaji, & Nosek, 2004; Jost & Hunyady, 2002), the protestant work ethic (Furnham, 1990; Quinn & Crocker, 1999), belief in a just world (Furnham, 2003; Furnham & Procter, 1989; Rohan, 2000), and religious fundamentalism (Altemeyer, 2003; Rowatt & Franklin, 2004). Uncertainty–identity theory does not distinguish among specific ideologies, but focuses on common features of ideological systems that suit them to uncertainty reduction through group identification. The greater the uncertainty, the more likely it is that people will identify very strongly, as true believers, with high entitativity, perhaps totalistic, groups that have a tightly integrated, consensual, and explanatory system of absolutist beliefs, attitudes, and values.
104
MICHAEL A. HOGG
E. UNCERTAINTY, IDENTITY, AND SOCIAL MOBILIZATION Uncertainty about outcomes and about the behavior of others is often viewed as a significant obstacle to social mobilization. The topic of social mobilization focuses on people’s participation in social protest or social action groups (Reicher, 2001; Stu¨rmer & Simon, 2004; Tyler & Smith, 1998), and addresses the question of how individual discontents are transformed into collective action. How and why do sympathizers become mobilized as activists or participants? Klandermans (1997) argues that mobilization reflects the attitude–behavior relation. Sympathizers hold sympathetic attitudes toward an issue, yet these attitudes do not readily translate into behavior. From this perspective, one barrier to mobilization is that people are not sure how, or if at all, their behavior will achieve their attitudinal goals (cf. Ajzen, 1989). Mobilization also resembles a social dilemma (cf. de Cremer & van Vugt, 1999). Protest is generally for a social good (e.g., equality) or against a social ill (e.g., oppression). Success benefits everyone irrespective of participation but failure harms participants more. Because you cannot be sure that others will participate, it is tempting to free ride, to remain a sympathizer rather than become a participant. In both cases, uncertainty is a potent barrier to participation. However, uncertainty–identity theory suggests that uncertainty may sometimes have the opposite eVect. To the extent that uncertainty reflects on or is focused on self‐conception, people identify strongly with the group, and the group’s norms are applied to self via self‐categorization and depersonalization. The group’s attitudes become one’s own, or one’s own attitudes become prototypical of membership. According to the social identity analysis of the attitude–behavior relation, it is precisely under these circumstances that people are most likely to behave in accordance with their attitudes (Terry & Hogg, 1996; Terry et al., 2000; also see Hogg & Smith, in press)—thus, it is group identification that increases the probability of social action and collective protest (Stu¨rmer & Simon, 2004). There is some evidence for this analysis. Two of the studies of uncertainty and extremism reported above (Hogg et al., 2006, Studies 2 and 3) found that intentions to actually engage in behavior on behalf of the group were fully mediated by identification. The extent to which people reported they would engage in attitude‐congruent group normative behavior was entirely determined by the degree to which they identified with the group. Smith et al. (in press) report two studies adopting a conventional attitude– behavior paradigm to focus on the impact of self‐uncertainty on conformity to group norms. In both studies, self‐uncertainty was manipulated using a deliberative mindset manipulation (McGregor et al., 2001). In Study 1 (N ¼ 106), student participants were exposed to information about a relevant student
105
UNCERTAINTY–IDENTITY THEORY
Willingness to engage in behavior
9 8 7 6 5
5.28b
4.77b 4.89ab 3.78a
4
Incongruent norm Congruent norm
3 2 1 Low
High Uncertainty
Fig. 9. Smith et al. (in press, Study 1): EVect of uncertainty and normative support on willingness to engage in attitude‐consistent behavior, F(1, 97) ¼ 5.72, p ¼ .019. Note: Willingness to engage in attitude‐consistent behavior is a four‐item scale ( ¼ .85) that can take values between one and nine. Means not sharing a subscript are significantly diVerent by simple main eVects.
issue that portrayed normative student attitudes as being congruent, or incongruent, with the participant’s own attitude. In Study 2 (N ¼ 83), there was a third norm condition in which the normative information was ambiguous. The dependent variable comprised a range of measures of intentions to engage in attitude‐consistent behaviors. The results of Study 1 were clearest. Participants’ intentions to engage in attitude‐consistent behavior (four items, ¼ .85) strengthened under uncertainty where the group provided normative support for their attitude, but weakened where the group’s norm was contrary to the participants’ attitude (Fig. 9).
VII. Uncertainty–Identity Theory in Relation to Other Ideas The previous section discussed some extensions and implications of uncertainty–identity theory. In this section, I briefly touch on, sometimes revisit, the relation between uncertainty–identity theory and some related ideas. A. CONTEXT AND PERSONALITY An important credo for uncertainty–identity theory is that self‐uncertainty is produced by contextual factors. Thus, enduring uncertainty reflects an enduring uncertainty–producing context rather than an individual diVerence or personality factor—an argument famously used by others to attribute
106
MICHAEL A. HOGG
authoritarianism and prejudice to context rather than personality (Minard, 1952; Pettigrew, 1958). For example, variations in uncertainty across the life span can reflect developmental or life‐span factors such as adolescence, empty‐nest, retirement, and so forth. This is not to say that people do not diVer in terms of their experience of uncertainty, their tolerance for uncertainty, their need to reduce uncertainty, and their mode of reducing or avoiding uncertainty. Rather, much of this individual variation may be due to variation in enduring contextual factors; and in uncertainty–identity theory, individual variation is treated as ‘‘error variance.’’ Nevertheless, as discussed earlier, the social psychology literature on uncertainty and uncertainty‐related constructs is dominated by personality and individual diVerences approaches. One particularly relevant personality approach to uncertainty is the work of Kruglanski and colleagues on need for cognitive closure (Kruglanski, 1989, 2004; Kruglanski & Webster, 1996; Neuberg & Newson, 1993; Webster & Kruglanski, 1998). Kruglanski talks about need for cognitive closure as a generic individual diVerence variable. Some people have a strong need for structure and cognitive closure. They are more concerned to reduce uncertainty quickly than to be correct, and therefore ‘‘seize’’ on information and ‘‘freeze’’ on their judgments—they are closed‐ minded. Other people have a greater fear of invalidity and are able to tolerate substantial uncertainty while they engage in a prolonged search for validity. There is no doubt that need for cognitive closure is essentially a personality and individual diVerences construct. A 42‐item need for cognitive closure scale has been developed (Webster & Kruglanski, 1994), and research shows that individual diVerences in need for closure are associated with a wide array of other personality constructs. However, Kruglanski has reintroduced the notion that situational factors can have an influence (Kruglanski, Pierro, Mannetti, & De Grada, 2006). For example, need for closure has been elevated by time pressure, ambient noise, fatigue, alcoholic intoxication, and other manipulations that vary the costs or benefits of reaching closure quickly and completely. Kruglanski et al. (2006) also introduce a new concept—a ‘‘syndrome’’ they call ‘‘group‐centrism,’’ which emerges ‘‘When people care a lot about sharing opinions with others in their group; when they endorse central authority that sets uniform norms and standards; when they suppress dissent, shun diversity, and show in‐group favoritism; when they venerate their group’s norms and traditions, and display fierce adherence to its views; when above all, they exhibit all these as a package’’ (p. 84).
They go on to argue that need for closure as a trait or a state can foster group‐centrism. For instance, high need for closure has been found to be
UNCERTAINTY–IDENTITY THEORY
107
associated with a number of more pronounced group membership‐related behaviors such as stereotyping (Dijksterhuis, van Knippenberg, Kruglanski, & Schaper, 1996) and intergroup discrimination (Shah, Kruglanski, & Thompson, 1998). Although Kruglanski and associates only make brief passing reference to social identity theory and to the role of uncertainty reduction in group identification, their analysis has clear convergence with aspects of uncertainty–identity theory, and therefore provides some conceptual validation for uncertainty– identity theory. Specifically, the idea that a drive for closure can cause people to become group‐centric to the extent that they prefer highly groupy groups, resonates with the uncertainty–identity thesis that self‐uncertainty ‘‘drives’’ people to categorize themselves as members of (identify with) entitative groups, and that more extreme uncertainty can produce true believers and zealotry associated with totalistic group structures characterized by ideology, intolerance, extreme ethnocentrism, homogeneity, suppression of dissent, hierarchical structure, and so forth. Another particularly relevant personality approach to uncertainty is Sorrentino’s research on uncertainty orientation (Sorrentino & Roney, 1986, 1999; Sorrentino & Short, 1986). Sorrentino diVerentiates between (1) uncertainty‐oriented people, who seek out information that may raise uncertainty, in order to resolve uncertainty (‘‘need to know,’’ or scientific types); and (2) certainty‐oriented people who avoid uncertainty if they can, and when confronted with uncertainty defer to others or use heuristics to resolve uncertainty quickly. Sorrentino and colleagues believe that the social identity analysis of uncertainty reduction and group identification applies better to certainty‐ than uncertainty‐oriented people, and report a modified replication of Mullin and Hogg (1998) to support their argument (Sorrentino et al., 2001). Two comments can be made in relation to this work (see Hogg, 2001a). First Sorrentino separates social identity processes and self‐categorization processes, when in fact self‐categorization is the cognitive process responsible for social identification. Second, as explained above, we would expect individual diVerences to strengthen or weaken the uncertainty–identity relation, but this does not invalidate the uncertainty–identity process. B. CULTURE Sorrentino’s work does, however, invite the question of how culture might impact the uncertainty–identity relation. Of particular relevance is the fact that one of Hofstede’s (1980) four dimensions to diVerentiate cultures was uncertainty avoidance. Western cultures (e.g., Denmark) are low on
108
MICHAEL A. HOGG
uncertainty avoidance, whereas Eastern cultures (e.g., Japan) are high. Relatedly, Schwartz (1992) reported that Eastern cultures are less open to change. Overall, relative to Western cultures, Eastern cultures are not only more uncertainty avoidant and less open to change, but they are more collectivist (Hofstede, 1980) and they are associated with a more interdependent (Markus & Kitayama, 1991) and relational (Brewer & Gardner, 1996; Yuki, 2003) self‐concept (see Oyserman et al., 2002). There appears to be an association between a stronger desire to reduce or avoid uncertainty, and self‐ conception in more collectivist, interdependent, or relational terms. This is consistent with uncertainty–identity theory. Where the need to reduce uncertainty is elevated, people identify more strongly with groups, particularly high‐entitativity groups. Hogg (2006a) reports a preliminary study, by Hogg and Alit, testing the hypothesis that uncertainty‐induced identification with high‐entitativity groups would be accentuated among people with a stronger inclination toward collectivist/interdependent/relational self‐construal. Participants were 115 Indonesian tertiary students from the island of Bali and 105 Australian tertiary students from the city of Brisbane (N ¼ 220). They completed a pencil‐and‐paper role playing exercise in which independent– interdependent self‐construal was measured (median split produced ‘‘independents’’ and ‘‘interdependents’’), performed a diYcult or easy perception task (a standard uncertainty manipulation), and then role‐played being in a high‐ or low‐entitativity discussion group (entitativity manipulation). The dependent measure was an eight‐item group identification scale ( ¼ .96). The three‐way interaction was significant, F(1, 212) ¼ 4.97, p ¼ .027. Under high uncertainty, participants identified more strongly with a high‐ than a low‐entitativity group, and the eVect was stronger for interdependent than independent participants.
C. TERROR MANAGEMENT, COMPENSATORY CONVICTION, SELF‐VERIFICATION, AND SYSTEM JUSTIFICATION Over the past decade, terror management theory has attracted substantial attention in social psychology (Greenberg, Solomon, & Pyszczynski, 1997; Pyszczynski et al., 1999, 2004). The key point is that existential anxiety, fear of death, motivates aYliation and other behaviors aimed at buVering this anxiety. One way in which people can buVer existential anxiety and create symbolic immortality is by constructing, adhering to, and protecting a cultural worldview (cf. ideology) that provides them with a sense of order, stability, and predictability. Mortality salience has been shown to increase
UNCERTAINTY–IDENTITY THEORY
109
aYliation and belongingness needs (Mikulincer, Florian, & Hirschberger, 2004; Pyszczynski et al., 1996), and group and worldview protective behaviors (Greenberg et al., 1990; Harmon‐Jones, Greenberg, & Solomon, 1996). Existential anxiety is, however, a messy variable. It certainly involves anxiety about death, but also a significant degree of uncertainty about one’s own death and, perhaps most importantly, about what there is after death, the afterlife. Not surprisingly, mortality salience has been shown to increase people’s desire for certainty (Landau et al., 2004; van den Bos, 2001), and to have very similar eVects to those produced by uncertainty manipulations (van den Bos & Miedema, 2000; van den Bos, Poortvliet, Maas, Miedema, & van den Ham, 2005; see van den Bos & Lind, 2002). In addition, low state or trait uncertainty has been shown to reduce the impact of mortality salience (Dechesne, Janssen, & van Knippenberg, 2000; Jost, Fitzsimons, & Kay, 2004). This suggests that existential anxiety may have an epistemic dimension, that is, in this case focused on death and the afterlife, but obeys more general principles of uncertainty reduction through group identification as specified by uncertainty–identity theory. The afterlife certainly is the big unknown, and so it is not surprising, from an uncertainty–identity point of view, that religious identities and associated ideologies have a long history of tending toward fundamentalism, extremism, and zealotry. Uncertainty–identity theory has implications for McGregor’s compensatory conviction model. On the basis of self‐aYrmation theory (Steele, 1988), compensatory conviction is a hydraulic motivational model. It argues that when people feel uncertain in one domain, they compensate for this by ‘‘spontaneously emphasizing certainty and conviction about unrelated attitudes, values, personal goals, and identifications’’ (McGregor et al., 2001, p. 473; also see McGregor & Marigold, 2003). Compensatory conviction research focuses mainly on uncertainty‐sponsored hardening of expressed attitudes toward social issues, and has (McGregor & Marigold, 2003) investigated the moderating role of self‐esteem as an individual diVerence variable. Although it provides some data showing uncertainty‐sponsored hardening of intergroup attitudes, accentuation of consensual worldviews, and aYrmation that particular activities define self (McGregor et al., 2001), compensatory conviction is not primarily a motivational theory of psychological identification with a social group. However, a group of studies by McGregor et al. (2005) shows more directly that group identification and exaggerated opinion consensus can be consequences of uncertainty‐related threats. Because compensatory conviction states that people compensate for uncertainty in one domain by expressing conviction in another, it cannot so easily
110
MICHAEL A. HOGG
account for the uncertainty–identity theory findings that the uncertainty– identification relationship is moderated by being categorized (Grieve & Hogg, 1999) and by the entitativity of the group (Hogg et al., 2007). Uncertainty– identity theory focuses on group identification, not aYrmation or conviction, as a response to self‐uncertainty, and thus it predicts that factors related to group membership, such as inclusion, categorization, and entitativity, will influence how strongly people identify with a group as a response to self‐uncertainty. Another theory that relates to aspects of uncertainty–identity theory is Swann’s self‐verification theory (Swann, 1977; Swann, Polzer, Seyle, & Ko, 2004; Swann, Rentfrow, & Guinn, 2003). Self‐verification theory ‘‘. . . assumes that stable self‐views provide people with a crucial source of coherence, and invaluable means of defining their existence, organizing experience, predicting future events, and guiding social interaction’’ (Swann et al., 2003, p. 369).
It goes on to argue that interpersonal relations provide ‘‘the steady supply of self‐verifying feedback from others’’ (Swann et al., 2003, p. 369) that is required to construct a stable sense of self. Having achieved a stable and validated sense of self, people are invested in maintaining and protecting this self‐view and are likely to pursue a range of strategies to confirm and verify their self‐concept. Using a slightly diVerent language, self‐verification theory shares with uncertainty–identity theory the assumption that people need to reduce uncertainty about themselves and the world they live in. The key diVerence lies in the source and process of self‐uncertainty reduction. For self‐verification, the source is feedback from other people we interact with and the process is information comparison. For uncertainty–identity theory, the source is group prototypes (about which we can gather information from other people) and the process is self‐categorization and prototype‐based depersonalization. Another key diVerence concerns the types of groups to which people seek out to belong. For self‐verification theory, it is ‘‘groups that confirm (people’s) negative or positive self‐views’’ (Swann et al., 2004, p. 11); for uncertainty–identity theory, it is groups that provide a clearly defined and distinctive identity, high‐entitativity groups. Finally, system justification theory (Jost, Banaji et al., 2004; Jost & Hunyady, 2002) touches on issues that are relevant to uncertainty–identity theory. System justification theory maintains that ‘‘people are motivated to perceive existing social and political arrangements as fair, legitimate, and justifiable . . ., even sometimes at the expense of personal and group interests and esteem,’’ and goes on to argue that ‘‘people who are most disadvantaged by a given social system should paradoxically be the most likely to provide
UNCERTAINTY–IDENTITY THEORY
111
ideological support for it, insofar as they have the greatest need to justify their suVering’’ (Jost, Fitzsimons, & Kay, 2004, p. 267, emphasis in original). Uncertainty reduction is nominated as one motivation for system justification, along with cognitive consistency, conservation of eVort and prior beliefs, fear of equality, illusion of control, belief in a just world, and reduction of dissonance over failing to combat oppression. However, the needs to reduce uncertainty and threat are considered to be perhaps the primary needs—the ones most closely associated with ideological justification of the status quo (Jost et al., 2003). Jost, Fitsimons, and Kay (2004) write that ‘‘the management of uncertainty and threat would be most closely linked to the two core components of conservative thought . . ., resistance to change and acceptance of inequality’’ and ‘‘(N)eeds to reduce uncertainty and threat are well served by ideological resistance to change, insofar as change (by its very nature) upsets existing realities and is fraught with epistemic insecurity’’ (p. 271, parenthesis in original). In this way, system justification theory attributes inequality‐enhancing conservative ideologies primarily to the management of uncertainty, which is consistent with uncertainty–identity theory. However, uncertainty–identity theory is more basic—it specifies precisely how uncertainty reduction is social cognitively accomplished by self‐categorization and group identification and the associated process of prototype‐based depersonalization. It is also more encompassing—it shows how any ideological belief system, not just a politically conservative one, grounded in a high‐entitativity group is ideally suited to self‐uncertainty reduction through group identification.
VIII. Concluding Comments This chapter describes uncertainty–identity theory—its origins, its concepts, its implications and extensions, and its relation to related ideas and theories. The emphasis has been conceptual but empirical support was also described and assessed. Uncertainty–identity theory is an extension of social identity theory that postulates uncertainty reduction as a key motivation for social identity processes and group and intergroup behaviors. It is a theory that attributes particular forms of group attachment, self‐definition, and group structure to people’s striving to reduce, via group identification, self‐ categorization and prototype‐based depersonalization, feelings of uncertainty about and related to themselves. The key features of uncertainty–identity theory can be captured by four broad premises. Premise 1. People are motivated to reduce or avoid feelings of uncertainty about themselves, and about their perceptions, judgments, attitudes, and
112
MICHAEL A. HOGG
behaviors that relate to themselves, their interactions with other people, and their place in social context. Premise 2. Social categorization reduces or protects from uncertainty because it depersonalizes perception to conform to one’s in‐group and out‐ group prototypes, such that one ‘‘knows’’ how others will behave. Prototypes define and prescribe people’s identities and therefore their perceptions, attitudes, feelings, and behaviors, and how they interact with and treat other people, including oneself. Social categorization of self, self‐categorization, assigns one an identity with all its associated in‐group prototypical attributes. There is usually substantial agreement within a group on the in‐group prototype and on prototypes of relevant out‐groups—further reducing uncertainty through consensual validation for one’s behaviors and sense of self. Premise 3. Prototypes are better at resolving uncertainty to the extent that they are simple, clear, unambiguous, prescriptive, focused, and consensual, as well as coherently integrated, self‐contained, and explanatory. These kinds of prototypes circumscribe clear identities and define or are associated with distinctive, well‐structured groups that are high in entitativity. Under uncertainty, people identify more strongly with high‐entitativity groups— they seek them out to join, they create them anew, or they transform existing groups to be more entitative. Premise 4. Where uncertainty is extreme and enduring, the motivation to reduce uncertainty and the quest for high‐entitativity groups and clear prototypes are accentuated. Under these circumstances, people may identify passionately as true believers or zealots, seeking rigidly and hierarchically structured totalistic groups with closed boundaries, homogenous and ideological belief structures, and inflexible customs—ethnocentric, insular, and somewhat narcissistic groups that suppress dissent and are intolerant of outsiders. These kinds of groups provide all‐embracing identities that are powerful buVers against self‐uncertainty. Direct and indirect tests of uncertainty–identity theory provide good support for Premise 2, across diVerent paradigms and diVerent manipulations of uncertainty. The studies show that uncertainty’s motivational role is independent of self‐enhancement and self‐esteem concerns, and provide some evidence for reduced uncertainty after identification. There is also good support for Premise 3. Entitativity moderates the uncertainty–identification relationship as predicted, as do the self‐relevance of the group and the subjective importance of the dimension of uncertainty. Studies of ways in which uncertainty can lead group members to accentuate entitativity remain to be published. A series of unpublished preliminary studies provide some support for this by finding that under uncertainty people accentuate how diVerent they see in‐group and out‐group to be—they polarize their representations of in‐group and out‐group attitudes (Sherman, Hogg, Maitner, & MoYtt, 2006).
UNCERTAINTY–IDENTITY THEORY
113
Direct support for Premise 4 comes from a series of studies that were described. However, there is indirect support from a wide array of literatures that associate uncertainty with aspects of extremism or conservative ideology and praxis. Unlike these other literatures, uncertainty–identity theory specifies the social cognitive process and places self‐conception and social identity at its heart. Much research remains to be done—for example, experiments and surveys with really extreme groups and really extreme uncertainty, closer investigations of the role of ideology and values in fending oV or reducing extreme uncertainty, a focus on leadership in totalistic groups, and a focus on uncertainty and active minorities. Uncertainty–identity theory embraces some concepts that are also dealt with by other theories and constructs—for example, terror management theory, self‐verification theory, compensatory conviction, system justification theory, need for cognitive closure, group‐centrism, and uncertainty orientation. This furnishes convergent validity for the theory. However, as was discussed, uncertainty–identity theory diVers from these other constructs and theories in a number of ways—it is broader than they are, it identifies social identity processes as the mediator in the relation between uncertainty and group phenomena, it specifies the way in which these social identity processes operate to translate uncertainty into group phenomena, it places self‐conception and identity at the heart of the problematic, it treats uncertainty as a state rather than a trait, and it describes how extremism and totalism can emerge from ordinary processes associated with ordinary groups and ordinary identities. Uncertainty is a pervasive part of life—we get excited and stimulated by it, we get frightened and oppressed by it, and we do what we can to reduce, control, or avoid it. We can never be truly certain, so we are always more or less uncertain. In this chapter, I have sketched out a model of how uncertainty may be related to why and how we identify with groups and to the types of groups that we identify with—suggesting that extreme uncertainty may encourage zealotry and totalism. These last are the bane of human existence—at best, producing ineYcient and oppressive groups and at worst, causing immeasurable human suVering.
Acknowledgments The research program reported in this chapter has been generously supported by the award of an Australian Research Council Professorial Fellowship, three research grants from the Australian Research Council, a research grant from The Leverhulme Trust, and travel grants from the British Academy and the British Psychological Society.
114
MICHAEL A. HOGG
References Aberson, C. L., Healy, M., & Romero, V. (2000). Ingroup bias and self‐esteem: A meta‐analysis. Personality and Social Psychology Review, 4, 157–173. Abrams, D., & Hogg, M. A. (1988). Comments on the motivational status of self‐esteem in social identity and intergroup discrimination. European Journal of Social Psychology, 18, 317–334. Abrams, D., & Hogg, M. A. (2004). Metatheory: Lessons from social identity research. Personality and Social Psychology Review, 8, 98–106. Abrams, D., Hogg, M. A., Hinkle, S., & Otten, S. (2005). The social identity perspective on small groups. In M. S. Poole & A. B. Hollingshead (Eds.), Theories of small groups: Interdisciplinary perspectives (pp. 99–137). Thousand Oaks, CA: Sage. Abrams, D., Wetherell, M., Cochrane, S., Hogg, M. A., & Turner, J. C. (1990). Knowing what to think by knowing who you are: Self‐categorization and the nature of norm formation, conformity and group polarization. British Journal of Social Psychology, 29, 97–119. Adorno, T. W., Frenkel‐Brunswik, E., Levinson, D. J., & Sanford, R. M. (1950). The authoritarian personality. New York: Harper. Ajzen, I. (1989). Attitude structure and behavior. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp. 241–274). Hillsdale, NJ: Erlbaum. Ajzen, I., & Fishbein, M. (1977). Attitude‐behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletin, 84, 888–918. Altemeyer, B. (1998). The other ‘‘authoritarian personality’’. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 47–92). Orlando, FL: Academic Press. Altemeyer, B. (2003). Why do religious fundamentalists tend to be prejudiced. International Journal for the Psychology of Religion, 13, 17–28. Arrow, H., McGrath, J. E., & Berdahl, J. L. (2000). Small groups as complex systems: Formation, coordination, development, and adaptation. Thousand Oaks, CA: Sage. Barber, B. R. (1995). Jihad vs. McWorld. New York: Random House. Baron, R. S., Crawley, K., & Paulina, D. (2003). Aberrations of power: Leadership in totalist groups. In D. van Knippenberg & M. A. Hogg (Eds.), Leadership and power: Identity processes in groups and organizations (pp. 169–183). London: Sage. Bartlett, F. C. (1932). Remembering. Cambridge, UK: Cambridge University Press. Bashevkin, S. B. (1991). True patriot love: The politics of Canadian nationalism. Toronto: Oxford University Press. Batson, C. D., Schoenrade, P., & Ventis, W. L. (1993). Religion and the individual: A social‐ psychological perspective. New York: Oxford University Press. Baumeister, R. F. (1987). How the self became a problem: A psychological review of historical research. Journal of Personality and Social Psychology, 52, 163–176. Baumeister, R. F. (1998). The self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 1, pp. 680–740). New York: McGraw‐Hill. Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism to violence and aggression: The dark side of high self‐esteem. Psychological Review, 103, 5–33. Bavelas, A. (1968). Communications patterns in task‐oriented groups. In D. Cartwright & A. Zander (Eds.), Group dynamics: Research and theory (3rd ed., pp. 503–511). London: Tavistock. Berger, C. E., & Calabrese, R. J. (1975). Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1, 99–112.
UNCERTAINTY–IDENTITY THEORY
115
Berger, C. R. (1987). Communicating under uncertainty. In M. E. RoloV & G. R. Miller (Eds.), Interpersonal processes: New directions in communication research (pp. 39–62). Newbury Park, CA: Sage. Billig, M. (1978). Fascists: A social psychological view of the National Front. London: Harcourt Brace Jovanovich. Billig, M. (1982). Ideology and social psychology: Extremism, moderation and contradiction. London: Sage. Billig, M. (1985). Prejudice, categorization and particularization: From a perceptual to a rhetorical approach. European Journal of Social Psychology, 15, 79–103. Bourhis, R. Y., Sachdev, I., & Gagnon, A. (1994). Intergroup research with the Tajfel matrices: Methodological notes. In M. Zanna & J. Olson (Eds.), The psychology of prejudice: The Ontario symposium (Vol. 7, pp. 209–232). Hillsdale, NJ: Erlbaum. Bradac, J. J. (2001). Theory comparison: Uncertainty reduction, problematic integration, uncertainty management, and other curious constructs. Journal of Communication, 51, 456–476. Brewer, M. B. (1991). The social self: On being the same and diVerent at the same time. Personality and Social Psychology Bulletin, 17, 475–482. Brewer, M. B. (1999). The psychology of prejudice: Ingroup love or outgroup hate? Journal of Social Issues, 55, 429–444. Brewer, M. B., & Campbell, D. T. (1976). Ethnocentrism and intergroup attitudes: East African evidence. New York: Sage. Brewer, M. B., & Gardner, W. (1996). Who is this ‘We’? Levels of collective identity and self representation. Journal of Personality and Social Psychology, 71, 83–93. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116–131. Cadinu, M. R., & Rothbart, M. (1996). Self‐anchoring and diVerentiation processes in the minimal group setting. Journal of Personality and Social Psychology, 70, 661–677. Campbell, D. T. (1958). Common fate, similarity, and other indices of the status of aggregates of persons as social entities. Behavioral Science, 3, 14–25. Campbell, J. D. (1990). Self‐esteem and the clarity of the self‐concept. Journal of Personality and Social Psychology, 59, 538–549. Campbell, J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., Lavalle, L. F., & Lehman, D. R. (1996). Self‐concept clarity: Measurement, personality correlates, and cultural boundaries. Journal of Personality and Social Psychology, 70, 141–156. Castano, E. (2004). On the advantages of reifying the ingroup. In V. Yzerbyt, C. M. Judd, & O. Corneille (Eds.), The psychology of group perception: Perceived variability, entitativity, and essentialism (pp. 381–400). New York: Psychology Press. Castano, E., Yzerbyt, V. Y., & Bourguignon, D. (2003). We are one and I like it: The impact of ingroup entitativity on ingroup identification. European Journal of Social Psychology, 33, 735–754. Clement, R. W., & Krueger, J. (2000). The primacy of self‐referent information in perceptions of social consensus. British Journal of Social Psychology, 39, 279–299. Crisp, R. J., Farr, M. J., Beck, S. R., Hogg, M. A., & Cortez, B. (2006). Self‐stereotyping and self‐projection: Definitional processes in novel and established groups. Manuscript submitted for publication, University of Birmingham. Crisp, R. J., & Hogg, M. A. (2006). Depersonalization and projection in groups: Two paths to uncertainty reduction and self‐construal. Manuscript submitted for publication: University of Birmingham. Crocker, J., & Luhtanen, R. (1990). Collective self‐esteem and ingroup bias. Journal of Personality and Social Psychology, 58, 60–67.
116
MICHAEL A. HOGG
Crocker, J., & Major, B. (1994). Reactions to stigma: The moderating role of justifications. In M. P. Zanna & J. M. Olson (Eds.), The psychology of prejudice: The Ontario symposium (Vol. 7, pp. 289–314). Hillsdale, NJ: Erlbaum. Crockett, W. H. (1965). Cognitive complexity and impression formation. In B. A. Maher (Ed.), Progress in experimental personality research (Vol. 2, pp. 47–90). New York: Academic Press. Curtis, J. M., & Curtis, M. J. (1993). Factors related to susceptibility and recruitment by cults. Psychological Reports, 73, 451–460. Dechesne, M., Janssen, J., & van Knippenberg, A. (2000). Derogation and distancing as terror management strategies: The moderating role of need for closure and permeability of group boundaries. Journal of Personality and Social Psychology, 79, 923–932. Deconchy, J.‐P. (1984). Rationality and social control in orthodox systems. In H. Tajfel (Ed.), The social dimension: European developments in social psychology (Vol. 2, pp. 425–445). Cambridge, UK: Cambridge University Press. de Cremer, D., & van Vugt, M. (Cremer 1999). Social identification eVects in social dilemmas: A transformation of motives. European Journal of Social Psychology, 29, 871–893. Diehl, M. (1990). The minimal group paradigm: Theoretical explanations and empirical findings. European Review of Social Psychology, 1, 263–292. Dijksterhuis, A., van Knippenberg, A., Kruglanski, A. W., & Schaper, C. (1996). Motivated social cognition: Need for closure eVects on memory and judgment. Journal of Experimental Social Psychology, 32, 254–270. Doty, R. M., Peterson, B. E., & Winter, D. G. (1991). Threat and authoritarianism in the United States, 1978–1987. Journal of Personality and Social Psychology, 61, 629–640. Duckitt, J. (2001). A dual‐process cognitive‐motivational theory of ideology and prejudice. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 33, pp. 41–113). San Diego, CA: Academic Press. Dunn, R. G. (1998). Identity crises: A social critique of postmodernity. Minneapolis, MN: University of Minnesota Press. Ellemers, N., de Gilder, D., & Haslam, S. A. (2004). Motivating individuals and groups at work: A social identity perspective on leadership and group performance. Academy of Management Review, 29, 459–478. Festinger, L. (1954a). A theory of social comparison processes. Human Relations, 7, 117–140. Festinger, L. (1954b). Motivation leading to social behavior. In M. R. Jones (Ed.), Nebraska symposium on motivation (Vol. 2, pp. 121–218). Lincoln, NE: University of Nebraska Press. Fiske, A. P., Kitayama, S., Markus, H. R., & Nisbett, R. E. (1998). The cultural matrix of social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 2, pp. 915–981). New York: McGraw‐Hill. Fletcher, G. J. O., Danilovics, P., Fernandez, G., Peterson, D., & Reeder, G. D. (1986). Attributional complexity: An individual diVerences measure. Journal of Personality and Social Psychology, 51, 875–884. Ford, L. A., Babrow, A. S., & Stohl, C. (1996). Social support and the management of uncertainty: An application of problematic integration theory. Communication Monographs, 63, 189–207. Freeman, J. (1973). The tyranny of structurelessness. In A. Koedt, E. Levine, & A. Rapone (Eds.), Radical feminism (pp. 285–299). New York: Quadrangle Books. Fromm, E. (1947). Man for himself: An inquiry into the psychology of ethics. New York: Rinehart. Furnham, A. (1990). The Protestant work ethic: The psychology of work‐related beliefs and behaviors. New York: Routledge.
UNCERTAINTY–IDENTITY THEORY
117
Furnham, A. (2003). Belief in a just world: Research progress over the past decade. Personality and Individual DiVerences, 34, 795–817. Furnham, A., & Procter, E. (1989). Belief in a just world: Review and critique of the individual diVerence literature. British Journal of Social Psychology, 28, 365–384. Gaertner, L., Sedikides, C., & Graetz, K. (1999). In search of self‐definition: Motivational primacy of the individual self, motivational primacy of the collective self, or contextual primacy? Journal of Personality and Social Psychology, 76, 5–18. Galanter, M. (Ed.) (1989). Cults and new religious movements. Washington, DC: American Psychiatric Association. Gergen, K. J. (1991). The saturated self: Dilemmas of identity in contemporary life. New York: Basic Books. Gollwitzer, P. M. & Bargh, J. A. (Eds.) (1996). The psychology of action: Linking cognition and motivation to behavior. New York: Guilford Press. Greco, V., & Roger, D. (2003). Uncertainty, stress, and health. Personality and Individual DiVerences, 34, 1057–1068. Greenberg, J., Koole, S. L., & Pyszczynski, T. (Eds.) (2004). Handbook of experimental existential psychology. New York: Guilford. Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S., et al. (1990). Evidence for terror management theory II: The eVects of mortality salience on reactions to those who threaten or bolster the cultural worldview. Journal of Personality and Social Psychology, 58, 308–318. Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory of self‐esteem and cultural worldviews: Empirical assessments and conceptual refinements. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 61–139). San Diego, CA: Academic Press. Grieve, P., & Hogg, M. A. (1999). Subjective uncertainty and intergroup discrimination in the minimal group situation. Personality and Social Psychology Bulletin, 25, 926–940. Gudykunst, W. B. (1985). The influence of cultural similarity, type of relationship, and self‐monitoring on uncertainty reduction processes. Communication Monographs, 52, 203–217. Hackman, J. R. (2002). Leading teams: Setting the stage for great performances. Boston, MA: Harvard Business School Press. Hains, S. C., Hogg, M. A., & Duck, J. M. (1997). Self‐categorization and leadership: EVects of group prototypicality and leader stereotypicality. Personality and Social Psychology Bulletin, 23, 1087–1100. Hamilton, D. L., & Sherman, S. J. (1996). Perceiving persons and groups. Psychological Review, 103, 336–355. Hamilton, D. L., Sherman, S. J., & Rodgers, J. S. (2004). Perceiving the groupness of groups: Entitativity, homogeneity, essentialism, and stereotypes. In V. Yzerbyt, C. M. Judd, & O. Corneille (Eds.), The psychology of group perception: Perceived variability, entitativity, and essentialism (pp. 39–60). New York: Psychology Press. Harmon‐Jones, E., Greenberg, J., & Solomon, S. (1996). The eVects of mortality salience on intergroup bias between minimal groups. European Journal of Social Psychology, 26, 677–681. Haslam, N., Rothschild, L., & Ernst, D. (1998). Essentialist beliefs about social categories. British Journal of Social Psychology, 39, 113–127. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Heine, S. J., Proulx, T., & Vohs, K. D. (2006). The meaning maintenance model: On the coherence of social motivations. Personality and Social Psychology Review, 10, 88–110.
118
MICHAEL A. HOGG
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 1–46). New York: Academic Press. HoVer, E. (1951). The true believer. New York: Time. Hofstede, G. (1980). Culture’s consequences: International diVerences in work‐related values. Beverly Hills, CA: Sage. Hogg, M. A. (2000a). Subjective uncertainty reduction through self‐categorization: A motivational theory of social identity processes. European Review of Social Psychology, 11, 223–255. Hogg, M. A. (2000b). Social identity and social comparison. In J. Suls & L. Wheeler (Eds.), Handbook of social comparison: Theory and research (pp. 401–421). New York: Kluwer/ Plenum. Hogg, M. A. (2001a). Self‐categorization and subjective uncertainty resolution: Cognitive and motivational facets of social identity and group membership. In J. P. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behavior (pp. 323–349). New York: Cambridge University Press. Hogg, M. A. (2001b). A social identity theory of leadership. Personality and Social Psychology Review, 5, 184–200. Hogg, M. A. (2003). Social identity. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 462–479). New York: Guilford. Hogg, M. A. (2004). Uncertainty and extremism: Identification with high entitativity groups under conditions of uncertainty. In V. Yzerbyt, C. M. Judd, & O. Corneille (Eds.), The psychology of group perception: Perceived variability, entitativity, and essentialism (pp. 401–418). New York: Psychology Press. Hogg, M. A. (2005a). Uncertainty, social identity, and ideology. In S. R. Thye & E. J. Lawler (Eds.), Advances in group processes (Vol. 22, pp. 203–229). San Diego, CA: Elsevier. Hogg, M. A. (2005b). The social identity perspective. In S. A. Wheelan (Ed.), The handbook of group research and practice (pp. 133–157). Thousand Oaks, CA: Sage. Hogg, M. A. (2005c). All animals are equal but some animals are more equal than others: Social identity and marginal membership. In K. D. Williams, J. P. Forgas, & W. von Hippel (Eds.), The social outcast: Ostracism, social exclusion, rejection, and bullying (pp. 243–261). New York: Psychology Press. Hogg, M. A. (2006a). Self‐conceptual uncertainty and the lure of belonging. In R. J. Brown & D. Capozza (Eds.), Social identities: Motivational, emotional, cultural influences (pp. 33–49). New York: Psychology Press. Hogg, M. A. (2006b). Social identity theory. In P. J. Burke (Ed.), Contemporary social psychological theories (pp. 111–136). Palo Alto, CA: Stanford University Press. Hogg, M. A. (2007). Organizational orthodoxy and corporate autocrats: Some nasty consequences of organizational identification in uncertain times. In C. A. Bartel, S. Blader, and A. Wrzesniewski (Eds.), Identity and the modern organization (pp. 35–59). Mahwah, NJ: Erlbaum. Hogg, M. A. (in press). Personality, individuality and social identity. In F. Rhodewalt (Ed.), Personality and social behavior. New York: Psychology Press. Hogg, M. A., & Abrams, D. (1993). Towards a single‐process uncertainty‐reduction model of social motivation in groups. In M. A. Hogg & D. Abrams (Eds.), Group motivation: Social psychological perspectives (pp. 173–190). New York: Prentice Hall. Hogg, M. A., & Abrams, D. (2003). Intergroup behavior and social identity. In M. A. Hogg & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 407–431). London: Sage. Hogg, M. A., Fielding, K. S., & Darley, J. (2005). Fringe dwellers: Processes of deviance and marginalization in groups. In D. Abrams, M. A. Hogg, & J. M. Marques (Eds.), The social psychology of inclusion and exclusion (pp. 191–210). New York: Psychology Press.
UNCERTAINTY–IDENTITY THEORY
119
Hogg, M. A., & Grieve, P. (1999). Social identity theory and the crisis of confidence in social psychology: A commentary, and some research on uncertainty reduction. Asian Journal of Social Psychology, 2, 79–93. Hogg, M. A., & Hains, S. C. (1996). Intergroup relations and group solidarity: EVects of group identification and social beliefs on depersonalized attraction. Journal of Personality and Social Psychology, 70, 295–309. Hogg, M. A., Meehan, C., Parsons, A., Farquharson, J., & Svensson, A. (2006). Identification with moderate and extreme groups: The role of self‐uncertainty. Manuscript submitted for publication: Claremont Graduate University. Hogg, M. A., & Mullin, B.‐A. (1999). Joining groups to reduce uncertainty: Subjective uncertainty reduction and group identification. In D. Abrams & M. A. Hogg (Eds.), Social identity and social cognition (pp. 249–279). Oxford, UK: Blackwell. Hogg, M. A., Sherman, D. K., Dierselhuis, J., Maitner, A. T., & MoYtt, G. (2007). Uncertainty, entitativity, and group identification. Journal of Experimental Social Psychology, 43, 135–142. Hogg, M. A., & Smith J. R. (in press). Attitudes in social context: A social identity perspective. European Review of Social Psychology. Hogg, M. A., & Svensson, A. (2006). Uncertainty reduction, self‐esteem and group identification. Manuscript submitted for publication: Claremont Graduate University. Hogg, M. A., & van Knippenberg, D. (2003). Social identity and leadership processes in groups. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 35, pp. 1–52). San Diego, CA: Academic Press. Hornsey, M. J. (2005). Why being right is not enough: Predicting defensiveness in the face of group criticism. European Review of Social Psychology, 16, 301–334. Isenberg, D. J. (1986). Group polarization: A critical review. Journal of Personality and Social Psychology, 50, 1141–1151. Janis, I. L. (1972). Victims of groupthink: A psychological study of foreign policy decisions and fiascoes. Boston, MA: Houghton‐MiZin. Jetten, J., Hogg, M. A., & Mullin, B.‐A. (2000). Ingroup variability and motivation to reduce subjective uncertainty. Group Dynamics: Theory, Research, and Practice, 4, 184–198. Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology, 25, 882–920. Jost, J. T., Fitzsimons, G., & Kay, A. C. (2004). The ideological animal: A comparison of system justification, terror management, and just world theories. In J. Greenberg, S. L. Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 263–283). New York: Guilford. Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as motivated social cognition. Psychological Bulletin, 129, 339–375. Jost, J. T., & Hunyady, O. (2002). The psychology of system justification and the palliative function of ideology. European Review of Social Psychology, 13, 111–153. Jost, J. T. & Major, B. (Eds.) (2001). The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations. New York: Cambridge University Press. Kenrick, D., Li, N. P., & Butner, J. (2003). Dynamical evolutionary psychology: Individual decision rules and emergent social norms. Psychological Review, 110, 3–28. Klandermans, B. (1997). The social psychology of protest. Oxford, UK: Blackwell. Kosterman, R., & Feshbach, S. (1989). Towards a measure of patriotic and nationalistic attitudes. Political Psychology, 10, 257–274. Kramer, R. M., & Wei, J. (1999). Social uncertainty and the problem of trust in social groups: The social self in doubt. In T. R. Tyler, R. M. Kramer, & O. P. John (Eds.), The psychology of the social self (pp. 145–168). Mahwah, NJ: Erlbaum.
120
MICHAEL A. HOGG
Kraus, S. J. (1995). Attitudes and the prediction of behavior: A meta‐analysis of the empirical literature. Personality and Social Psychology Bulletin, 21, 58–75. Krueger, J., & Clement, R. W. (1996). Inferring category characteristics from sample characteristics: Inductive reasoning and social projection. Journal of Experimental Psychology: General, 128, 52–68. Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and motivational bases. New York: Plenum Press. Kruglanski, A. W. (2004). The psychology of closed mindedness. New York: Psychology Press. Kruglanski, A. W., Pierro, A., Mannetti, L., & De Grada, E. (2006). Groups as epistemic providers: Need for closure and the unfolding of group‐centrism. Psychological Review, 113, 84–100. Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: ‘Seizing’ and ‘freezing’. Psychological Review, 103, 263–283. Kuhn, T. (1962). The structure of scientific revolutions. Chicago, IL: University of Chicago Press. Lambert, A. J., Burroughs, T., & Nguyen, T. (1999). Perceptions of risk and the buVering hypothesis: The role of just world beliefs and right‐wing authoritarianism. Personality and Social Psychology Bulletin, 25, 643–656. Landau, M. J., Johns, M., Greenberg, J., Pyszczynski, T., Martens, A., Goldenberg, J. L., et al. (2004). A function of form: Terror management and structuring the social world. Journal of Personality and Social Psychology, 87, 190–210. Larrain, J. (1979). The concept of ideology. London: Hutchinson. Leary, M. R. & Tangney, J. P. (Eds.) (2003). Handbook of self and identity. New York: Guilford. Lester, R. E. (1987). Organizational culture, uncertainty reduction, and the socialization of new organizational members. In S. Thomas (Ed.), Culture and communication: Methodology, behavior, artifacts, and institutions (pp. 105–113). Norwood, NJ: Ablex. Levine, J. M., & Moreland, R. L. (2002). Group reactions to loyalty and disloyalty. In E. Lawler & S. Thye (Eds.), Advances in group processes (Group cohesion, trust and solidarity) (Vol. 19, pp. 203–228). Amsterdam: Elsevier Science. Lewis, B. (2004). The crisis of Islam: Holy war and unholy terror. London: Phoenix. Lickel, B., Hamilton, D. L., Wieczorkowska, G., Lewis, A., Sherman, S. J., & Uhles, A. N. (2000). Varieties of groups and the perception of group entitativity. Journal of Personality and Social Psychology, 78, 223–246. Lickel, B., Rutchick, A. M., Hamilton, D. L., & Sherman, S. J. (2006). Intuitive theories of group types and relational principles. Journal of Experimental Social Psychology, 42, 28–39. Linville, P. W. (1987). Self‐complexity as a buVer against stress‐related illness and depression. Journal of Personality and Social Psychology, 52, 663–676. Lippitt, R., & White, R. (1943). The ‘social climate’ of children’s groups. In R. G. Barker, J. Kounin, & H. Wright (Eds.), Child behavior and development (pp. 485–508). New York: McGraw‐Hill. Long, M. K., & Spears, R. (1997). The self‐esteem hypothesis revisited: DiVerentiation and the disaVected. In R. Spears, P. J. Oakes, N. Ellemers, & S. A. Haslam (Eds.), The social psychology of stereotyping and group life (pp. 296–317). Oxford, UK: Blackwell. Lopes, L. L. (1987). Between hope and fear: The psychology of risk. Advances in Experimental Psychology, 20, 255–295. Lorenz, E. N. (1993). The essence of chaos. London: University College London Press. Luhtanen, R., & Crocker, J. (1992). A collective self‐esteem scale: Self‐evaluation of one’s social identity. Personality and Social Psychology Bulletin, 18, 302–318. Major, B., Quinton, W. J., & McCoy, S. (2002). Antecedents and consequences of attributions to discrimination: Theoretical and empirical advances. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 251–330). San Diego, CA: Academic Press.
UNCERTAINTY–IDENTITY THEORY
121
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253. Marques, J. M., Abrams, D., Pa´ez, D., & Hogg, M. A. (2001). Social categorization, social identification, and rejection of deviant group members. In M. A. Hogg & R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes (pp. 400–424). Oxford, UK: Blackwell. Marques, J. M., Abrams, D., & Serodio, R. (2001). Being better by being right: Subjective group dynamics and derogation of in‐group deviants when generic norms are undermined. Journal of Personality and Social Psychology, 81, 436–447. Marris, P. (1996). The politics of uncertainty: Attachment in private and public life. London: Routledge. Martin, R., & Hewstone, M. (2003). Social influence processes of control and change: Conformity, obedience to authority, and innovation. In M. A. Hogg & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 347–366). London: Sage. Maslow, A. H. (1987). Motivation and personality (3rd ed.). New York: Harper Collins. McGarty, C., Turner, J. C., Oakes, P. J., & Haslam, S. A. (1993). The creation of uncertainty in the influence process: The roles of stimulus information and disagreement with similar others. European Journal of Social Psychology, 23, 17–38. McGregor, I., & Marigold, D. C. (2003). Defensive zeal and the uncertain self: What makes you so sure? Journal of Personality and Social Psychology, 85, 838–852. McGregor, I., Nail, P. R., Marigold, D. C., & Kang, S. (2005). Defensive pride and consensus: Strength in imaginary numbers. Journal of Personality and Social Psychology, 89, 978–996. McGregor, I., Zanna, M. P., Holmes, J. G., & Spencer, S. J. (2001). Compensatory conviction in the face of personal uncertainty: Going to extremes and being oneself. Journal of Personality and Social Psychology, 80, 472–488. Medin, D. L., & Ortony, A. (1989). Psychological essentialism. In S. Vosnaidou & A. Ortony (Eds.), Similarity and analogical reasoning (pp. 179–195). Cambridge, UK: Cambridge University Press. Mikulincer, M., Florian, V., & Hirschberger, G. (2004). The terror of death and the quest for love: An existential perspective on close relationships. In J. Greenberg, S. L. Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 287–304). New York: Guilford. Minard, R. D. (1952). Race relations in the Pocahontas coal field. Journal of Social Issues, 8, 29–44. Moghaddam, F. M. & Marsella, A. J. (Eds.) (2004). Understanding terrorism: Psychosocial roots, consequences, and interventions. Washington, DC: American Psychological Association. Montoya, R. M., Rosenthal, S. A., & Pittinsky, T. L. (2006). It’s all about we: Collective narcissism predicts ingroup bias and outgroup denigration. Unpublished data. Harvard University. Moscovici, S. (1976). Social influence and social change. London: Academic Press. Moscovici, S. (1980). Toward a theory of conversion behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 13, pp. 202–239). New York: Academic Press. Moscovici, S., & Zavalloni, M. (1969). The group as a polarizer of attitudes. Journal of Personality and Social Psychology, 12, 125–135. Mugny, G., & Pe´rez, J. A. (1991). The social psychology of minority influence. Cambridge, UK: Cambridge University Press. Mullin, B.‐A., & Hogg, M. A. (1998). Dimensions of subjective uncertainty in social identification and minimal intergroup discrimination. British Journal of Social Psychology, 37, 345–365.
122
MICHAEL A. HOGG
Mullin, B.‐A., & Hogg, M. A. (1999). Motivations for group membership: The role of subjective importance and uncertainty reduction. Basic and Applied Social Psychology, 21, 91–102. Neuberg, S. L., & Newson, J. T. (1993). Personal need for structure: Individual diVerences in the desire for simpler structure. Journal of Personality and Social Psychology, 65, 113–131. Nisbet, R. (1959). The decline and fall of social class. Pacific Sociological Review, 2, 11–17. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood CliVs, NJ: Prentice Hall. Oakes, P. J. (1987). The salience of social categories. In J. C. Turner, M. A. Hoggce, P. J. Oakes, S. D. Reicher, & M. S. Wetherell (Eds.), Rediscovering the social group: A self‐categorization theory (pp. 117–141). Oxford, UK: Blackwell. Otten, S. (2002). ‘‘Me and us’’ or ‘‘us and them’’? The self as a heuristic for defining minimal ingroups. European Review of Social Psychology, 13, 1–33. Otten, S., & Moskowitz, G. B. (2000). Evidence for implicit evaluative in‐group bias: AVect‐ based spontaneous trait inference in a minimal group paradigm. Journal of Experimental Social Psychology, 36, 77–89. Otten, S., & Wentura, D. (1999). About the impact of automaticity in the minimal group paradigm: Evidence from aVective priming tasks. European Journal of Social Psychology, 29, 1049–1071. Otten, S., & Wentura, D. (2001). Self‐anchoring and ingroup favoritism: An individual‐profiles analysis. Journal of Experimental Social Psychology, 37, 525–532. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta‐analyses. Psychological Bulletin, 128, 3–72. Pettigrew, T. F. (1958). Personality and sociocultural factors in intergroup attitudes: A cross‐ national comparison. Journal of Conflict Resolution, 2, 29–42. Pickett, C. L., & Brewer, M. B. (2001). Assimilation and diVerentiation needs as motivational determinants of perceived ingroup and outgroup homogeneity. Journal of Experimental Social Psychology, 37, 341–348. Pickett, C. L., Silver, M. D., & Brewer, M. B. (2002). The impact of assimilation and diVerentiation needs on perceived group importance and judgments of ingroup size. Personality and Social Psychology Bulletin, 28, 546–558. Pittinsky, T. L., Montoya, R. M., & Rosenthal, S. A. (2006). Individual and collective narcissism: The influence of culture. Unpublished data. Harvard University. Platow, M. J., Reid, S. A., & Andrew, S. (1998). Leadership endorsement: The role of distributive and procedural behavior in interpersonal and intergroup contexts. Group Processes and Intergroup Relations, 1, 35–47. Platow, M. J., & van Knippenberg, D. (2001). A social identity analysis of leadership endorsement: The eVects of leader ingroup prototypicality and distributive intergroup fairness. Personality and Social Psychology Bulletin, 27, 1508–1519. Pollock, H. N. (2003). Uncertain science . . . uncertain world. Cambridge, UK: Cambridge University Press. Popper, K. R. (1959). The logic of scientific discovery. New York: Basic Books. Pyszczynski, T., Greenberg, J., & Solomon, S. (1999). A dual‐process model of defense against conscious and unconscious death‐related thoughts: An extension of terror management theory. Psychological Review, 106, 835–845. Pyszczynski, T., Greenberg, J., & Solomon, S. (2004). Why do people need self‐esteem? A theoretical and empirical review. Psychological Bulletin, 130, 435–468. Pyszczynski, T., Wicklund, R. A., Floresku, S., Koch, H., Gauch, G., Solomon, S., et al. (1996). Whistling in the dark: Exaggerated consensus estimates in response to incidental reminders of mortality. Psychological Science, 7, 332–336.
UNCERTAINTY–IDENTITY THEORY
123
Quinn, D. M., & Crocker, J. (1999). When ideology hurts: EVects of belief in the Protestant work ethic and feeling overweight on the psychological well‐being of women. Journal of Personality and Social Psychology, 77, 402–414. Reicher, S. D. (2001). The psychology of crowd dynamics. In M. A. Hogg & R. S. Tindale (Eds.), Blackwell handbook of social psychology: Group processes (pp. 182–207). Oxford, UK: Blackwell. Reid, S. A., & Hogg, M. A. (2005). Uncertainty reduction, self‐enhancement, and ingroup identification. Personality and Social Psychology Bulletin, 31, 804–817. Rohan, M. J. (2000). A rose by any name? The values construct. Personality and Social Psychology Review, 4, 255–277. Rokeach, M. (1960). The open and closed mind. New York: Basic Books. Rokeach, M. (1973). The nature of human values. New York: Free Press. Rothbart, M., & Taylor, M. (1992). Category labels and social reality: Do we view social categories as natural kinds? In G. R. Semin & K. Fiedler (Eds.), Language, interaction and sociul cognition (pp. 11–36). London: Sage. Rowatt, W. C., & Franklin, L. M. (2004). Christian orthodoxy, religious fundamentalism, and right‐wing authoritarianism as predictors of implicit racial prejudice. International Journal for the Psychology of Religion, 14, 125–138. Rubin, M., & Hewstone, M. (1998). Social identity theory’s self‐esteem hypothesis: A review and some suggestions for clarification. Personality and Social Psychology Review, 2, 40–62. Saks, A. M., & Ashforth, B. E. (1997). Organizational socialization: Making sense of the past and present as a prologue for the future. Journal of Vocational Behavior, 51, 234–279. Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 cultures. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 1–65). San Diego, CA: Academic Press. Sedikides, C., & Gregg, A. P. (2003). Portraits of the self. In M. A. Hogg & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 110–138). London: Sage. Sedikides, C., & Strube, M. J. (1997). Self‐evaluation: To thine own self be good, to thine own self be sure, to thine own self be true, and to thine own self be better. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 29, pp. 209–296). New York: Academic Press. Shah, J. Y., Kruglanski, A. W., & Thompson, E. P. (1998). Membership has its (epistemic) rewards: Need for closure eVects on in‐group bias. Journal of Personality and Social Psychology, 75, 383–393. Sherman, D. K., & Cohen, G. L. (2006). The psychology of self‐defense: Self‐aYrmation theory. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 38, pp. 183–242). San Diego, CA: Academic Press. Sherman, D. K., Hogg, M. A., Maitner, A. T., & MoYtt, G. (2006). Polarization under uncertainty: Democrats, Republicans and strikers. Unpublished data. University of California: Santa Barbara. Showers, C. (1992). Compartmentalization of positive and negative self‐knowledge: Keeping bad apples out of the bunch. Journal of Personality and Social Psychology, 62, 1036–1049. Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and oppression. New York: Cambridge University Press. Singer, M. (1995). Cults in our midst. San Francisco, CA: Jossey‐Bass. Smith, J. R., Hogg, M. A., Martin, R., & Terry, D. J. (in press). Uncertainty and the influence of group norms in the attitude‐behaviour relationship. British Journal of Social Psychology. Smith, P. B., Bond, M. H., & Kag˘itc¸ibasi, C. (2006). Understanding social psychology across cultures. London: Sage.
124
MICHAEL A. HOGG
Sorrentino, R. M., Hodson, G., & Huber, G. L. (2001). Uncertainty orientation and the social mind: Individual diVerences in the interpersonal context. In J. P. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behavior (pp. 199–227). New York: Cambridge University Press. Sorrentino, R. M., & Roney, C. J. R. (1986). Uncertainty orientation, achievement‐related motivation and task diagnosticity as determinants of task performance. Social Cognition, 4, 420–436. Sorrentino, R. M., & Roney, C. J. R. (1999). The uncertain mind: Individual diVerences in facing the unknown. Philadelphia, PA: Psychology Press. Sorrentino, R. M., & Short, J. C. (1986). Uncertainty orientation, motivation and cognition. In R. M. Sorrentino & E. T. Higgins (Eds.), The handbook of motivation and cognition: Foundations of social behavior (Vol. 1, pp. 379–403). New York: Guilford. Staub, E. (1989). The roots of evil: The psychological and cultural origins of genocide and other forms of group violence. New York: Cambridge University Press. Staub, E. (1997). Blind versus constructive patriotism: Moving from embeddedness in the group to critical loyalty and action. In D. Bar‐Tal & E. Staub (Eds.), Patriotism: In the lives of individuals and nations (pp. 213–228). Chicago: Nelson‐Hall. Steele, C. M. (1988). The psychology of self‐aYrmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261–302). New York: Academic Press. Stephan, W. G., & Stephan, C. W. (1985). Intergroup anxiety. Journal of Social Issues, 41, 157–175. Stu¨rmer, S., & Simon, B. (2004). Collective action: Towards a dual‐pathway model. European Review of Social Psychology, 15, 59–99. Suls, J. & Wheeler, L. (Eds.) (2000). Handbook of social comparison: Theory and research. New York: Kluwer/Plenum. Swann, W. B., Jr., Polzer, J. T., Seyle, C., & Ko, S. (2004). Finding value in diversity: Verification of personal and social self‐views in diverse groups. Academy of Management Review, 29, 9–27. Swann, W. B., Jr., Rentfrow, P. J., & Guinn, J. S. (2003). Self‐verification: The search for coherence. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 367–383). New York: Guilford. Swann, W. R. (1977). The trouble with change: Self‐verification and allegiance to the self. Psychological Science, 8, 177–180. Tajfel, H. (1959). Quantitative judgement in social perception. British Journal of Psychology, 50, 16–29. Tajfel, H. (1969). Cognitive aspects of prejudice. Journal of Social Issues, 25, 79–97. Tajfel, H. (1972). Social categorization. English manuscript of ‘La cate´gorisation sociale’. In S. Moscovici (Ed.), Introduction a` la psychologie sociale (Vol. 1, pp. 272–302). Paris: Larousse. Tajfel, H. (1974). Intergroup behaviour, social comparison and social change. Unpublished Katz‐ Newcomb lectures, University of Michigan: Ann Arbor. Tajfel, H., & Billig, M. (1974). Familiarity and categorization in intergroup behavior. Journal of Experimental Social Psychology, 10, 159–170. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole. Taylor, S. E. (1998). The social being in social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed.,Vol. 1, pp. 58–95). New York: McGraw‐Hill. Terry, D. J., Carey, C. J., & Callan, V. J. (2001). Employee adjustment to an organizational merger: An intergroup perspective. Personality and Social Psychology Bulletin, 27, 267–280.
UNCERTAINTY–IDENTITY THEORY
125
Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude–behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22, 776–793. Terry, D. J., Hogg, M. A., & White, K. M. (2000). Attitude–behavior relations: Social identity and group membership. In D. J. Terry & M. A. Hogg (Eds.), Attitudes, behavior, and social context: The role of norms and group membership (pp. 67–93). Mahwah, NJ: Erlbaum. Thompson, J. B. (1990). Ideology and modern culture: Critical social theory in the era of mass communication. Stanford, CA: Stanford University Press. Triandis, H. C. (1989). The self and social behavior in diVering cultural contexts. Psychological Review, 96, 506–520. Trope, Y., & Gaunt, R. (2003). Attribution and person perception. In M. A. Hogg & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 190–208). London: Sage. Turner, J. C. (1975). Social comparison and social identity: Some prospects for intergroup behaviour. European Journal of Social Psychology, 5, 5–34. Turner, J. C. (1985). Social categorisation and the self‐concept: A social cognitive theory of the group. In E. J. Lawler (Ed.), Advances in group processes: Theory and Research (Vol. 2, pp. 77–122). Greenwich, CT: JAI Press. Turner, J. C. (1991). Social influence. Milton‐Keynes: Open University Press. Turner, J. C. (1999). Some current issues in research on social identity and self‐categorization theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity (pp. 6–34). Oxford, UK: Blackwell. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self‐categorization theory. Oxford, UK: Blackwell. Turner, J. C., & Onorato, R. (1999). Social identity, personality and the self‐concept: A self‐ categorization perspective. In T. R. Tyler, R. M. Kramer, & O. Johns (Eds.), The psychology of the social self (pp. 11–46). Mahwah, NJ: Erlbaum. Tyler, T. R. (1997). The psychology of legitimacy: A relational perspective on voluntary deference to authorities. Personality and Social Psychology Review, 1, 323–345. Tyler, T. R., & Lind, E. A. (1992). A relational model of authority in groups. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 25, pp. 115–191). New York: Academic Press. Tyler, T. R., & Smith, H. J. (1998). Social justice and social movements. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., pp. 595–629). Boston: McGraw‐Hill. van den Bos, K. (2001). Reactions to perceived fairness: The impact of mortality salience and self‐esteem on ratings of negative aVect. Social Justice Research, 14, 1–23. van den Bos, K., & Lind, E. A. (2002). Uncertainty management by means of fairness judgments. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 1–60). San Diego, CA: Academic Press. van den Bos, K., & Miedema, J. (2000). Toward understanding why fairness matters: The influence of mortality salience on reactions to procedural fairness. Journal of Personality and Social Psychology, 79, 355–366. van den Bos, K., Poortvliet, M., Maas, M., Miedema, J., & van den Ham, E. (2005). An enquiry concerning the principles of cultural norms and values: The impact of uncertainty and mortality salience on reactions to violations and bolstering of cultural worldviews. Journal of Experimental Social Psychology, 41, 91–113. van Knippenberg, D., & Hogg, M. A. (2003). A social identity model of leadership in organizations. In R. M. Kramer & B. M. Staw (Eds.), Research in organizational behavior (Vol. 25, pp. 243–295). Greenwich, CT: JAI Press. Van Vugt, M., & Hart, C. M. (2004). Social identity as social glue: The origins of group loyalty. Journal of Personality and Social Psychology, 86, 585–598.
126
MICHAEL A. HOGG
Vignoles, V. L., Regalia, C., Manzi, C., Golledge, J., & Scabini, E. (2006). Beyond self‐esteem: Influence of multiple motives on identity construction. Journal of Personality and Social Psychology, 90, 308–333. Webster, D. M., & Kruglanski, A. W. (1994). Individual diVerences in need for cognitive closure. Journal of Personality and Social Psychology, 67, 1049–1062. Webster, D. M., & Kruglanski, A. W. (1998). Cognitive and social consequences of the motivation for closure. European Review of Social Psychology, 8, 133–173. Wills, T. A. (1991). Similarity and self‐esteem in downward comparison. In J. Suls & T. A. Wills (Eds.), Social comparison: Contemporary theory and research (pp. 51–78). Hillsdale, NJ: Erlbaum. Yuki, M. (2003). Intergroup comparison versus intragroup relationships: A cross‐cultural examination of social identity theory in North American and East Asian cultural contexts. Social Psychology Quarterly, 66, 166–183. Yzerbyt, V., Castano, E., Leyens, J.‐P., & Paladino, M.‐P. (2000). The primacy of the ingroup: The interplay of entitativity and identification. European Review of Social Psychology, 11, 257–295. Yzerbyt, V., Judd, C. M., & Corneille, O. (2004). Perceived variability, entitativity, and essentialism: Introduction and overview. In V. Yzerbyt, C. M. Judd, & O. Corneille (Eds.), The psychology of group perception: Perceived variability, entitativity, and essentialism (pp. 1–22). New York: Psychology Press.
METACOGNITIVE EXPERIENCES AND THE INTRICACIES OF SETTING PEOPLE STRAIGHT: IMPLICATIONS FOR DEBIASING AND PUBLIC INFORMATION CAMPAIGNS Norbert Schwarz Lawrence J. Sanna Ian Skurnik Carolyn Yoon
I. Introduction Decades of psychological research documented that human judgment often falls short of normative ideals. Social and cognitive psychologists discovered an ever increasing number of systematic biases and illustrated their pervasive role in judgment and decision making (for reviews see Gilovich, GriYn, & Kahneman, 2002; Kahneman, Slovic, & Tversky, 1982; Kerr, MacCoun, & Kramer, 1996; Nisbett & Ross, 1980; Pohl, 2005). Similarly, researchers in applied fields, like health and consumer behavior, identified numerous erroneous beliefs that impair good decisions and prevent people from doing what would be in their best interest (Christensen, Moran, & Wiebe, 1999; Webley, Burgoyne, Lea, & Young, 2001). In both cases, the remedy seems obvious: If people only thought enough about the issues at hand, considered all the relevant information and employed proper reasoning strategies, their decision making would surely improve. This assumption is at the heart of numerous strategies that attempt to debias human judgment (for a review see Larrick, 2004); it is likewise central to public information campaigns designed to dispel erroneous beliefs and to replace them with more accurate information (for a review see Rice & Atkin, 2001). Unfortunately, these attempts to improve decision making often fail to achieve their goals, even under conditions assumed to foster rational judgment. 127 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39003-X
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
128
NORBERT SCHWARZ et al.
Models of rational choice assume that people will expend more time and eVort on getting it right when the stakes are high; hence, providing proper incentives should improve judgment. Empirically, it rarely does and Camerer and Hogarth (1999, p. 33) concluded after an extensive review that ‘‘there is no replicated study in which a theory of rational choice was rejected at low stakes in favor of a well‐specified behavioral alternative, and accepted at high stakes.’’ Similarly, increasing people’s accountability for their decisions improves performance in some cases, but impedes it in others (for a review see Lerner & Tetlock, 1999). One piece of this puzzle is that increased eVort will only improve performance when people already possess strategies that are appropriate for the task at hand; in the absence of such strategies, they will just do the wrong thing with more gusto (see Shafir & LeBoeuf, 2002). But even when no particularly sophisticated strategy is required, trying harder does not necessarily result in any improvement—in fact, it may often backfire. This is the case for one of the most widely recommended debiasing strategies: encouraging people to ‘‘consider the opposite,’’ or to counterargue their initial response, by asking themselves, ‘‘What are some reasons that my initial judgment might be wrong?’’ (Larrick, 2004, p. 323). Ironically, the more people try to consider the opposite, the more they often convince themselves that their initial judgment was right on target. The strategy of consider the opposite produces this unintended eVect because it ignores the second piece of the puzzle: the metacognitive experiences that accompany the reasoning process. Similar surprises arise in the domain of public information campaigns. Presumably, erroneous beliefs can be dispelled by confronting them with contradictory evidence. Yet attempts to do so often increase later acceptance of the erroneous beliefs, as known since Allport and Lepkin’s pioneering research (1945) into rumor transmission. Again, the unintended eVect arises because the educational strategy focuses solely on information content and ignores the metacognitive experiences that are part and parcel of the reasoning process. This chapter draws attention to the role of metacognitive experiences in judgment and decision making and explores their implications for debiasing strategies and public information campaigns. It is organized as follows. The first section introduces key metacognitive experiences and summarizes their principles of operation. The second section addresses a core assumption of most debiasing techniques: If people only thought enough about the right inputs, they would arrive at a less biased judgment. We identify the conditions under which this assumption holds as well as the conditions under which this strategy backfires. The third section addresses public information campaigns and illuminates why attempts to discredit erroneous beliefs often achieve the opposite. Throughout, we identify open issues for future research
SETTING PEOPLE STRAIGHT
129
and the chapter concludes with a discussion of theoretical and applied implications.
II. Metacognitive Experiences Most theories of judgment and decision making (for reviews see Koehler & Harvey, 2004; Lopes, 1994; Plous, 1993; Wyer & Srull, 1989) focus on the role of declarative information, that is on what people think about, and on the inference rules they apply to accessible thought content. Not surprisingly, theories of debiasing share this focus (for reviews see Arkes, 1991; FischhoV, 1982; Larrick, 2004; Wilson & Brekke, 1994). However, human reasoning is accompanied by a variety of metacognitive experiences, which provide experiential information that people systematically use in forming a judgment. These experiences qualify the implications of accessible declarative information, with the result that we can only accurately predict people’s judgments by taking the interplay of declarative and experiential information into account (for reviews see Schwarz, 1998, 2004; Schwarz & Clore, in press). Two of these experiences are of particular interest to this chapter, namely, the ease or diYculty with which information can be brought to mind and thoughts can be generated, and the fluency with which new information can be processed. A. ACCESSIBILITY EXPERIENCES Accessibility experiences refer to the ease or diYculty with which information can be recalled and thoughts can be generated. According to most models of judgment, we should evaluate an object more favorably when we bring many rather than few positive attributes to mind, should consider an event more likely when we generate many rather than few reasons for its occurrence, and so on. Empirically, this is often not the case (Sanna, Schwarz, & Stocker, 2002b; Wa¨nke, Bohner, & Jurkowitsch, 1997). Recalling many attributes or generating many reasons is more diYcult than recalling or generating only a few and these metacognitive accessibility experiences are informative in their own right. What people conclude from them depends on which of many naive theories of mental functioning they apply (Schwarz, 2004). The naive theory that is most relevant in the present context links recall experiences to characteristics of the external world. It holds that the more exemplars exist, the easier it is to bring some to mind. This correct belief is at the heart of Tversky and Kahneman’s availability heuristic (1973) and
130
NORBERT SCHWARZ et al.
people infer higher frequency and probability when examples are easy rather than diYcult to recall. Because frequent exemplars are also more typical for their category, ease of recall further suggests high typicality. Accordingly, people infer that they use their bicycles more often after recalling few rather than many instances (Aarts & Dijksterhuis, 1999), rate themselves as more assertive after recalling few rather than many of their own assertive behaviors (Schwarz et al., 1991), hold an attitude with more confidence after generating few rather than many supporting arguments (Haddock, Rothman, Reber, & Schwarz, 1999), and are more likely to choose a product after generating few rather than many reasons for this choice (Novemsky, Dhar, Schwarz, & Simonson, in press). Throughout, people’s inferences and decisions are consistent with the implications of what comes to mind when recall or thought generation is experienced as easy, but opposite to these implications when it is experienced as diYcult. In addition, people are more confident in their judgments when the relevant information was easy to bring to mind (Haddock et al., 1999; Tormala, Petty, & Brinol, 2002). From a content‐focused perspective, these findings may simply reflect that the examples become less compelling as people attempt to recall or generate an increasing number of them. Several lines of research converge on ruling out this possibility. First, when the informational value of experienced diYculty of recall or thought generation is undermined through misattribution manipulations, people draw on the recalled examples. For example, Schwarz et al. (1991, Experiment 3) told participants that unfamiliar music played in the background may interfere with the recall task. In this case, participants reported higher assertiveness the more examples of assertive behavior they brought to mind, reversing the otherwise observed pattern (see also Haddock et al., 1999; Sanna & Schwarz, 2003). This indicates that participants’ themselves found the examples compelling, unless their implications were qualified by an informative experience of diYculty. Second, yoked participants, who merely read the thoughts generated by another and are hence deprived of the generation experience, are more influenced when their partner lists many rather than few arguments, in contrast to the person who lists them (Wa¨nke, Bless, & Biller, 1996). This increasing influence on others’ judgments would not be observed if the quality of participants’ thoughts declined with the number of thoughts generated. Finally, the same influence of accessibility experiences can be observed when all participants list the same number of thoughts and their subjective experience of diYculty is manipulated through facial feedback in the form of corrugator contraction, an expression associated with mental eVort (Sanna, Schwarz, & Small, 2002a; Stepper & Strack, 1993). In combination, these findings highlight the role of subjective accessibility experiences and rule out alternative accounts in terms of the quality of the examples recalled or arguments generated.
SETTING PEOPLE STRAIGHT
131
As our review will indicate, experienced diYculty of thought generation can thwart the success of popular debiasing strategies, which encourage people to guard against overconfidence, hindsight bias, and similar fallacies by thinking about alternative possibilities. Such strategies only work when generating alternatives is experienced as easy, but backfire when it is experienced as diYcult. Variations in the perceived informational value of the experience and in the naive theory applied to the experience introduce additional complexities, as addressed in a later section.
B. PROCESSING FLUENCY Processing fluency refers to the ease or diYculty with which new, external information can be processed. Variables like figure–ground contrast, presentation duration, or the amount of previous exposure to the stimulus aVect the speed and accuracy of low‐level processes concerned with the identification of a stimulus’ physical identity and form; they influence perceptual fluency (Jacoby, Kelley, & Dywan, 1989). Variables like the consistency between the stimulus and its context or the availability of appropriate mental concepts for stimulus classification aVect the speed and accuracy of high‐level processes concerned with the identification of stimulus meaning and its relation to semantic knowledge structures; they influence conceptual fluency (Whittlesea, 1993). Empirically, both types of fluency show parallel influences on judgment (for a review see Winkielman, Schwarz, Fazendeiro, & Reber, 2003) and can be subsumed under the general term processing fluency. Because the diVerent variables that aVect processing fluency result in similar phenomenological experiences, the meaning of the experience is open to interpretation. Which interpretation people choose, and which inferences they draw, depends on the naive theory they bring to bear (Schwarz, 2004). Some naive theories pertain to presentation conditions; they give rise to illusions of perception. For example, people assume that material is easier to process when shown for long rather than short durations or with high rather than low clarity. Applying these theories, they infer that a stimulus was presented for a longer duration or with higher clarity when it is easy to process due to previous exposures (Jacoby, & Girard, 1990; Whittlesea, Witherspoon & Allan, 1985). Other theories pertain to one’s state of knowledge; they give rise to illusions of knowledge. People assume, for example, that familiar material is easier to process than unfamiliar material. Hence, they erroneously infer that they have seen a stimulus before when it is easy to process due to favorable presentation conditions (Whittlesea et al., 1990). For reasons discussed later, this fluency–familiarity link fosters the
132
NORBERT SCHWARZ et al.
acceptance of new information as true when it is easy to process, but flags it for closer scrutiny when it is diYcult to process. In addition, high fluency is experienced as hedonically positive (as captured by psychophysiological measures, Winkielman & Cacioppo, 2001). This aVective response can itself serve as a basis of judgment and fluently processed stimuli are evaluated more favorably than less fluently processed ones (for a review see Reber, Schwarz, & Winkielman, 2004). As is the case for other subjective experiences, the impact of fluency experiences is eliminated when people, correctly or incorrectly, attribute fluency to an irrelevant source, thus undermining its informational value for the judgment at hand (for a review see Kelley & Rhodes, 2002).
C. SUMMARY As this summary of core insights from research into the interplay of declarative and experiential information indicates, we cannot predict people’s judgments by knowing solely what comes to mind (for more extended reviews see Kelley & Rhodes, 2002; Schwarz, 2004; Schwarz & Clore, in press). By the same token, we cannot design strategies to improve people’s judgments by focusing solely on their use of declarative information. Instead, strategies of debiasing need to take the interplay of declarative and experiential information into account.
III. Accessibility Experiences and the Emergence and Attenuation of Bias According to content‐focused models of judgment, many biases arise because people focus narrowly on focal features of the issue and fail to consider a wider range of information. For example, people are overconfident about future success because they focus on behaviors that will lead to success and fail to consider variables that may impede success (Koriat, Lichtenstein, & FischhoV, 1980); once they learn about the outcome of an event, people assume that they knew all along that this would happen because they focus on outcome congruent knowledge and fail to consider variables that may have given rise to alternative outcomes (FischhoV, 1975); when predicting the time by which a task will be completed, people’s estimates are too optimistic because they focus on goal‐directed behaviors and fail to consider variables that may impede progress (Buehler, GriYn, & Ross, 1994). The strength of these biases is assumed to increase with the number of focal
SETTING PEOPLE STRAIGHT
133
thoughts and to decrease with the number of alternatives considered. From this perspective, any strategy that succeeds in encouraging people to consider information about alternatives should attenuate the respective bias (Larrick, 2004). Research into the interplay of accessible declarative information and subjective accessibility experiences leads to diVerent predictions regarding the role of focal and alternative thoughts in the emergence as well as attenuation of bias. It predicts (1) that the focal thoughts that are assumed to give rise to bias will only do so when they come to mind easily. (2) When focal thoughts are diYcult to bring to mind, the otherwise observed bias is attenuated or eliminated (and sometimes even reversed) despite the numerous ‘‘biasing’’ focal thoughts generated. These contingencies have been overlooked in previous research because people are likely to truncate the search process early (see Bodenhausen & Wyer, 1987), before any diYculties of recall and thought generation are likely to be experienced. Hence, their judgments are usually consistent with spontaneously generated focal thoughts, which led to the erroneous conclusion that we can predict people’s judgments on the basis of thought content alone. Empirically this is not the case (Schwarz et al., 1991), as reviewed above. The same logic applies to thoughts about alternatives. (3) Thinking about alternatives will only attenuate bias when alternatives come to mind easily. Hence, (4) trying to think of many alternatives will increase rather than decrease bias, in contrast to the predictions of content‐focused models. Ironically, this suggests that successful debiasing may become the less likely the harder people try to avoid bias: the more they search for information that may argue against their initial judgment, the harder they will find their task, convincing them that their initial judgment was indeed right on target. Finally, subjective accessibility experiences are only used as a source of information when their informational value is not discredited (for reviews see Schwarz, 2004; Schwarz & Clore, in press). When people attribute the experienced ease or diYculty to an irrelevant source, they turn to accessible thought content instead. In this case, the judgment is solely based on accessible declarative information and (5) bias increases with the number of focal thoughts and (6) decreases with the number of alternative thoughts, as predicted by content‐focused models. Figure 1 summarizes this logic; it is a generalized version of a metacognitive model of hindsight bias proposed by Sanna and Schwarz (2006, in press). We assume that judgments are always a joint function of thought content (accessible declarative information) and accompanying metacognitive experiences. When the informational value of the metacognitive experience is discredited, judgments are based on declarative information and hence consistent with thought content (lower right‐hand oval). When the informational value of
134
NORBERT SCHWARZ et al.
Thought content (e.g., Supportive or contrary to bias) + Metacognitive experiences (e.g., Emotional reactions, conceptual or perceptual fluency, ease of recall, or thought generation) Naive therories (e.g., Implicit beliefs about self or external world) Informational value (e.g., Applicability to judgment, discounting, or augmenting) Informative
Metacognitive experiences qualify thought content (e.g., Judgments consistent with content when fluent or easily generated but inconsistent with content when disfluent or difficultly generated)
Uniformative
Metacognitive experiences uninformative and judgments consistent with thought content
Fig. 1. The interplay of thought content and metacognitive experiences.
the experience is not discredited (lower left‐hand oval), the metacognitive experience qualifies the implications of declarative information. What exactly people conclude from the experience depends on the nature of the experience (ease or diYculty of thought generation, fluency of processing, and so on) and the naive theory of mental processes applied. In the case of accessibility experiences, judgments are consistent with the implications of declarative information when it was easy to bring to mind. However, the content congruent judgments at which people arrive under this condition are likely to be more extreme than is the case when the metacognitive experience is uninformative (lower right‐hand oval): When experienced ease of recall or thought generation is discredited, two arguments are just two arguments; when it is not discredited, the experienced ease suggests that there are many more arguments where those two came from. Finally, judgments are opposite to the implications of accessible declarative information when it was diYcult to bring to mind. In sum, this model converges with content‐based models by predicting content congruent judgments (1) when the accessibility experience is discredited or (2) when recall or thought generation is experienced as easy.
SETTING PEOPLE STRAIGHT
135
It diVers from content‐based models by predicting (3) easily recalled or generated thoughts are more influential when the experience is considered informative than when it is not. Finally, and most important, it (4) arrives at predictions that are opposite to the predictions of content‐based models when recall or thought generation is diYcult.
A. THE INTERPLAY OF DECLARATIVE AND EXPERIENTIAL INFORMATION Consistent support for this model comes from a study that took advantage of a real‐world event, namely students’ first exam in a psychology class (Sanna & Schwarz, 2004). In a between‐subjects design, students made a variety of judgments either 28 days or a few minutes before their exam, or right after they received their grades. In addition, they listed either 0, 3, or 12 thoughts about succeeding or failing on the exam, as described below. Manipulation checks indicated that listing 3 thoughts was experienced as easy, whereas listing 12 thoughts was experienced as diYcult, irrespective of whether thoughts were about success or failure. We focus on this study because it bears on a number of diVerent biases and note additional experimental work along the way. 1. Confidence Changes Previous research showed that people are overconfident about their future success at a distance, but become more realistic as the moment of truth approaches. For example, students taking an immediate test predict a lower likelihood of success than those taking the test in 4 weeks (Nisan, 1972); similarly, college seniors provide more muted estimates of first‐job salaries than sophomores and juniors (Shepperd, Ouellette, & Fernandez, 1996). These diVerences presumably reflect a focus on success‐related thoughts at a temporal distance, which gives way to worries and awareness that one may not be as well prepared as one hoped as the moment of truth comes closer. The left‐hand panel of Fig. 2 supports this reasoning. Twenty‐eight days before the exam, students who listed no exam‐related thoughts were as optimistic as students who had just listed three reasons for success. Conversely, 5 minutes before the exam, students who listed no thoughts were as pessimistic as students who had just listed 3 reasons for why they may fail. Thus, spontaneous thoughts about success and failure seem to account for distal confidence and proximal pessimism, respectively—but only when these thoughts are easy to generate. When students had to list 12 thoughts, which they found diYcult, this pattern reversed. In this case, listing success‐related
136
NORBERT SCHWARZ et al.
Likelihood of success
10
9
Control 3-Success 12-Success 3-Failure 12-Failure
8
7
6 28 days
5 min
Prior to exam
Success
Failure
After exam
Fig. 2. Likelihood‐of‐success ratings by thought‐listing and time: Confidence changes and hindsight bias. Shown are the mean ratings of the likelihood of success (10 ¼ very likely) as a function of thought‐listing condition and time. The thought‐listing conditions were as follows: control, no thoughts listing; 3‐success, generate 3‐success thoughts; 3‐failure, generate 3‐failure thoughts; 12‐success, generate 12‐success thoughts; 12‐failure, generate 12‐failure thoughts. Thoughts were listed either 28 days or approximately 5 minutes prior to the exam, or right after learning grades. Participants in the latter condition were grouped according to whether they viewed their grade as a success or failure. Adapted from Sanna and Schwarz (2004), ß 2004 American Psychological Society.
thoughts undermined their confidence long before the exam, whereas listing failure‐related thoughts boosted their confidence right before the exam. In combination, this pattern of findings indicates that confidence changes over time are a joint function of thought content and the ease with which these thoughts can be brought to mind. This pattern bears directly on the success and failure of debiasing strategies. Presumably, overconfidence long before the exam can be reduced by prompting participants to consider reasons why the exam may not go well, whereas their pessimism right before the exam can be alleviated by prompting them to consider reasons for success. As the left‐hand panel of Fig. 2 shows, this is the case—but only when the respective thoughts come to mind easily. Listing 3 failure‐related thoughts 28 days before the exam reduced confidence to the level of control participants’ pessimism 5 minutes before the exam, whereas listing 12 failure‐related thoughts failed to do so. Conversely, listing 3 success‐related thoughts 5 minutes prior to the exam brought confidence back to its distant level, yet 12 failure‐related thoughts failed to do so.
SETTING PEOPLE STRAIGHT
137
Throughout, listing few success (failure)‐related thoughts was functionally equivalent to listing many failure (success)‐related thoughts. Confidence shifts over time have been observed in many diVerent domains, including the forecasts of professional market analysts (for a review see Kadous, Krische, & Sedor, 2006). Paralleling the temporal shifts discussed above, the optimism of market analysts’ earnings forecasts is a function of their time horizon: ‘‘analysts’ two‐year‐ahead forecasts are relatively more optimistic than their 1‐year‐ahead forecasts’’ and the ‘‘optimism in quarterly earnings forecasts decreases as the earnings announcement date approaches’’ (Kadous et al., 2006, p. 380). To reduce the optimism of analysts’ forecasts, analysts are often encouraged to consider ways in which the scenarios presented by a company’s management may fail (Heiman, 1990; Koonce, 1992). Kadous et al. (2006) found that the success of this strategy depends on analysts’ ease of thought generation. They presented professional analysts with scenarios that outlined a company’s plans and asked them to generate 0, 2 (easy), or 12 (diYcult) ways in which the plan may fail. Replicating the above findings, this strategy reduced analysts’ optimism when thought generation was easy, but not when it was diYcult. This finding also highlights that the influence of subjective accessibility experiences is not limited to lay judgment; it can also be observed with professional analysts, whose forecasts move markets and whose optimism hurts investors (Dechow, Hutton, & Sloan, 2000). 2. Hindsight Bias Once people know the outcome of an event, they believe that it was relatively inevitable and that they ‘‘knew all along’’ what would happen. This hindsight bias (FischhoV, 1975) has been observed across many content domains (for reviews see Christensen‐Szalanski & Willham, 1991; Hawkins & Hastie, 1990). The usually recommended remedy is to think about alternatives to the known outcome in an attempt ‘‘to convince oneself that it might have turned out otherwise’’ (FischhoV, 1982, p. 343). To assess hindsight bias, Sanna and Schwarz (2004) asked students right after they received their grades to report which likelihood of success they would have predicted 28 days earlier, at the beginning of class. The right‐ hand panel of Fig. 2 shows the results. Replicating numerous earlier studies, students who listed no thoughts assumed that they knew the outcome all along and provided higher retrospective ratings of success when they succeeded (after exam/success) rather than failed (after exam/failure). Moreover, their retrospective estimates were more extreme than the ratings provided by their peers 28 days earlier, even compared to peers who had just generated three reasons why they might succeed or fail at that time.
138
NORBERT SCHWARZ et al.
More important, asking successful students to think of three reasons why they might have failed eliminated their hindsight bias; so did asking failing students to list three reasons why they might have succeeded. Yet, generating 12 thoughts about alternative outcomes did not further attenuate hindsight bias, in contrast to what content‐focused theorizing would predict (FischhoV, 1982)—to the contrary, it undermined the eVectiveness of this debiasing strategy. Finally, generating many thoughts in support of the obtained outcome attenuated hindsight, again in contrast to what content‐ focused theorizing would predict: Successful students who were asked to list 12 reasons for why they succeeded concluded that they would not have expected their success; neither did failing students think they would have expected their failure after listing 12 reasons for failing. In sum, hindsight bias was successfully reduced when participants found it easy to think of (a few) reasons for alternative outcomes or found it diYcult to think of (many) reasons for the obtained outcome. Other studies reiterate this point. Sanna et al. (2002b) presented participants with a story about the British– Gurkha war (taken from FischhoV, 1975), which informed them that either the British or the Gurkha had won. Next, they listed few or many reasons for how the war could have come out otherwise. Generating few reasons for the alternative outcome attenuated hindsight bias, whereas generating many reasons backfired and increased hindsight bias above the otherwise observed level. The same pattern was observed when participants’ accessibility experiences were manipulated through facial feedback (Sanna et al., 2002a). Specifically, all participants listed five thoughts about how the war could have come out otherwise and some were induced to contract the currugator muscle while doing so, which conveys a feeling of mental eVort (Stepper & Strack, 1993). As predicted, listing five thoughts about alternative outcomes while tensing the currogator muscle increased hindsight bias, resulting in a significant backfire eVect. Finally, listing many thoughts about why the war had to turn out the way it did, or tensing the currogator muscle while doing so, attenuated hindsight bias (Sanna et al., 2002b), again paralleling the above findings (Sanna & Schwarz, 2004). Throughout, these results highlight that generating many focal thoughts is functionally equivalent to generating few alternative thoughts, whereas generating many alternative thoughts is functionally equivalent to generating few focal thoughts. Any attempt to debias hindsight needs to take this interplay of declarative and experiential information into account. Because previous theorizing has not done so, it is not surprising that a comprehensive meta‐analysis concluded that ‘‘eVorts to reduce hindsight bias have generally been unsuccessful’’ (Guilbault, Bryant, Brockway, & Posavac, 2004, p. 112). Our findings suggest that a clearer pattern is likely to emerge once we take participants’ accessibility experiences into account; unfortunately, virtually
SETTING PEOPLE STRAIGHT
139
all prior studies fail to report the details needed to consider this variable in a meta‐analysis. 3. Planning Fallacy As a final example, consider the planning fallacy (Buehler et al., 1994; Kahneman & Tversky, 1979). At a temporal distance, people usually predict that task completion will need less time than is actually the case. This bias has been observed in diverse settings ranging from household chores to school assignments. Incentives worsen the planning fallacy, as people expecting tax refunds or other monetary rewards for speedy completion are even more optimistic than those who have no incentive for speedy completion (Buehler, GriYn, & MacDonald, 1997; Buehler et al., 1994). Again, this fallacy presumably reflects a focus on acts that facilitate task completion at the expense of hurdles that impair it. Accordingly, it should be attenuated when attention is drawn to reasons for slow progress (Newby‐Clark, Ross, Buehler, Koehler, & GriYn, 2000). Twenty‐eight days before the exam, control participants predicted that they would complete their exam preparation 4.4 days prior to the exam, although their peers who responded to the questionnaire 5 minutes prior to the exam reported that were not done until 0.3 days prior to the exam (Fig. 3). This replicates the usual planning fallacy. Again, control participants’ early optimism matched the optimism of those who generated three success‐related thoughts well before the exam, whereas those who generated
Reported study completion (days)
5 Control 3-Success 12-Success 3-Failure 12-Failure
4 3 2 1 0 28 days
5 min
Time prior to exam
Fig. 3. Predicted and actual study completion time by thought‐listing and time: Planning fallacy. Shown are the predicted (reported 28 days prior to the exam) and actual (reported 5 minutes prior to the exam) completion times in days. Adapted from Sanna and Schwarz (2004), ß 2004 American Psychological Society.
140
NORBERT SCHWARZ et al.
three failure‐related thoughts provided more realistic estimates. More important, listing 12 failure‐related thoughts failed to attenuate the planning fallacy, whereas listing 12 success‐related thoughts, ironically, did attenuate it. Once again, listing few success (failure)‐related thoughts, an easy task, was functionally equivalent to listing many failure (success)‐related thoughts, a diYcult task.1 4. Conclusions In sum, these findings illustrate that we can only predict the emergence and attenuation of bias by paying attention to the joint influence of accessible declarative information and subjective accessibility experiences. First, focal thoughts give rise to bias when they are easy to bring to mind. This is usually the case when people are left to their own devices as they truncate the search process early (Bodenhausen & Wyer, 1987), before any diYculty is experienced. Accordingly, the judgments of control participants, who did not list any thoughts, converged with the judgments of participants who listed three focal thoughts (see Figs. 2 and 3). Second, focal thoughts attenuate or eliminate bias when they are diYcult to bring to mind. Hence, bias is more likely to arise when people generate few rather than many focal thoughts, in contrast to what content‐focused models would predict. Conversely, third, thoughts about alternatives attenuate or eliminate bias when they come to mind easily, but, fourth, increase bias when they are diYcult to bring to mind. As a result, the frequently recommended debiasing strategy of ‘‘consider the opposite!’’ is only successful when people do not try too hard to follow it—and backfires when people are overly zealous in their eVorts to protect themselves against bias. We propose that this previously overlooked contingency accounts for the mixed success of the consider‐the‐opposite strategy (see Arkes, 1991; Guilbault et al., 2004; Hawkins & Hastie, 1990).
B. DISCREDITING THE INFORMATIONAL VALUE OF THE EXPERIENCE Theoretically, the observed interaction of declarative information and subjective accessibility experiences should only be observed when the informational value of the accessibility experience is not called into question (see Fig. 1). When it is called into question, people discount the accessibility experience as a source of information for the judgment at hand and rely on 1 No impact of the thought manipulation is observed 5 minutes prior to the exam, reflecting that actual study completion times were known at that point.
SETTING PEOPLE STRAIGHT
141
thought content to arrive at a judgment (Schwarz et al., 1991). Unfortunately, this does not imply that debiasing is necessarily successful when people do not take their metacognitive experiences into account. 1. Could Gore Have Won? Taking advantage of the 2000 presidential election in the United States, Sanna and Schwarz (2003) asked participants to predict the outcome of the popular vote 1 day prior to the November 7 election. Following an extended court battle over disputed election outcomes in Florida, the Democratic candidate Gore conceded the election on December 13, 2000. On December 14, participants were asked to recall their preelection predictions made on November 6. The intervening 5 weeks were filled with extensive media coverage of legal and political debates ranging from voting irregularities in Florida to the role of state legislatures in national elections and the responsibilities of the Supreme Court. Table I shows the results. In the popular vote, Gore–Lieberman led over Bush–Cheney by a small diVerence of 0.32%. Prior to the election, participants predicted a clear victory for Gore, with a lead of close to 5%. After the election, participants who were merely asked to recall their preelection prediction recalled that they did predict a Gore win, but at a much smaller margin of 0.58%. This hindsight bias is consistent with previous observations in election studies (Leary, 1982; Powell, 1988). Asking participants to list 12 thoughts about how Gore could have won before they recalled their earlier predictions did not attenuate hindsight bias. Nor did listing 12 thoughts, which participants experienced as diYcult, result in a backfire eVect, in contrast to earlier studies (Sanna & Schwarz, 2004; Sanna et al., 2002a,b). This probably reflects that a backfire eVect would have implied a qualitative change in form of recalling that one predicted a Bush victory, which imposes some constraints. More important, the experienced diYculty of listing 12 thoughts in favor of a Gore victory was discredited for a third group of participants. Right after listing their thoughts, they were asked to rate how knowledgeable they are about politics. The instructions read, ‘‘Thank you for listing those thoughts. We realize this was an extremely diYcult task that only people with a good knowledge of politics may be able to complete. As background information, may we therefore ask you how knowledgeable you are about politics?’’ This manipulation suggests that any experienced diYculty may not imply that there were no ways for Gore to win—it may merely imply that one’s own knowledge is not suYcient to come up with them. Hence, these participants should discount their diYculty experience and draw on the thoughts they generated. The results indicate that they did: These participants recalled that they predicted a large victory for Gore, with
142
NORBERT SCHWARZ et al.
TABLE I ACTUAL ELECTION OUTCOME, MEAN PREDICTED OUTCOME, AND MEAN RECALLED PREDICTIONS BY THOUGHT‐LISTING AND ATTRIBUTION Actual outcome Gore–Lieberman
48.31
Bush–Cheney
47.99
DiVerence
0.32 Participants’ estimates Preelection prediction
Postelection recall
Gore–Lieberman
49.54
48.22
Bush–Cheney
45.08
47.54
4.45
0.58
Gore–Lieberman
49.69
48.80
Bush–Cheney
44.42
48.19
5.26
0.61
No Thoughts‐Listing
DiVerence 12‐Thoughts
DiVerence
12‐Thoughts, Attribution (Experience Uninformative) Gore–Lieberman
49.85
52.09
Bush–Cheney
45.14
44.57
4.71
7.52
DiVerence
Note: Values are in predicted percentages of popular vote and recalled predictions. DiVerence ¼ Gore–Lieberman – Bush–Cheney. Adapted from Sanna and Schwarz (2003, Experiment 2), ß 2002 Elsevier Science.
a margin of 7.52%—in fact, significantly larger than the 4.71% margin they had predicted prior to the election. In sum, once the outcome of the popular vote was known, it seemed inevitable and participants erroneously ‘‘recalled’’ that they had predicted it all along. This hindsight bias was not attenuated by generating 12 thoughts about alternative outcomes. More important, inducing participants to attribute the experienced diYculty to their own lack of knowledge undermined its informational value. As observed in earlier studies (Schwarz et al., 1991), these participants drew on the pro‐Gore thoughts they had just generated and inferred that they must have predicted a large victory for Gore. Far from improving the accuracy of recall, this successful elimination of hindsight bias came at the price of a significant bias in the opposite direction. A parallel study, using the outcome of football games as the content domain, produced comparable results (see Sanna & Schwarz, 2003).
SETTING PEOPLE STRAIGHT
143
2. Conclusions In combination, the reviewed research highlights the crucial role of subjective accessibility experiences in the production and reduction of judgmental biases. Thoughts about focal aspects of a given issue produce bias when they are easy to bring to mind, but reduce bias when they are diYcult to bring to mind; conversely, thoughts about alternatives reduce bias when they are easy to bring to mind, but produce bias when they are diYcult to bring to mind. Hence, encouraging people to ‘‘consider the opposite’’ can be a successful debiasing strategy when consideration of the opposite is experienced as easy; the strategy backfires when consideration of the opposite is diYcult—and even furrowing one’s brow can be enough to produce a backfire eVect. Finally, encouraging people to generate many focal thoughts can, paradoxically, be a successful debiasing strategy, provided that thought generation is diYcult. The latter strategy has not been examined in previous research and its potential deserves future attention. Overall, this systematic pattern of findings cannot be derived from models that focus exclusively on the role of declarative information.
IV. Fluency, Familiarity, and Truth: Implications for Public Information Campaigns So far, we considered metacognitive experiences that arise when people elaborate on an issue. We now turn to a metacognitive experience that accompanies exposure to new information, namely the fluency with which the information can be processed. As discussed in Section II, people correctly assume that familiar information is easier to process than novel information. Applying this naive theory of mental processes, they infer from experienced processing fluency that the information is familiar—even when the fluency derives from presentation variables, like good figure–ground contrast or long exposure times, or from contextual influences, like preceding semantic primes (for reviews see Kelley & Rhodes, 2002; Winkielman et al., 2003). In memory experiments, this fluency–familiarity link gives rise to the erroneous ‘‘recognition’’ of fluently processed but previously unseen stimuli (see Kelley & Rhodes, 2002). More important for the present purposes, the perceived familiarity of information influences the likelihood that the information is accepted as true or flagged for closer scrutiny. In terms of the model presented in Fig. 1, judgments are likely to be consistent with the implications of fluently processed declarative information, but inconsistent with the implications of disfluently processed declarative information. We review selected results to
144
NORBERT SCHWARZ et al.
illustrate these eVects and subsequently turn to their implications for public information campaigns. A. FLUENCY, FAMILIARITY, AND BIAS 1. Fluency and Hindsight Bias If fluently processed information seems more familiar, we may expect that fluently processed information about an outcome gives rise to more pronounced hindsight bias—after all, if it seems familiar, we probably ‘‘knew it all along.’’ Empirically, this is the case. Werth and Strack (2003) gave participants general knowledge questions (e.g., ‘‘How high is the EiVel tower?’’) along with answers (300 m). Next, they asked participants what they would have answered, had they not been given solutions. To manipulate processing fluency, questions and answers were presented in colors that were either easy or diYcult to read against the background. As expected, high processing fluency increased the size of hindsight bias, and participants’ confidence in their answers, whereas low processing fluency attenuated hindsight bias. Presumably, participants found the easy‐to‐read material more familiar and hence concluded that they knew this information all along and ‘‘would have’’ provided the correct answer. 2. Fluency and the Detection of Distorted Questions Conversely, Song and Schwarz (2007) observed that low processing fluency flags material for closer scrutiny. When asked, ‘‘How many animals of each kind did Moses take on the Ark?’’ most people responded ‘‘two’’ despite knowing that Noah rather than Moses was the actor in the biblical story (Erickson & Mattson, 1981). Previous research addressed a variety of plausible accounts for this ‘‘Moses illusion’’ (for a comprehensive review see Park & Reder, 2003), including the possibility that recipients are cooperative communicators (Grice, 1975; Schwarz, 1994, 1996) who notice the distortion, but simply correct for it by responding to what the questioner ‘‘must have meant.’’ Yet making participants aware that the text may be distorted, or asking them to identify such distortions, does not eliminate the eVect (Bredart & Modolo, 1988; Reder & Kusbit, 1991), in contrast to what a Gricean account would predict. The account that is most compatible with the available data holds that ‘‘distortion detection involves a two‐pass process—the first to flag a potential mismatch and the second to invoke a careful inspection that might confirm an erroneous term in the question’’ (Park & Reder, 2003, p. 282). Distorted questions pass the first stage when the semantic overlap between the question and the person’s knowledge
SETTING PEOPLE STRAIGHT
145
provides a suYcient match (e.g., Moses and Noah are both characters in the Old Testament, who received commands from God that were related to water), but not when the semantic overlap is low (e.g., when Moses is replaced by ‘‘Nixon’’ in the above question; see Erickson & Mattson, 1981; van Oostendorp & de Mul, 1990). From the present perspective, semantic overlap is only one of many variables that influences whether the content of the question is likely to seem familiar. If so, any other manipulation that increases perceived familiarity should also aVect the size of the Moses illusion, including manipulations of processing fluency. To test this prediction, Song and Schwarz (2007) presented the Moses question and related ones in either an easy‐to‐read or diYcult‐to‐read print font. As expected, many more participants answered ‘‘two’’ in response to the Moses question when the font was easy to read (15 of 17) than when it was diYcult to read (8 of 15). Other questions replicated this eVect, indicating that low processing fluency flags material for closer scrutiny because it seems less familiar, much like high processing fluency conveys that one knew it all along (Werth & Strack, 2003).
B. FLUENCY, CONSENSUS, AND TRUTH When the objective state of aVairs is diYcult to determine, people often resort to social consensus information to judge the truth value of a belief: if many people believe it, there’s probably something to it (Festinger, 1954). Because one is more frequently exposed to widely shared beliefs than to highly idiosyncratic ones, the familiarity of a belief is often a valid indicator of social consensus. But, unfortunately, information can seem familiar for the wrong reason, leading to erroneous perceptions of social consensus. For example, Weaver, Garcia, Schwarz, and Miller (in press) exposed participants to multiple iterations of the same statement, provided by the same communicator. When later asked to estimate how widely the conveyed belief is shared, participants estimated higher consensus the more often they had read the identical statement coming from the same single source. Findings of this type indicate that repeated exposure to a statement influences perceptions of social consensus, presumably because the statement seems more familiar. This inferred consensus contributes to the observation that repeated exposure to a statement increases its acceptance as true. In a classic study of rumor transmission, Allport and Lepkin (1945) observed that the strongest predictor of belief in wartime rumors was simple repetition. Numerous subsequent studies confirmed this conclusion and demonstrated that a given statement is more likely to be judged ‘‘true’’ the more often it is repeated. This illusion of truth eVect has been obtained with trivia statements and words from a foreign
146
NORBERT SCHWARZ et al.
language (Begg, Anas, & Farinacci, 1992; Hasher, Goldstein, & Toppino, 1977) as well as advertising materials (Hawkins & Hoch, 1992). Theoretically, any other variable that increases processing fluency should have the same eVect as message repetition. Supporting this prediction, Reber and Schwarz (1999) found that participants were more likely to accept statements like ‘‘Osorno is a city in Chile’’ as true when the statements were presented in colors that made them easy (e.g., dark blue) rather than diYcult (e.g., light blue) to read against the background. Similarly, McGlone and Tofighbakhsh (2000) manipulated processing fluency by presenting substantively equivalent novel aphorisms in a rhyming (e.g., ‘‘woes unite foes’’) or nonrhyming form (e.g., ‘‘woes unite enemies’’). As expected, participants judged substantively equivalent aphorisms as more true when they rhymed than when they did not. In combination, these findings indicate that processing fluency serves as an experiential basis of truth judgments. In the absence of more diagnostic information, people draw on the apparent familiarity of the statement to infer its likely truth value—if it seems they heard it before, there is probably something to it (Festinger, 1954). This inference involves the (over)application of the correct naive theory that ‘‘familiar material is easy to process’’—and the application of any other theory [like ‘‘Good figure–ground contrast makes things easy to read’’ in the Reber and Schwarz (1999) study] would presumably render the fluency experience uninformative for truth judgments.
C. IMPLICATIONS FOR PUBLIC INFORMATION CAMPAIGNS The fluency–familiarity–truth link suggested by the above studies has important implications for the design of public information campaigns, including some obvious ones and some rather counterintuitive ones. On the obvious side, it suggests that frequent repetition of the information that one wants to convey to the public is a good idea. All information campaigns attempt to do so, although usually based on the assumption that frequent exposure facilitates successful learning and message retention (McQuail, 2000; Rice & Atkin, 2001; Tellis, 2004). From a metacognitive perspective, frequent exposure also facilitates increasingly fluent processing of the message and increased perceptions of familiarity, which, in turn, increase the likelihood of message acceptance. Rhyming slogans and presentation formats that facilitate fluent processing will further enhance this eVect. On the counterintuitive side, this logic implies that false information is better left alone. Any attempt to explicitly discredit false information necessarily involves a repetition of the false information, which may contribute to its later
SETTING PEOPLE STRAIGHT
147
familiarity and acceptance. Although this problem has been known since Allport and Lepkin’s research (1945) into wartime rumors, the idea that false information needs to be confronted is so appealing that it is still at the heart of many information campaigns. Like the debiasing strategy of consider‐the‐ opposite, it derives its appeal from the assumption that judgments are based on declarative information—and it fails because it underestimates the power of metacognitive experiences. 1. Spreading Myths by Debunking Them Figure 4 shows a flyer published by the Centers for Disease Control (CDC). The flyer is available online, provided under materials that physicians can download to educate patients. It illustrates a common format of information campaigns that counter misleading information by confronting ‘‘myths’’ with ‘‘facts.’’ In this case, the myths are erroneous beliefs about flu vaccination (e.g., ‘‘The side eVects are worse than the flu’’), which are confronted with a number of facts. From the perspective of content‐focused models of judgment, the facts present strong arguments and should decrease the acceptance of the myths. Yet the flyer repeats the myths, which may contribute to their fluency and perceived familiarity when they are encountered again, possibly increasing rather than decreasing their later acceptance. Skurnik, Yoon, and Schwarz (2007) tested this possibility by giving participants the CDC’s ‘‘Facts & Myths’’ flyer shown in Fig. 4 or a parallel ‘‘Facts’’ version that presented only the facts. Of interest was how these flyers aVect participants’ beliefs about the flu and their intention to receive flu vaccination. These measures were assessed either immediately after participants read the respective flyer or 30 minutes later. Participants who read the ‘‘Facts & Myths’’ flyer received a list of statements that repeated the facts and myths and indicated for each statement whether it was true or false. Right after reading the flyer, participants had good memory for the presented information and made only a few random errors, identifying 4% of the myths as true and 3% of the facts as false. Thirty minutes later, however, their judgments showed a systematic error pattern: They now misidentified 15% of the myths as true, whereas their misidentification of facts as false remained at 2%. This is the familiar pattern of illusion‐of‐truth eVects: Once memory for substantive details fades, familiar statements are more likely to be accepted as true than to be rejected as false. This familiarity bias results in a higher rate of erroneous judgments when the statement is false rather than true, as observed in the present study. On the applied side, these findings illustrate how the attempt to debunk myths facilitates their acceptance after a delay of only 30 minutes.
148
NORBERT SCHWARZ et al.
Fig. 4. Flyer distributed by the Centers for Disease Control (http://www.cdc.gov/flu/professionals/flugallery/flyers.htm; accessed 28 Dec 2006).
Participants’ attitude judgments and behavioral intentions showed a parallel pattern. Right after reading the flyer, flyers both improved participants’ attitudes toward flu vaccination and increased their intention to get vaccinated, relative to participants who had not read a flyer. The same positive
SETTING PEOPLE STRAIGHT
149
influence was observed after a 30‐minute delay for participants who had only read the facts. In contrast, the ‘‘Facts & Myths’’ flyer backfired after a delay: these participants reported less favorable attitudes toward flue vaccination and lower behavioral intentions than control participants who read no flyer at all. In sum, providing participants only with the facts had the intended eVects on participants’ attitudes and intentions, both immediately and after a short delay. The classic ‘‘Facts & Myths’’ format, on the other hand, was only eVective immediately and backfired after a short delay. A mere 30 minutes later, the ‘‘Facts & Myth’’ flyer facilitated acceptance of myths as facts, impaired participants’ attitudes toward vaccination, and undermined their vaccination intentions, relative to controls who read no flyer at all. In combination, these findings suggest that participants drew on the declarative information provided by the flyers when it was highly accessible (no‐delay condition). As this information faded from memory, they increasingly relied on the perceived familiarity of the information to determine its truth value, resulting in the observed backfire eVects. 2. Turning Warnings into Recommendations If confronting a given myth with several facts backfires, we may expect that backfire eVects are even more pronounced when misleading claims are merely identified as false. Moreover, the observed delay eVects suggest that older adults may be particularly vulnerable to the backfire eVects of information campaigns. Numerous studies indicate that explicit memory declines with age, whereas implicit memory stays largely intact [see Park (2000) for a review]. If so, older adults may be unlikely to remember the details of previously seen information (a function of explicit memory), but may still find previously seen statements more easy to process and may experience them as familiar (a function of implicit memory). Skurnik, Yoon, Park, and Schwarz (2005) tested this possibility in the context of health product claims. They exposed older and younger adults once or thrice to statements like, ‘‘Shark cartilage is good for your arthritis;’’ each statement was explicitly marked as ‘‘true’’ or ‘‘false.’’ Either immediately after the learning phase or after a 3‐day delay, participants were shown the statements again and were asked to identify each one as true or false. Figure 5 shows the results. Without a delay (top panel), young participants were equally likely to misidentify a true statement as false and a false statement as true, indicating some random error. Older participants, on the other hand, were more likely to misidentify a false statement as true than a true statement as false. Not surprisingly, this illusion‐of‐truth eVect was more pronounced after a single exposure than after three exposures,
150
NORBERT SCHWARZ et al.
0.50
Mean proportions of responses
0.45
No-delay condition
“True” to false “False” to true
0.40 0.35
0.28
0.30 0.25
0.17
0.20 0.15
0.12
0.12
0.10
0.10
0.07
0.10
0.04
0.05 0.00 1 Presentation 3 Presentations 1 Presentation 3 Presentations Younger adults
Older adults
0.50
Mean proportions of responses
0.45
3-days delay condition
“True” to false “False” to true
0.40
0.40 0.28
0.35 0.24 0.30
0.19
0.25 0.17 0.20
0.16 0.14 0.10
0.15 0.10 0.05 0.00 1 Presentation
3 Presentations 1 Presentation 3 Presentations
Younger adults
Older adults
Fig. 5. Truth judgments for true and false statements as a function of repetition, age, and delay. Shown is the proportion of false statements identified as true (‘‘true’’ to false) and true statements identified as false (‘‘false’’ to true) after no delay (top panel) and a delay of 3 days (bottom panel). Adapted from Skurnik, Yoon, Park, and Schwarz (2005), ß 2005 Journal of Consumer Research, Inc.
SETTING PEOPLE STRAIGHT
151
indicating that three exposures resulted in better memory for the presented information. After a 3‐day delay (bottom panel), young participants in the single exposure condition showed a sizeable illusion of truth eVect and misidentified 24% of the false statements as true; this eVect was less pronounced after three exposures, which improved memory for the details. Finally, older participants misidentified 29% of the false statements as true after a single exposure and this error increased to a full 40% when they had been told three times that the statements are false. Note that the overall pattern shows an increasing reliance on familiarity as memory fades. Without a delay, young participants remember much of the presented information and make some random errors; older participants’ memory is less good and they resort to familiarity as a basis for truth judgments, resulting in a bias to accept false statements as true that exceeds the error of rejecting true statements as false. Over the course of a 3‐day delay, younger participants’ memory for details fades as well, putting them in the position that older participants experienced without a delay. Finally, older participants are particularly likely to accept false statements as true after three warnings, suggesting that repeated warnings primarily increase the perceived familiarity of the statements for this age group. As a result, repeating false information as part of a public information campaign may put older adults at a particular risk, essentially turning warnings into recommendations. 3. Lending Credibility to Myths As these studies illustrate, attempts to inform people that a given claim is false may increase acceptance of the misleading claim. In addition, such attempts may also have the unintended eVect that the false claim is eventually associated with a highly credible source. Because messages from high credibility sources are more influential, as known since Hovland and Weiss (1951), this will further enhance the acceptance of the false claim—including its acceptance by others, who are told that one learned it from a credible source. Two diVerent processes are likely to contribute to this. First, consider participants who read the CDC’s ‘‘Facts & Myths’’ flyer in Skurnik et al.’s study (2007). As the results indicate, some of the participants came to accept the myths as true. After some delay, these participants may recall (and report to others) that they learned this information from a truly credible source—the CDC. Such cases of source confusion have been repeatedly observed in rumor transmission. For example, Emery (2000) reported a case where an Internet rumor about flesh‐eating bananas became attributed to increasingly more credible sources over time, including the CDC and the Los Angeles Times, which had both made explicit eVorts to debunk it.
152
NORBERT SCHWARZ et al.
Second, people may infer the credibility of the source from the strength with which they hold a belief, as Fragale and Heath (2004) proposed. Highly credible sources usually elicit high confidence in the information they convey. Drawing on this relationship, people may consult their confidence in their beliefs as a source of information that bears on the likely source. To test this possibility, Fragale and Heath (2004) exposed participants to statements like, ‘‘The wax used to line Cup‐o‐Noodles cups has been shown to cause cancer in rats.’’ They manipulated participants’ acceptance of these statements by presenting them either two or five times. Next, participants learned that some statements were taken from the National Enquirer (a low‐ credibility source) and some from Consumer Reports (a high‐credibility source). Their task was to guess which source had originally reported which statement. As predicted, a given statement was more likely to be attributed to Consumer Reports than to the National Enquirer the more often it had been presented. Thus, frequent exposure does not only increase the acceptance of a statement as true, as reviewed above, but also facilitates the attribution of the ‘‘true’’ statement to a highly credible source. This source attribution, in turn, may increase the likelihood that recipients convey the information to others, who themselves are more likely to accept (and spread) it, given its alleged credible source (Rosnow & Fine, 1976). These examples suggest that countering false information in ways that repeat it may further contribute to its dissemination by associating the information with a credible source, either through source confusion or through erroneous inferences of source credibility.
D. CONCLUSIONS In combination, the reviewed findings dovetail with research into rumor transmission (Allport & Postman, 1947; Koenig, 1985; Rosnow & Fine, 1976) by highlighting the risks of repeating erroneous information. Public information campaigns that confront myths with facts, or warn people that a given claim is false, necessarily reiterate the information they want to discredit. This strategy is successful as long as people remember what is true and what is false. Unfortunately, memory for these details fades quickly. When the false claims are encountered again on a later occasion, all that is left may be the vague feeling that ‘‘I heard something like this before.’’ This sense of familiarity, in turn, will foster the acceptance of statements as true. Once a statement is accepted as true, people are likely to attribute it to a credible source—which, ironically, may often be the source that attempted to discredit it (Emery, 2000)—lending the statement additional credibility when conveyed to others.
SETTING PEOPLE STRAIGHT
153
Perceived familiarity exerts the observed influence because, under natural conditions, frequent exposure to an opinion is often a valid cue that many people share it, providing the social consensus information that figures prominently in Festinger’s ‘‘secondary reality tests’’ (1954). Moreover, daily conversational conduct is based on the assumption that communicated information is truthful and relevant (Grice, 1975), again fostering acceptance, in particular when it seems that one has ‘‘heard this repeatedly.’’ Unfortunately, people are not good at tracking where they heard what how often and any variable that facilitates fluent processing may elicit erroneous estimates of high consensus and the acceptance of the statement as true—even when all repetitions came from the same single source (Weaver et al., in press) or the fluency of processing is solely due to the quality of the print font (Reber & Schwarz, 1999). To counteract this powerful influence of the fluency–familiarity–truth link, it is not suYcient that the correct information is compelling and highly memorable. It also needs to be closely linked to the false statement to ensure that exposure to the ‘‘myth’’ prompts recall of the ‘‘fact.’’ This is diYcult to achieve, although memorable slogans that link the myth and fact may provide a promising avenue. In most cases, however, it will be safer to refrain from any reiteration of the myths and to focus solely on the facts. The more the facts become familiar and fluent, the more likely it is that they will be accepted as true and serve as the basis of people’s judgments and intentions.
V. Implications and Future Directions Guided by the information processing paradigm and its computer metaphor (Lachman, Lachman, & Butterfield, 1979), psychologists emphasized the ‘‘cold’’ cognitive processes of information encoding, storage, and retrieval. In social psychology, this emphasis was soon complemented by an exploration of processes that do not easily fit the computer metaphor, including the use of experiential information. While the initial work addressed the role of moods and emotions, later work extended the analysis to metacognitive experiences and bodily sensations (for a review see Schwarz & Clore, in press). Paralleling these developments, cognitive psychologists rediscovered the role of subjective experiences in memory, which figured prominently in early theorizing, but went out of fashion with the behaviorist as well as cognitive revolution (for reviews see Brewer, 1992; Roediger, 1996). After decades of pervasive ‘‘neglect of conscious experience’’ (Tulving, 1989, p. 4), it is now widely accepted that an understanding of memory and judgment
154
NORBERT SCHWARZ et al.
requires the consideration of the phenomenal experiences that accompany cognitive processes. Nevertheless, many substantive areas of research remained untouched by these developments. As the reviewed research illustrates, we cannot understand the emergence of many judgmental biases without taking the interplay of declarative and experiential information into account, nor can we design strategies to debias judgment without doing so. Similarly, many well‐intentioned public information campaigns may be counterproductive when they are solely informed by theories that emphasize the role of declarative information. Much remains to be learned about the role of metacognitive experiences in these areas. This final section addresses open issues and likely complications, and identifies avenues for future research.
A. NAIVE THEORIES Our discussion of accessibility experiences focused on the naive theory identified in Tversky and Kahneman’s availability heuristic (1973): when there are many (few) examples or reasons, it is easy (diYcult) to bring some to mind. Applying this naive theory, people infer from the experienced ease or diYculty that there are many or few reasons of the sought after type. A growing body of research indicates, however, that people hold numerous other naive theories about the diYculty of recall and thought generation (Schwarz, 2004; Skurnik, Schwarz, & Winkielman, 2000). They assume that information that is well represented in memory is easier to recall than information that is poorly represented, making ease of recall a cue for memory judgments (Winkielman, Schwarz, & Belli, 1998); that recent events are easier to recall than distant events, making ease of recall a cue for temporal distance (Sanna, Chang, & Carter, 2004; Schwarz, Cho, & Xu, 2005); that important events are easier to recall than unimportant ones; and that thought generation is easier when one has high rather than low expertise, making ease a cue for importance and expertise (Schwarz et al., 2005). Drawing on these naive theories, people may consider ease of thought generation more informative, and diYculty less informative, when the event is distant rather than recent, unimportant rather than important, and when they lack rather than have domain expertise. Hence, diVerent naive theories of mental processes suggest variables that may moderate the size of bias, and the success of debiasing, by influencing the perceived informational value of metacognitive experiences. We consider this a particularly promising area for future research. Importantly, once a specific naive theory is applied, it renders the metacognitive experience uninformative for later judgments that require application of a diVerent theory (Schwarz, 2004). Schwarz et al.’s participants (2005), for example, inferred from the diYculty of listing many ‘‘fine Italian
SETTING PEOPLE STRAIGHT
155
restaurants’’ that there are few in town when asked for a frequency judgment, but that they did not know much about town when asked for a knowledge judgment. Each of these judgments, however, entails an attribution of the recall experience, either to the number of restaurants in town or to one’s own expertise. Once this implicit attribution is made, the experience is uninformative for the next judgment that requires a diVerent theory and people turn to thought content instead. Hence, those who first concluded that their diYculty reflects a lack of knowledge subsequently inferred that there are many fine Italian restaurants in town—after all, they listed quite a few and they do not even know much about town. Conversely, those who first concluded that there are not many restaurants in town subsequently reported high expertise—after all, there are not many such restaurants and they nevertheless listed quite a few, so they must know a lot about town. We anticipate similar judgment‐order eVects in debiasing. Whenever a preceding judgment entails an attribution of one’s metacognitive experience, it may render the experience uninformative for other judgments, paralleling the misattribution eVects reviewed above. Similar considerations apply to people’s inferences from experienced fluency of processing (see Schwarz, 2004). The model shown in Fig. 1 incorporates naive theories and assumes that they supply the inference rules that are applied to a given accessibility experience. Once an experience‐based inference is drawn, the experience is uninformative for judgments that require the application of a diVerent inference rule.
B. TEMPORAL CHANGES Theoretically, it is likely that accessibility experiences change over time, with diVerential eVects on immediate and delayed judgments. Suppose, for example, that participants in a hindsight bias experiment are asked to generate many thoughts about alternatives. Experiencing this as diYcult, they will conclude that the event was rather inevitable, as seen above (Sanna et al., 2002a,b). Yet a couple of days later the previously generated thoughts about alternatives may come to mind easily, potentially turning the initial backfire eVect into a delayed debiasing success. Conversely, suppose that the outcome of an event initially elicits high surprise. Surprise is ‘‘a cognitive state having to do with unexpectedness’’ (Ortony, Clore, & Collins, 1988, p. 33) and its experience curtails hindsight bias (Ofir & Mazursky, 1997). Yet surprising events also elicit explanatory activity and sense‐making (Pezzo, 2003), which renders outcome congruent thoughts easily accessible downstream. As a result, a metacognitive experience that curtails hindsight bias initially may give rise to increased hindsight bias after a delay.
156
NORBERT SCHWARZ et al.
Temporal shifts may be particularly likely when outcomes are especially important, striking, or impactful. At first, the shock of the outcome elicits a strong sense of surprise, and events appear to have been very unpredictable. However, as people strive to understand what happened, the search for explanations makes potential causes highly accessible, resulting in the conclusion that it could have been foreseen and should have been prevented. Public discourse following the 9/11 terror attacks is consistent with this conjecture (for a review see Wirtz, 2006). Media coverage may further change the metacognitive experiences associated with events through frequent repetition of key event scenes, rendering them highly accessible and fluent, with far reaching implications for public opinion, calls for relevant policy, and individual coping strategies. To date, little is known about such temporal trajectories.
C. METACOGNITIVE EXPERIENCES IN GROUPS Many decisions are made in groups and theories of group decision making share the usual focus on declarative information (for reviews see Kerr & Tindale, 2004; Kerr et al., 1996). At present, little is known about the role of metacognitive experiences in group decision making (but see Sanna, Parks, Chang, & Carter, 2005). For example, does the ease of collectively generating thoughts influence groups in the same way as individuals? Which naive theories do people hold about thought generation in groups? They probably assume that pooled resources make collective thought generation easier than individual thought generation; but do they also assume that group pressure or distraction may impede thought generation under some conditions? Which theory do they apply under which conditions? Assuming that people believe that coming up with many thoughts is easier in groups than individually, would group members find a collective experience of ease less informative than an individual one? Conversely, would they find a collective experience of diYculty more informative than an individual one? Exploring these and related issues will contribute to our understanding of group decision making and may fruitfully extend the exploration of metacognitive processes beyond the individual domain.
Acknowledgments We thank Daniel Kahneman, Johannes Keller, Ruth Mayo, and Daphna Oyserman for critical discussions of an earlier draft of this chapter.
SETTING PEOPLE STRAIGHT
157
References Aarts, H., & Dijksterhuis, A. (1999). How often did I do it? Experienced ease of retrieval and frequency estimates of past behavior. Acta Psychologica, 103, 77–89. Allport, F. H., & Lepkin, M. (1945). Wartime rumors of waste and special privilege: Why some people believe them. Journal of Abnormal and Social Psychology, 40, 3–36. Allport, F. H., & Postman, L. J. (1947). The psychology of rumor. New York: Holt, Rinehart, & Winston. Arkes, H. R. (1991). Costs and benefits of judgmental errors: Implications for debiasing. Psychological Bulletin, 110, 486–498. Begg, I. M., Anas, A., & Farinacci, S. (1992). Dissociation of processes in belief: Source recollection, statement familiarity, and the illusion of truth. Journal of Experimental Psychology: General, 121, 446–458. Bodenhausen, G. V., & Wyer, R. S. (1987). Social cognition and social reality: Information acquisition and use in the laboratory and the real world. In H. J. Hippler, N. Schwarz, & S. Sudman (Eds.), Social information processing and survey methodology (pp. 6–41). New York: Springer‐Verlag. Bredart, S., & Modolo, K. (1988). Moses strikes again: Focalization eVect on a semantic illusion. Acta Psychologica, 67, 135–144. Brewer, W. F. (1992). Phenomenal experience in laboratory and autobiographical memory. In M. A. Conway, D. C. Rubin, H. Spinner, & W. A. Wagenaar (Eds.), Theoretical perspectives in autobiographical memory (pp. 31–51). Netherlands: Kluwer. Buehler, R., GriYn, D., & MacDonald, H. (1997). The role of motivated reasoning in optimistic time predictions. Personality and Social Psychology Bulletin, 23, 238–247. Buehler, R., GriYn, D., & Ross, M. (1994). Exploring the ‘‘planning fallacy’’: Why people underestimate their task completion times. Journal of Personality and Social Psychology, 67, 366–381. Camerer, C. F., & Hogarth, R. M. (1999). The eVects of financial incentives in experiments: A review and capital‐labor‐production framework. Journal of Risk and Uncertainty, 19, 7–42. Christensen, A. J., Moran, P. J., & Wiebe, J. S. (1999). Assessment of irrational health beliefs. Health Psychology, 18, 169–176. Christensen‐Szalanski, J. J. J., & Willham, C. F. (1991). The hindsight bias: A meta‐analysis. Organizational Behavior and Human Decision Processes, 48, 147–168. Dechow, P. M., Hutton, A. P., & Sloan, R. G. (2000). The relation between analysts’ forecasts of long‐term earnings growth and stock price performance following equity oVerings. Contemporary Accounting Research, 17, 1–32. Emery, D. (2000, February 23). The great banana scare of 2000. Retrieved July 17, 2004, from http://urbanlegends.about.com/od/fooddrink/a/killer_bananas.htm Erickson, T. A., & Mattson, M. E. (1981). From words to meaning: A semantic illusion. Journal of Verbal Learning and Verbal Behavior, 20, 540–552. Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 123–146. FischhoV, B. (1975). Hindsight 6¼ foresight: The eVect of outcome knowledge on judgments under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1, 288–299. FischhoV, B. (1982). Debiasing. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgment under uncertainty: Heuristics and biases (pp. 442–444). New York: Cambridge University Press. Fragale, A. R., & Heath, C. (2004). Evolving information credentials: The (mis)attribution of believable facts to credible sources. Personality and Social Psychology Bulletin, 30, 225–236.
158
NORBERT SCHWARZ et al.
Gilovich, T., GriYn, D., & Kahneman, D. (2002). Heuristics and biases: The psychology of intuitive judgment. New York: Cambridge University Press. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semantics (Vol. 3: Speech acts, pp. 41–58). New York: Academic Press. Guilbault, R. L., Bryant, F. B., Brockway, H. H., & Posavac, E. J. (2004). A meta‐analysis of research on hindsight bias. Basic and Applied Social Psychology, 26, 103–117. Haddock, G., Rothman, A. J., Reber, R., & Schwarz, N. (1999). Forming judgments of attitude certainty, importance, and intensity: The role of subjective experiences. Personality and Social Psychology Bulletin, 25, 771–782. Hasher, L., Goldstein, D., & Toppino, T. (1977). Frequency and the conference of referential validity. Journal of Verbal Learning and Verbal Behavior, 16, 107–112. Hawkins, S. A., & Hastie, R. (1990). Hindsight: Biased judgments of past events after the outcomes are known. Psychological Bulletin, 107, 311–327. Hawkins, S. A., & Hoch, S. J. (1992). Low‐involvement learning: Memory without evaluation. Journal of Consumer Research, 19, 212–225. Heiman, V. B. (1990). Auditors’ assessments of the likelihood of error explanations in analytical review. The Accounting Review, 65, 875–890. Hovland, C. I., & Weiss, W. (1951). The influence of source credibility on communication eVectiveness. Public Opinion Quarterly, 15, 635–650. Jacoby, L. L., Kelley, C. M., & Dywan, J. (1989). Memory attributions. In H. L. Roediger & F. I. M. Craik (Eds.), Varieties of memory and consciousness: Essays in honour of Endel Tulving (pp. 391–422). Hillsdale, NJ: Erlbaum. Kadous, K., Krische, S. D., & Sedor, L. M. (2006). Using counter‐explanation to limit analysts’ forecast optimism. The Accounting Review, 81, 377–397. Kahneman, D., Slovic, P., & Tversky, A. (Eds.) (1982). Judgment under uncertainty: Heuristics and biases. New York: Cambridge University Press. Kahneman, D., & Tversky, A. (1979). Intuitive prediction: Biases and corrective procedures. Management Science, 12, 313–327. Kelley, C. M., & Rhodes, M. G. (2002). Making sense and nonsense of experience: Attributions in memory and judgment. The Psychology of Learning and Motivation, 41, 293–320. Kerr, N. L., MacCoun, R. J., & Kramer, G. P. (1996). Bias in judgment: Comparing groups and individuals. Psychological Review, 103, 687–719. Kerr, N. L., & Tindale, R. S. (2004). Group performance and decision making. Annual Review of Psychology, 55, 623–655. Koehler, D. J. & Harvey, N. (Eds.) (2004). Blackwell handbook of judgment and decision making. Oxford: Blackwell Publishing. Koenig, F. (1985). Rumor in the market place: The social psychology of commercial hearsay. Dover, MA: Auburn House. Koonce, L. (1992). Explanation and counter‐explanation during audit analytical reviews. The Accounting Review, 67, 59–76. Koriat, A., Lichtenstein, S., & FischhoV, B. (1980). Reasons for confidence. Journal of Experimental Psychology: Human Learning & Memory, 6, 107–118. Lachman, R., Lachman, J. T., & Butterfield, E. C. (1979). Cognitive psychology and information processing. Hillsdale, NJ: Erlbaum. Larrick, R. P. (2004). Debiasing. In D. J. Koehler & N. Harvey (Eds.), Blackwell handbook of judgment and decision making (pp. 316–337). Oxford: Blackwell Publishing. Leary, M. R. (1982). Hindsight distortion and the 1980 presidential election. Personality and Social Psychology Bulletin, 8, 257–263. Lerner, J. S., & Tetlock, P. E. (1999). Accounting for the eVects of accountability. Psychological Bulletin, 125, 225–275.
SETTING PEOPLE STRAIGHT
159
Lopes, L. L. (1994). Psychology and economics: Perspectives on risk, cooperation, and the marketplace. Annual Review of Psychology, 45, 197–227. McGlone, M. S., & Tofighbakhsh, J. (2000). Birds of a feather flock conjointly (?): Rhyme as reason in aphorisms. Psychological Science, 11, 424–428. McQuail, D. (2000). McQuail’s mass communication theory. Newbury Park, CA: Sage. Newby‐Clark, I. R., Ross, M., Buehler, R., Koehler, D. J., & GriYn, D. (2000). People focus on optimistic scenarios and disregard pessimistic scenarios while predicting task completion times. Journal of Experimental Psychology: Applied, 6, 171–182. Nisan, M. (1972). Dimension of time in relation to choice behavior and achievement orientation. Journal of Personality and Social Psychology, 21, 175–182. Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. New York: Prentice Hall. Novemsky, N., Dhar, R., Schwarz, N., & Simonson, I. (in press). Preference fluency in consumer choice. Journal of Marketing Research. Ofir, C., & Mazursky, D. (1997). Does a surprising outcome reinforce or reverse the hindsight bias? Organziational Behavior and Human Decision Processes, 69, 51–57. Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive structure of emotions. London: Cambridge University Press. Park, D. C. (2000). The basic mechanisms accounting for age‐related decline in cognitive function. In D. C. Park & N. Schwarz (Eds.), Cognitive aging: A primer (pp. 3–22). Philadelphia, PA: Psychology Press. Park, H, & Reder, L. M. (2003). Moses illusion. In R. F. Pohl (Ed.), Cognitive illusions (pp. 275–292). New York, NY: Psychology Press. Pezzo, M. V. (2003). Surprise, defence, or making sense: What removes the hindsight bias? Memory, 11, 421–441. Plous, S. (1993). The psychology of judgment and decision making. New York: McGraw‐Hill. Pohl, R. F. (2005). Cognitive illusions: A handbook on fallacies and biases in thinking, judgment and memory. New York: Routledge. Powell, J. L. (1988). A test of the knew‐it‐all‐along eVect in the 1984 presidential statewide elections. Journal of Applied Social Psychology, 18, 760–773. Reber, R., & Schwarz, N. (1999). EVects of perceptual fluency on judgments of truth. Consciousness and Cognition, 8, 338–342. Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience? Personality and Social Psychology Review, 8, 364–382. Reder, L. M., & Kusbit, G. W. (1991). Locus of the Moses illusion: Imperfect encoding, retrieval or match? Journal of Memory and Language, 30, 385–406. Rice, R. & Atkin, C. (Eds.) (2001). Public communication campaigns (3rd ed.). Newbury Park, CA: Sage. Roediger, H. L. (1996). Memory illusions. Journal of Memory and Language, 35, 76–100. Rosnow, R. L., & Fine, G. A. (1976). Rumor and gossip: The social psychology of hearsay. New York: Elsevier. Sanna, L. J., Chang, E. C., & Carter, S. E. (2004). All our troubles seem so far away: Temporal pattern to accessible alternatives and retrospective team appraisals. Personality and Social Psychology Bulletin, 30, 1359–1371. Sanna, L. J., Parks, C. D., Chang, E. C., & Carter, S. E. (2005). The hourglass is half full or half empty: Temporal framing and the group planning fallacy. Group Dynamics: Theory, Research, and Practice, 9, 173–188. Sanna, L. J., & Schwarz, N. (in press). Metacognitive experiences and hindsight bias: It’s not just the thought (content) that counts! Social Cognition.
160
NORBERT SCHWARZ et al.
Sanna, L. J., & Schwarz, N. (2006). Metacognitive experiences and human judgment: The case of hindsight bias and its debiasing. Current Directions in Psychological Science, 15, 172–176. Sanna, L. J., & Schwarz, N. (2003). Debiasing the hindsight bias: The role of accessibility experiences and (mis)attributions. Journal of Experimental Social Psychology, 39, 287–295. Sanna, L. J., & Schwarz, N. (2004). Integrating temporal biases: The interplay of focal thoughts and accessibility experiences. Psychological Science, 15, 474–481. Sanna, L. J., Schwarz, N., & Small, E. M. (2002a). Accessibility experiences and the hindsight bias: I knew it all along versus it could never have happened. Memory & Cognition, 30, 1288–1296. Sanna, L. J., Schwarz, N., & Stocker, S. L. (2002b). When debiasing backfires: Accessible content and accessibility experiences in debiasing hindsight. Journal of Experimental Psychology: Learning, Memory, and Cognition, 28, 497–502. Schwarz, N. (1994). Judgment in a social context: Biases, shortcomings, and the logic of conversation. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 26, pp. 123–162). San Diego: Academic Press. Schwarz, N. (1996). Cognition and communication: Judgmental biases, research methods, and the logic of conversation. Hillsdale, NJ: Erlbaum. Schwarz, N. (1998). Accessible content and accessibility experiences: The interplay of declarative and experiential information in judgment. Personality and Social Psychology Review, 2, 87–99. Schwarz, N. (2004). Metacognitive experiences in consumer judgment and decision making. Journal of Consumer Psychology, 14, 332–348. Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer‐Schatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61, 195–202. Schwarz, N., Cho, H., & Xu, J. (2005). Diverging inferences from identical inputs: The role of naive theories. Paper presented at the European Association of Experimental Social Psychology, Wu¨rzburg, Germany. Schwarz, N., & Clore, G. L. (in press). Feelings and phenomenal experiences. In E. T. Higgins A. Kruglanski (Eds.), Social psychology: Handbook of basic principles (2nd ed.). New York: Guilford. Shafir, E., & LeBoeuf, R. A. (2002). Rationality. Annual Review of Psychology, 53, 491–517. Shepperd, J. A., Ouellette, J. A., & Fernandez, J. K. (1996). Abandoning unrealistic optimism: Performance estimates and the temporal proximity of self‐relevant feedback. Journal of Personality and Social Psychology, 70, 844–855. Skurnik, I., Schwarz, N., & Winkielman, P. (2000). Drawing inferences from feelings: The role of naive beliefs. In H. Bless & J. Forgas (Eds.), The message within: The role of subjective experience in social cognition and behavior (pp. 162–175). Philadelphia, PA: Psychology Press. Skurnik, I., Yoon, C., Park, D. C., & Schwarz, N. (2005). How warnings about false claims become recommendations. Journal of Consumer Research, 31, 713–724. Skurnik, I., Yoon, C., & Schwarz, N. (2007). Education about flu can reduce intentions to get a vaccination. Manuscript under review. Song, H., & Schwarz, N. (2007). Low processing fluency attenuates the Moses illusion. Manuscript under review. Stepper, S., & Strack, F. (1993). Proprioceptive determinants of emotional and nonemotional feelings. Journal of Personality and Social Psychology, 64, 211–220. Tellis, G. J. (2004). EVective advertising. Thousand Oaks, CA: Sage. Tormala, Z. L., Petty, R. E., & Brinol, P. (2002). Ease of retrieval eVects in persuasion: A self‐ validation analysis. Personality and Social Psychology Bulletin, 28, 1700–1712.
SETTING PEOPLE STRAIGHT
161
Tulving, E. (1989). Memory: Performance, knowledge, and experience. European Journal of Cognitive Psychology, 1, 3–26. Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207–232. van Oostendorp, H., & de Mul, S. (1990). Moses beats Adam: A semantic relatedness eVect on a semantic illusion. Acta Psychologica, 74, 35–46. Wa¨nke, M., Bless, H., & Biller, B. (1996). Subjective experiences versus content of information in the construction of attitude judgments. Personality and Social Psychology Bulletin, 22, 1105–1113. Wa¨nke, M., Bohner, G., & Jurkowitsch, A. (1997). There are many reasons to drive a BMW— Surely you know one: Ease of Argument Generation influences Brand Attitudes. Journal of Consumer Research, 24, 70–77. Weaver, K., Garcia, S. M., Schwarz, N., & Miller, D. T. (in press). Metacognitive beliefs about group opinion: Inferences of extensity from familiarity. Journal of Personality and Social Psychology. Webley, P., Burgoyne, C. B., Lea, S. E. G., & Young, B. M. (2001). The economic psychology of everyday life. New York: Psychology Press. Werth, L., & Strack, F. (2003). An inferential approach to the knew‐it‐all‐along phenomenon. Memory, 11, 411–419. Whittlesea, B. W. A. (1993). Illusions of familiarity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 1235–1253. Whittlesea, B. W. A., Jacoby, L. L., & Girard, K. (1990). Illusions of immediate memory: Evidence of an attributional basis for feelings of familiarity and perceptual quality. Journal of Memory and Language, 29, 716–732. Wilson, T. D., & Brekke, N. (1994). Mental contamination and mental correction: Unwanted influences on judgments and evaluations. Psychological Bulletin, 116, 117–142. Winkielman, P., & Cacioppo, J. T. (2001). Mind at ease puts a smile on the face: Psychophysiological evidence that processing facilitation leads to positive aVect. Journal of Personality and Social Psychology, 81, 989–1000. Winkielman, P., Schwarz, N., & Belli, R. F. (1998). The role of ease of retrieval and attribution in memory judgments: Judging your memory as worse despite recalling more events. Psychological Science, 9, 124–126. Winkielman, P., Schwarz, N., Fazendeiro, T, & Reber, R. (2003). The hedonic marking of processing fluency: Implications for evaluative judgment. In J. Musch & K. C. Klauer (Eds.), The psychology of evaluation: AVective processes in cognition and emotion (pp. 189–217). Mahwah, NJ: Erlbaum. Wirtz, J. J. (2006). Responding to surprise. Annual Review of Political Science, 9, 45–65. Witherspoon, D., & Allan, L. G. (1985). The eVects of a prior presentation on temporal judgments in a perceptual identification task. Memory and Cognition, 13, 103–111. Wyer, R. S., & Srull, T. K. (1989). Memory and cognition in its social context. Hillsdale, NJ: Erlbaum.
MULTIPLE SOCIAL CATEGORIZATION
Richard J. Crisp Miles Hewstone
Social categorization defines intergroup relations in ways that are not always simple and clear‐cut, but often complex and multifaceted. There are diverse, overlapping, and interwoven ways in which we can describe ourselves and others: young, British, Asian, woman, Muslim, engineer, and so on. We know that which of these identities is used in person perception depends upon a variety of factors, including context and motivation; but also that identities are not mutually exclusive. We can be, and often are, identified and we identify others according to a combination of group memberships: a disabled athlete, a young Briton, a female engineer. Research on multiple social categorization has found that our many and varied, cross‐cutting and convoluted, social identities have significant implications for understanding, and attenuating, prejudice and intergroup discrimination. This chapter reviews our own and others’ research on multiple social categorization and discusses when and how diVerent multiple categorization strategies can lead to more positive intergroup attitudes. We develop a model of the processes involved and explain how this work can help policy makers and practitioners in their eVorts to promote, encourage, and enhance harmonious intergroup relations. . . . there is no contradiction. I’m Muslim, I’m British, I’m Asian, I’m an imam, I’m a teacher. The Guardian, July 16, 2005 The hope of harmony in the contemporary world lies to a great extent in a clearer understanding of the pluralities of human identity, and in the appreciation that they cut 163 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39004-1
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
164
RICHARD J. CRISP AND MILES HEWSTONE across each other and work against a sharp separation along one single hardened line of impenetrable division. Amartya Sen, Identity and Violence: The Illusion of Destiny, 2006, p. xiv
I. Introduction In the opening scene of Ken Loach’s film ‘‘Ae Fond Kiss,’’ an ethnically Pakistani girl at a school in Glasgow, Scotland stands up in a debate and proclaims passionately: I reject the West’s simplification of a Muslim . . . I am a Glaswegian, Teenager, Woman, Woman of Muslim descent, who supports Glasgow Rangers in a Catholic school . . . cos’ I’m a dazzling mixture. I’m proud of it.
This sentiment illustrates a basic truth of social reality: the world presents us with a tremendously complex and intricate array of identities from which to construct a definition of ourselves and others. Think of your ethnicity, nationality, gender, religion, politics, occupation, friends, family, and sports allegiance. Social identities are complex and diVerentiated, fluid and dynamic, convoluted and multifaceted. We live in a climate of unprecedented intercultural exchange, a world where the geographical boundaries that segregated peoples of diVering color and culture in the twentieth century are being slowly but surely eroded. Many societies are increasingly multicultural, where awareness of our own, and others’, multiple aYliations is an unavoidable characteristic of social life. This complexity can sometimes be a source of challenges, both to those who are categorized in multiple terms and to those who must learn to categorize along multiple dimensions; as Edward Said wrote, ‘‘I have retained this unsettled sense of many identities—mostly in conflict with each other—all of my life, together with an acute memory of the despairing feeling that I wish we could have been all‐Arab, or all‐ European and American, or all‐Orthodox Christian, or all‐Muslim, or all‐ Egyptian, and so on’’ (Said, 1999, p. 5). But it can also be a source of great opportunities. As Sen (2006) argued, ‘‘A strong—and exclusive—sense of belonging to one group can in many cases carry with it the perception of distance and divergence from other groups’’ (pp. 1–2). This chapter is all about these multiple identities, what this multiplicity means for perceptions of the self and others, for behavior, and what implications an increasing awareness of crosscutting and complex bases for group membership has for intergroup relations.
MULTIPLE SOCIAL CATEGORIZATION
165
A. SOCIAL CATEGORIZATION The importance of the categorization process in defining social perception and behavior is well established (Brewer, 1979; Doise, 1978; Hewstone, Rubin, & Willis, 2002; Mullen, Brown, & Smith, 1992; Tajfel, 1978; Tajfel, Flament, Billig, & Bundy, 1971). Wilder (1986) provided a superb exposition of how mere categorization of persons into diVerent groups engages a series of assumptions that foster intergroup biases. Writing in the Advances, he reviewed evidence of consequences of social categorization including perceived intragroup similarity (but between‐group diVerence), biased causal attributions and memory processes, and in‐group favoritism and out‐group derogation. Among the more telling of his own findings was Wilder’s demonstration of accentuation eVects with regard to expectations: when we believe that individuals share a salient group membership we expect them to be similar; but when we believe that they are divided by categorization, then we expect them to be diVerent from each other (see Allen & Wilder, 1975). In many ways we intend our chapter to provide a timely extension to Wilder’s seminal theoretical exposition of categorization eVects, an extension that builds on, qualifies, and applies categorization theory to the complexity of intergroup relations in the twenty‐first century. Categorization is an integral and essential part of social perception (Allport, 1954; Chaiken & Trope, 1999; Fiske & Neuberg, 1990). Whether the people we meet are black or white, male or female, young or old, such social classifications have significant implications for how we think about ourselves and form impressions of others. In any given situation the categories that define ourselves and others can depend on varied factors such as context (Oakes, Turner, & Haslam, 1991) and motivation (Sinclair & Kunda, 1999). But our approach has a new emphasis: it is increasingly apparent that in many contexts multiple bases for social categorization can be salient, combined, and used simultaneously (see Brewer, Ho, Lee, & Miller, 1987; Stangor, Lynch, Duan, & Glass, 1992). So social perception is a complex business and in an increasingly multicultural, multiethnic, and multireligious world the social reality is that there are now multiple ways in which we can be the same as, or diVerent from, others. This characteristic of contemporary intergroup relations has direct consequences for critical social issues such as prejudice, social exclusion, and intergroup conflict. The increasing complexity of social categorization presents an intriguing challenge to theoretical accounts of group processes and intergroup relations. In this chapter, we examine how classification of ourselves, and others, along these multiple criteria, can have significant implications for psychological, social, and behavioral processes.
166
RICHARD J. CRISP AND MILES HEWSTONE
B. MULTIPLE CATEGORIZATION Updating and refining theories of categorization to consider the multiplicity of group life is critical because of the impact such processes have on intergroup relations. Social categorization matters especially because it is intimately linked to extreme forms of intergroup behavior and social conflict. Especially in highly segmented societies that are diVerentiated along a single primary categorization, such as ethnicity or religion (e.g., Rwanda, Northern Ireland), Brewer (2001) predicted a strong relationship between intense in‐group favoritism and out‐group antagonism, especially if categorization is dichotomous. The problem with such segmentation is that it promotes social comparison and perceptions of conflict of interest in zero‐ sum terms (see Esses, Jackson, & Armstrong, 1998; Sherif, 1966), leading to negative attitudes and high potential for conflict. By contrast, the potential for intergroup conflict may be reduced in societies that are more complex and diVerentiated along multiple dimensions that are not perfectly correlated, that is crosscutting category dimensions (Evans‐Pritchard, 1940; Horowitz, 1985; LeVine & Campbell, 1972; Murphy, 1957; Wallace, 1973; for a review see Hewstone, Turner, Kenworthy, & Crisp, 2006). These societal arrangements reduce the intensity of the individual’s dependence on any particular in‐group for meeting psychological needs for inclusion (see Brewer, 1991, 1993), this, in turn, reduces the potential for polarizing loyalties along any single cleavage or group distinction and perhaps increases tolerance for out‐groups in general. Although our knowledge of the importance of multiple categorization phenomena is not new (anthropologists and sociologists provided early analyses, e.g., Evans‐Pritchard, 1940; Murphy, 1957; Simmel, 1950; see also Crisp & Hewstone, 2006), we have only recently paid these important phenomena the attention they deserve (but see Allport’s, 1954, acknowledgement that people have multiple social identities that are hierarchically organized in terms of increasing inclusiveness). There is now extensive empirical support for the view that multiple identities exert a significant impact on social thought and action (Stryker & Statham, 1985; Tajfel, 1978; see Deaux, 1996 for a review). Society has changed: it is now the case that many societies are more multiethnic, multireligious, and multicultural than they were decades ago; an increasing number of individuals claim more than one ethnic identity, racial identification, and source of cultural competence (see Phinney & Alipuria, 2006). To deal with these new realities, respondents to some national census questions are now permitted to assert their multiple identities on oYcial forms (see Daniel, 2002). Reflecting these developments, social psychological research has now clearly established that, in many contexts, multiple bases for social categorization can be
MULTIPLE SOCIAL CATEGORIZATION
167
simultaneously salient and used in social perception (Crisp & Hewstone, 2000b; Crisp, Hewstone, & Cairns, 2001).
C. SOCIAL STRUCTURES AND PSYCHOLOGICAL PROCESS That people can be aYliated to multiple social categories is not a new idea, nor exclusively an idea explored by social psychologists. As we noted above, anthropologists, sociologists, and political scientists were among the first to observe some unique consequences associated with the apparent use of multiple categorization in social contexts. In a famous example, sociologist Georg Simmel (1950) observed that in the Australian Aborigine culture the whole population is divided into five ‘‘gentes’’ with members of the various gentes found across many diVerent tribes. Evans‐Pritchard (1940) and Murphy (1957) both documented cases of reduced conflict in cultures with crosscutting social structures (see also Coser, 1956; Deutsch, 1973; Guetzkow, 1955; Lipset, 1960; Rae & Taylor, 1970). LeVine and Campbell (1972) suggested that group members actively make use of crosscutting aYliations because they ensure security and stability (it is more diYcult, for instance, to have conflictual relations with a group based on territory that is simultaneously an ally according to common ancestry). Research has confirmed that crosscutting ties can be important at a societal level (Cairns & Mercer, 1984; Ross, 1985; see Phinney & Alipuria, 2006) and such category structures have emerged as a central concept for political scientists (see Carter, 2006; Horowitz, 1985; Wallace, 1973). Our own research has focused especially on defining the processes involved in multiple categorization. Our studies have shown how people use multiple criteria for categorization in both social and nonsocial judgment (e.g., Crisp, 2002; Crisp & Hewstone, 1999b), and how attention to these complex criteria predicts changes in how we evaluate people in diVerent groups (Crisp, Walsh, & Hewstone, 2006c; Hewstone, Islam, & Judd, 1993) as well as the extent to which we conform to our own groups’ stereotypes (Rosenthal & Crisp, 2006). We have established that the cognitive processes that form the basis for social category judgments vary as a function of the number of categories attended to with respect to perceived similarity (Crisp & Hewstone, 2000a) and memory representation (Crisp & Hewstone, 2001; Crisp et al., 2001), but also how motivational factors play a role, including self‐esteem (Crisp & Hewstone, 2000b) and emotional commitment (Crisp, Stone, & Hall, 2006b). We now know that thinking about multiple criteria can impact on social judgment and behavior in a number of specific ways, from ‘‘decategorizing’’ judgments (Crisp, Hewstone, & Rubin, 2001; Hall & Crisp, 2005) to inducing ‘‘recategorization’’ either by a blurring of category boundaries (Crisp, 2006; Crisp & Beck, 2005) or by triggering inconsistency
168
RICHARD J. CRISP AND MILES HEWSTONE
resolution processes (Hutter & Crisp, 2005). In this chapter, we review our work on the contextual antecedents, psychological processes, and social consequences of multiple social categorization; we integrate our past theorizing, and present a new inclusive theoretical model of why, when, and how multiple identities play a critical role in shaping our social worlds.
D. CHAPTER OVERVIEW We begin by introducing the basic conceptual and theoretical models that have been developed by social psychologists interested in multiple categorization. We then divide our review of existing theoretical and empirical work into two parts. In the first part we discuss research that has in common a focus on changing the representation of social categories. By this we mean any research that has investigated whether considering multiple bases for social categorization will change stereotypes of groups, change perceptions of how similar or diVerent groups are to one another, change how similar members of those groups are perceived to be to one another, and change how people evaluate and behave toward members of those diVerent groups. In this first part, we will consider cognitive models of how intra‐ and intercategory diVerentiation aVect evaluative judgment. We will use several terms to refer to such evaluative judgments in this chapter. Intergroup bias refers generally to the systematic tendency to evaluate one’s own membership group (the ‘‘in‐group’’) or its members more favorably than a nonmembership group (the ‘‘out‐group’’) or its members (Hewstone et al., 2002). Such intergroup bias can encompass behavioral discrimination, prejudiced attitudes, and stereotyped cognitions. More precisely, this group‐serving tendency can take the form of favoring the in‐ group (‘‘in‐group favoritism’’) and/or derogating the out‐group (‘‘out‐group derogation’’). Use of the term ‘‘bias’’ involves an interpretative judgment that the response is unfair, illegitimate, or unjustifiable, in the sense that it goes beyond the objective requirements or evidence of the situation (see Brewer & Brown, 1998; Fiske, 2000; Hewstone et al., 2002; Reynolds, Turner, & Haslam, 2000). We will argue that one mechanism is central to understanding multiple categorization eVects: category diVerentiation. We define category diVerentiation as the extent to which categories are distinct and nonoverlapping versus similar with blurred intergroup boundaries. The relationship between category diVerentiation and evaluation—a relationship that can be either positive or negative—is a key characteristic of the model we present. From this basic relationship, we discuss the impact of making either multiple shared or nonshared aYliations more or less salient and examine their impact on diVerentiation and, in turn, intergroup bias. We will discuss how a category
MULTIPLE SOCIAL CATEGORIZATION
169
diVerentiation mechanism underlies a range of previously observed consequences of multiple categorization. In the first part of this chapter we consider the common in‐group identity model (Gaertner & Dovidio, 2000) and the crossed categorization model (Crisp & Hewstone, 2006). We then examine more complex, extended, forms of crossed categorization, and the diVerent patterns of evaluation observed in such contexts (Brewer et al., 1987; Hewstone et al., 1993). We will show how a range of contextual antecedents interact with salient ‘‘converging’’ or ‘‘diverging’’ categorizations to predict judgment tendencies via the category diVerentiation mechanism (see Crisp & Hewstone, 1999a; Crisp, Ensari, Hewstone, & Miller, 2002; Migdal, Hewstone, & Mullen, 1998; Urban & Miller, 1998) (converging categorization describes where multiple intergroup diVerences coincide, and divergent categorization describes where multiple identities form the basis for both diVerentiation and assimilation; see Brewer & Campbell, 1976). These moderators can be both cognitive (categorization context) or aVective (mood) in nature. Finally, in this section, we will discuss how perceiver motivations to maintain, or retain, a positive and distinct social identity can moderate the impact that categorization has on diVerentiation. Such motivations are critical to our model because they determine whether intergroup diVerentiation leads to more, or less, intergroup bias. In the latter part of this chapter, we will argue that an equally important, but under‐researched, consequence of multiple categorization is its ability to change how perceivers form impressions of others. Most existing work has focused on when considering multiple bases for categorization changes how categories are represented—they become less homogeneous, or there is less category diVerentiation, due to the recognition of perceived similarities or diVerences when multiple shared and nonshared bases for aYliation are salient. We argue that multiple categorization can also change how people approach the task of impression formation. We discuss evidence that supports the idea that people can sometimes switch from a categorical to an individuated mode of perception in multiple category contexts, and argue that this process of ‘‘decategorization’’ can reduce intergroup bias in a diVerent way from blurring intergroup boundaries. We will discuss how multiple categorization can not only decategorize thought and judgment but also lead to more creative, divergent thought, and a deeper level of processing information. In this context, we will integrate developmental work on multiple classification skills, and argue that consideration of these diverse perspectives could become highly influential in future, elaborated strategies for improving intergroup relations. Our discussions of theory and research in this chapter will proceed in a relatively chronological order. Broadly speaking, we will discuss earlier studies before later studies, and develop our model accordingly, a model that reflects the culmination of our 10 years researching the psychological
170
RICHARD J. CRISP AND MILES HEWSTONE
and behavioral impact of multiple categorization. As our starting point we take a simple model of the relationship between categorization, diVerentiation, and category evaluation. At successive points in our review we will update this model, incorporating advances in the field as they were reflected in relevant research. The final model we present at the end of the chapter will bring together all of the research we review. We hope that our strategy of incrementally adding to the model will show how successive developments have helped us to refine our understanding of multiple categorization as both our own and other scholars’ research programs have progressed.
II. First Principles and Assumptions Categorization is ‘‘the process of understanding what something is by knowing what other things it is equivalent to, and what other things it is diVerent from’’ (McGarty, 1999, p. 1). This definition encapsulates much of what we will focus on in this chapter. We will discuss how intra‐ and intercategory diVerentiation—knowing whom we are the same as, and whom we are diVerent from—has an impact on social judgment and impression formation; and we will focus on the function of categorization—how it leads to understanding by providing meaning, establishing norms, and simplifying and aVording coherence to our social worlds. We can outline a taxonomy of category structure that captures the diVerent ways in which multiple identities can be combined contingent on the inherent hierarchical nature of social classification. Social categories, like categories more generally, can be hierarchically organized. For example, both Sunni and Shia communities are subcategories of Islam. People tend typically to categorize at the ‘‘basic level’’ (Rosch, 1978; Rosch, Mervis, Gray, Johnson, & Boyes‐Braem, 1976) which is in the middle of a conceptual hierarchy: social categories can be superordinate or subordinate to one another (Park & Rothbart, 1982; see Fig. 1; for an extended discussion of the nature of hierarchical categorization see McGarty, 2006). Multiple categorization, for our purposes, involves the use of classification criteria at varying levels of inclusiveness. Multiple categorization at the same level of inclusiveness is usually referred to as crossed categorization (Crisp & Hewstone, 1999a). Multiple categorization at diVerent levels of inclusiveness can take place with reference to superordinate categorization (Gaertner & Dovidio’s, 2000, common in‐group identity model) or subordinate categorization (subgrouping; Brewer, Dull, & Lui, 1981; Maurer, Park, & Rothbart, 1995). Thus, multiple categorization can be defined with respect to relations both between categories (in‐groups and out‐groups) and within hierarchies.
MULTIPLE SOCIAL CATEGORIZATION
171
European
British
English
German
Welsh
Fig. 1. An illustrative category hierarchy.
In this chapter, we discuss these multiple levels of category inclusiveness as they relate to our focus on representation, diVerentiation, recategorization, and decategorization. There are a number of factors that determine whether a category will be activated, including temporal primacy (we categorize on the basis of the features we encounter first; Jones & Goethals, 1972), contextual novelty or structural ‘‘fit’’ (when diVerences become salient, e.g. the only male in a room of females; Haslam, Oakes, Turner, & McGarty, 1995; Taylor, Fiske, EtkoV, & Ruderman, 1978), contextual accessibility (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), normative fit (i.e., whether categories have meaning in any given context; Oakes et al., 1991), and chronic accessibility (categorization in terms of some categories—race, age, gender—is so common that it can become automatized; Bargh & Pratto, 1986; Fiske & Neuberg, 1990). Categorization provides meaning to our social worlds, reducing uncertainty and providing clarity; it provides prescriptive norms that help us to understand ourselves in relation to others (Hogg, 2000). It also simplifies the task of impression formation, providing an eYcient mechanism for dealing with a complex world (Fiske & Neuberg, 1990; Macrae & Bodenhausen, 2001). In sum, social categorization is a psychological process that helps us navigate our social worlds by providing meaning, reducing uncertainty, and preserving cognitive resources, shaping our perceptions of, and behavior toward, social groups (Brewer, 1979; Brown & Turner, 1981; Macrae & Bodenhausen, 2001; Tajfel & Turner, 1979). There are two key assumptions we must consider before proceeding with our review: (1) people can and do use multiple categorization criteria in social judgment and (2) categorization can exert a direct influence on
172
RICHARD J. CRISP AND MILES HEWSTONE
category diVerentiation. The model we outline in this chapter rests on the assumption that considering multiple crosscutting or converging bases for categorization has a direct impact on how similar or diVerent others are perceived to be.
A. ASSUMPTION 1: PEOPLE CAN CATEGORIZE ALONG MULTIPLE DIMENSIONS SIMULTANEOUSLY The first assumption that underlies the research we will report later is that people (at least, sometimes) have both the cognitive resources and the motivation or inclination to classify themselves and others along multiple criteria simultaneously. Below, we first consider the conditions under which people do not classify people along multiple criteria, and second, when and how they do. 1. Category Selection Much of the research we will discuss in this chapter makes the implicit assumption that people can simultaneously think about multiple dimensions of categorization and can use category conjunctions when forming impressions of others (Crisp & Hewstone, 1999a; Migdal et al., 1998; Urban & Miller, 1998). This is not, however, to deny that category use is sometimes mutually exclusive, or that categories sometimes work against each other (Turner, 1999, p. 11; what has been referred to as ‘‘functional antagonism;’’ Turner et al., 1987, p. 49). Context is a powerful determinant of identity salience, and flexible self‐categorization along any one of multiple categorical criteria can occur as a function of situational cues (Turner, Oakes, Haslam, & McGarty, 1994). Contemporary examples of how context can prime particular social identities over others come from research into stereotype threat. Stereotype threat refers to the experience of becoming aware of one’s membership in a stigmatized group, and the subsequent impact this has on performance (Steele, 1997). For example, typically when women are told they will be compared with men on a mathematics test, they will perform worse than if no comparison had been made (for a review see Steele, Spencer, & Aronson, 2002). Shih, Pittinsky, and Ambady (1999) manipulated the salience of participants’ social identity (gender versus ethnic) and assessed its eVects on performance. In line with the stereotype threat model, Shih et al. (1999) found that Asian‐American women performed more poorly in a mathematics test when their gender identity rather than their Asian identity was made salient (because mathematics is related to negative stereotypes about women, but positive stereotypes about Asians) (see also Sinclair, Hardin, & Lowery, 2006). Other research has shown that the
MULTIPLE SOCIAL CATEGORIZATION
173
cultural context may simply make some categories more relevant than others in particular domains. Levin, Sinclair, Veniegas, and Taylor (2002), for instance, showed that Latinas and African‐American women did not diVer from their male counterparts in levels of perceived discrimination (i.e., ethnicity was dominant and gender was ignored) because ethnicity for these participants was a more salient basis for discrimination than gender. For white women in the same sample, gender was the dominant basis for assessing perceived discrimination, while ethnicity was ignored. Studies by Macrae, Bodenhausen, and Milne (1995) and van Twuyver and van Knippenberg (1999) also provide evidence that context leads to the dominance of some categories over others, but they also showed that activation of one categorization leads, concurrently, to inhibition of alternatives. Specifically, Macrae et al. (1995) found that the activation of one social category (e.g., Chinese) led to facilitated activation of the associated stereotype measured using response times (RTs) from a lexical decision task, but inhibition of the stereotype of a relevant, but not activated, second category (e.g., women) (see also Hugenberg & Bodenhausen, 2002). Van Twuyver and van Knippenberg observed similar eVects in a study employing the category confusion paradigm (Taylor et al., 1978). In this memory‐ based paradigm, participants are typically required to match statements with individuals who are members of either one group or another. The extent to which statements are misattributed to individuals within the same group, but not to members of the other group, is a measure of category use. Van Twuyver and van Knippenberg (1999) found that the use of a gender categorization was negatively correlated with the use of a student/teacher categorization. Such competitive activation of categories and stereotypes may be functional in many situations (i.e., enabling attentional focus to increase the eYciency of social perception; Dijksterhuis & van Knippenberg, 1998). Although context plays an important role in determining the categories that can be used to define the self and others (van Rijswijk & Ellemers, 2002), the multiple identities that people possess also oVer the perfect identity‐ management opportunity (Allen, Wilder, & Atkinson, 1983). People can strategically shift how they categorize themselves to maximize group‐derived self‐esteem gain (i.e., when threatened along one dimension of comparison, they can shift to compare themselves with others in more favorable terms using another dimension; Beach & Tesser, 1995; Tesser, 1988). For instance, Mussweiler, Gabriel, and Bodenhausen (2000) found that when Euro‐ American women were given false feedback that they had performed worse than an Asian‐American woman on a social perception task, they shifted identification away from their gender identity to their ethnic identity (relative to baseline identification with both). This can be explained in terms of Tesser’s self‐evaluation maintenance theory. When a similar other—in this case someone in your group, a woman—performs better than you, it is
174
RICHARD J. CRISP AND MILES HEWSTONE
possible to deflect the negative social comparison by redefining yourself along some alternative criterion that makes the comparison less relevant to your self‐concept. In this case, redefinition was achieved by shifting to an alternative social identity not shared with the comparison standard. People can also shift the identity they use to define others when under threat. Fein and Spencer (1997) demonstrated that targets who belonged to a negatively stereotyped group (Jews or Homosexuals) were more likely to be seen in terms of those negative stereotypes when participants received a threat to their self‐concept (in the form of negative feedback on an ostensibly unrelated intelligence test) than when they received no threat (and compared to positively stereotyped groups). Extending this work to consider multiply categorizable targets, Sinclair and Kunda (1999) found that following a criticism from a black doctor the stigmatized social category black was activated and the positive category, doctor, was inhibited. In contrast, positive feedback led to activation of the positive category, doctor, and inhibition of the negative category, black. Sinclair and Kunda (1999) argued that the negative stereotype was activated following negative feedback as a means of self‐esteem maintenance. This study supports the idea that not only context, but perceiver motivations, can determine which of multiple alternative categories can be used to define others. We are reminded of the New Yorker cartoon in which a doorstep pollster’s question is met with the reply, ‘‘How would you like me to answer that question? As a member of my ethnic group, educational class, income group, or religious category?’’ People are well aware of their multiple memberships, and can use them flexibly to define themselves and others depending on the situation at hand. To summarize, both cognitive and motivational factors contribute to the flexible redefinition of social identities in diVerent contexts. Categorization of self, and others, can vary for both contextual (salience) or motivational (self‐esteem maintenance) reasons. It is clear that sometimes just one of a range of multiple possible identities will be salient, relevant, and used. However, in the next section, we discuss research that has shown that whereas one category can dominate over another, in many situations peoples’ multiple categories are used simultaneously in social perception. 2. Simultaneous Categorization To examine whether people use multiple categorizations to guide their processing in social contexts, Arcuri (1982) used the category confusion paradigm (Taylor et al., 1978) in an experiment similar to that described earlier by van Twuyver and van Knippenberg (1999). By examining the inter‐ and intracategory errors on the matching task, Arcuri (1982) found that the ratio of intracategory errors to intercategory errors varied depending on the number of salient category dimensions (one or two). The way
MULTIPLE SOCIAL CATEGORIZATION
175
participants in this experiment represented the social context was therefore aVected by the presence of more than a single salient basis for categorization. Also investigating memory for categorization‐relevant information, Crisp et al. (2001) and Crisp and Hewstone (2001) reported further evidence for the use of multiple criteria in structuring person memory. In these experiments, we adapted a paradigm used by Park and Rothbart (1982) to allow subtle manipulation of multiple category memberships. In these experiments, undergraduate participants were presented with a short paragraph ostensibly taken from a local newspaper and were told that the experimenter was interested in how people read such stories. The eight stories concerned diVerent positive (e.g., citizens awards) and negative (e.g., drunk‐driving) events and were counterbalanced over participants. After the participant read the story and answered several filler questions, the experimenter left the room as if the ‘‘survey’’ were over. Forty‐five minutes later (at the end of a lecture), the experimenter reappeared and gave the participants a surprise cued recall test in which they were required to recall some specific information regarding the character in the story (e.g., age, occupation). Using this paradigm in Northern Ireland, we found that female and male, Catholic and Protestant participants recalled information relevant to both religion and gender categorizations, indicating that both had been attended to without any explicit prompting. We found similar eVects in a study that employed the same paradigm, but in a diVerent cultural context (Singapore), using an alternative category dimension (ethnicity: Chinese versus Malay). Even from an early age we can become aware of multiple identities, and think about them in complex ways, although this ability has a developmental trajectory. Allport (1954, p. 43) noted early work by Piaget and Weil (1951) that suggested children go through developmental stages reflecting an increasing understanding of the hierarchical (and nonmutually exclusive) nature of social categories. At age 7, children in this study indicated the relationship between town (Geneva) and country (Switzerland) as two circles side‐by‐side; they correspondingly pronounced identification with one or the other, but not both at the same time. They did not represent town and country by one circle enclosing another (and correspondingly recognize their simultaneous membership in both categories) until age 10 (see also Gaertner et al., in press; Markman & Callanan, 1984). Developmental research (Bigler & Liben, 1993) has shown that tendencies to notice multiple criteria for classification increase with age and that there are demonstrable individual diVerences in this ability (Averhart & Bigler, 1997). There are many other illustrations; for example studies by Stangor et al. (1992), van Knippenberg, van Twuyver, and Pepels (1994), and van Twuyver and van Knippenberg (1995) have all supported the idea that multiple dimensions of categorization can be used in social perception. Context and motivation can lead to selectivity (diVerential accessibility, van Knippenberg et al., 1994;
176
RICHARD J. CRISP AND MILES HEWSTONE
and diVerential situational relevance, ‘‘normative fit,’’ Oakes et al., 1991). However, in the absence of contextual factors making particular identities dominant, or perceiver motivations that consistently orient perceivers to some identities over others, people can and do notice and use their multiple identities for defining themselves and forming impressions of others. B. ASSUMPTION 2: CATEGORIZATION AFFECTS INTERGROUP DIFFERENTIATION According to models of category diVerentiation (Doise, 1978; Tajfel & Wilkes, 1963; Turner, 1982), classification of two groups along some continuum leads to an accentuation of the diVerences between categories (an interclass eVect) and an accentuation of the similarities within categories (an intraclass eVect), and it is this dichotomization that forms the basis for diVerential evaluation. Quite simply, divide anything into two categories (from pebbles to people) and there will be a tendency to then see all the members of the same category as more similar to each other, and members of the other category as more diVerent from each other, than they really are. Once you sort the brown pebbles from the gray pebbles, the brown pebbles will seem more diVerent from gray pebbles than they were before. The power of the categorization process has been consistently and reliably demonstrated in a number of domains including judgments of both physical stimuli (Crisp & Hewstone, 1999b; Deschamps, 1977; Tajfel & Wilkes, 1963) and social stimuli (Doise, Deschamps, & Meyer, 1978; Eiser, van der Pligt, & Gossop 1979; McGarty & Penny, 1988; Wilder, 1986). Campbell’s (1956) classic study presented evidence for the general principle of exaggeration of diVerence between categories and minimization of diVerence within categories as a fundamental process in concept formation. Tajfel and Wilkes (1963) demonstrated the diVerentiation process by presenting lines to participants. In one condition, the lines were classified by labeling the four smallest with the letter ‘‘A’’ and the four largest with the letter ‘‘B.’’ Relative to a control condition in which no labels were presented, participants in the classification condition judged the interclass diVerence to be larger than the intraclass diVerences. The eVect of imposing a system of classification on physical stimuli was that, despite the lines increasing at a constant rate, the diVerence between the largest of the ‘‘As’’ and the smallest of the ‘‘Bs’’ was judged as being larger than any other diVerences. When categorizations are perceived as dichotomous along some continuum, they provide the cognitive basis for intergroup bias (Tajfel, 1959; Tajfel & Turner, 1979; Tajfel & Wilkes, 1963; see also Doise, 1978). In Sherif’s classic summer camp studies, formation of two groups led to spontaneous
MULTIPLE SOCIAL CATEGORIZATION
177
social comparisons and the development of group ‘‘icons’’ (Sherif & Sherif, 1953; Sherif, White, & Harvey, 1955). These eVects occurred even without the presence of competition, suggesting mere categorization into two distinct groups had a critical role to play in the emergence of social identities, norms, and intergroup bias (LeVine & Campbell, 1972). Later, Tajfel’s studies using the minimal group paradigm demonstrated that mere categorization (with objectively trivial categories, no face‐to‐face interaction between group members, and no past history of conflict) was suYcient to elicit intergroup bias (Tajfel et al., 1971) and without even a basis in belief similarity (Billig & Tajfel, 1973). The mere categorization eVect on group evaluations is one of the most robust findings in research on intergroup relations (Brewer, 1979; Brewer & Brown, 1998; Messick & Mackie, 1989; Mullen et al., 1992).1 Categorization therefore accentuates the diVerences between two groups along any one dimension. This process is central to our model of multiple categorization: from category diVerentiation follow the cognitive, aVective, and motivational processes that shape our evaluations of groups. The reversablity of the diVerentiation process is critical: If strengthening category salience can accentuate diVerences, then weakening category salience should attenuate diVerences (Doosje, Spears, & Koomen, 1995; Gaertner & Dovidio, 2000). Below we discuss in detail models of multiple categorization and bias reduction that are based on this assumed positive relationship between diVerentiation and discrimination. This relationship will be central to how we construct the first part of our model, but it will not be where we end the model; this relationship is not always positive, motivations associated with diVerentiation and distinctiveness can exert a powerful moderating eVect, and this is something that will become important later on in the development of our model.
III. DiVerentiation–Discrimination In this section we turn to studies of multiple categorization that have focused on changing category representation—that is, the way categories are perceived in terms of structure (e.g., within and between category variability), the way they are evaluated (i.e., being thought of in positive or negative terms), feelings 1
We note that mere categorization here means contexts in which perceivers identify with the categories involved and the groups involved are comparable (Tajfel & Turner, 1979). In practical terms, these are arguably all qualities at least minimally descriptive to even the most basic intercategory dichotomy (although the extent to which they can apply can, of course, vary).
178
RICHARD J. CRISP AND MILES HEWSTONE
toward categories, stereotypes associated with categories, and so on. We will argue that a key underlying process that determines evaluations in multiple category contexts is category diVerentiation. However, we will also argue that the impact of diVerentiation due to salient converging (out‐group) or diverging (crosscutting, in‐group) aYliations varies depending on moderating factors such as categorization context and aVective state. A. CATEGORY DIFFERENTIATION MODELS In the section above we made two assumptions. The first was that people can simultaneously pay attention to multiple dimensions of categorization. The second was that the extent to which categories are perceived as similar or diVerent can vary. This latter assumption is important because of an observed relationship between category diVerentiation and evaluation. Sherif’s (1966) seminal summer camp studies illustrated the relative ease with which dividing people into groups can lead to in‐group favoritism, and this link has been an enduring aspect of work on intergroup relations. Tajfel’s later work with the minimal group paradigm (Tajfel et al., 1971) showed how even with trivial memberships, anonymity, and no prior intergroup contact, when ‘‘they’’ are diVerent from ‘‘us’’ then ‘‘they’’ are also evaluated less positively than ‘‘us’’ (see Brewer, 1979; Mullen et al., 1992). The extent to which in‐groups and out‐groups are diVerentiated and distinct determines how positively such groups are evaluated and this idea has been formalized in various theoretical accounts such as Doise’s (1978) category diVerentiation model and self‐categorization theory’s ‘meta‐contrast ratio’ (Oakes, Haslam, & Turner, 1994; see also Campbell, 1956; Tajfel, 1959; Tajfel & Wilkes, 1963). It is important to note that this relationship only applies when the categories involved are self‐ versus other defining (i.e., the participant is a member of one of the groups versus the other; see also Crisp & Hewstone, 2000a; Schaller & Maass, 1989). Otherwise it is meaningless to talk about in‐group favoritism. Second, the process of category diVerentiation is necessary, but not suYcient, for in‐group favoritism. In order to fully explain in‐group favoritism other processes that accompany the diVerentiation mechanism are needed; these can be both motivational (Hogg, 2000) and cognitive (Cadinu & Rothbart, 1996). However, whatever the accompanying process, bias appears to vary in proportion to category diVerentiation. In sum, these accounts all predict a positive correlation between category diVerentiation and evaluative discrimination: more intergroup diVerentiation initiates more in‐group favoritism. Categorization along multiple criteria has direct implications for category diVerentiation, and as such oVers the potential to weaken category boundaries, to increase intergroup similarity, and to reduce categorical diVerentiation; and
179
MULTIPLE SOCIAL CATEGORIZATION
in so doing also reduce intergroup bias. Initial work on multiple categorization was set within the framework of the category diVerentiation approach. Consistent with Doise’s (1978) argument that the act of (simple) categorization leads to an accentuation of diVerences between and similarities within categories, Deschamps and Doise (1978) suggested that this process would be ‘‘cancelled out’’ when a second category dimension was salient, along which both initial groups shared membership. Put another way, if multiple dimensions of categorization can sometimes be simultaneously salient (assumption 1), then this can aVect the extent to which such categories and their members are perceived to be similar or diVerent (assumption 2). Illustrative of the link between diVerentiation and evaluation is a study by Marcus‐Newhall, Miller, Holtz, and Brewer (1993; Study 1). Marcus‐ Newhall et al. (1993) investigated the eVects of crossing artificial groups and diVerential role assignments on intergroup bias and a composite measure of similarity. When the artificial in‐group and the out‐group shared a role assignment (e.g., a task that involved generating five traits important for dealing with stress), this led to an increase in perceived similarity (reduced category diVerentiation) and less‐biased intergroup reward allocations, and both measures were positively related. Perceived similarity statistically mediated the eVects of crosscutting categorization on reward allocations. This supported the idea that category diVerentiation can help explain the impact of crossed categorization on intergroup bias (for similar findings see Brown & Wade, 1987; Deschamps & Brown, 1983). These studies provided an early indication that category diVerentiation can vary with the introduction of crosscutting or convergent bases for categorization and that there is a positive relationship between diVerentiation and evaluation: increased diVerentiation is associated with increased bias, reduced diVerentiation is associated with reduced bias. Two established models exemplify how this basic link between category diVerentiation and evaluation has been used as the basis for reducing intergroup bias by blurring intergroup boundaries: the common in‐group identity and crossed categorization models of bias reduction. We discuss each model in turn. B. THE COMMON IN‐GROUP IDENTITY MODEL I against my brother I and my brother against my cousin I, my brother and our cousin against the foreigner Bedouin proverb
There are no Hutus or Tutsis any more, only Rwandans. The Economist, March 27, 2004
180
RICHARD J. CRISP AND MILES HEWSTONE
The common in‐group identity model (Gaertner & Dovidio, 2000; Gaertner, Dovidio, Anastasio, Bachman, & Rust, 1993; Gaertner, Mann, Dovidio, Murrell, & Pomare, 1990; Gaertner, Mann, Murrell, & Dovidio, 1989) rests on the idea that by changing the nature of categorical representation from ‘‘us’’ and ‘‘them’’ to a more inclusive ‘‘we’’ it is possible to reduce intergroup bias. There are a number of ways that such recategorization can occur. For instance, cooperative interdependence (i.e., common task or fate; see Brown & Abrams, 1986; Brown & Wade, 1987; Gaertner et al., 1993; Sherif, 1966; Worchel, 1979) can enhance evaluations of former out‐group members, and this is attributable to intergroup cooperation transforming members’ cognitive representations from two groups to one inclusive group. Goal relations and group interactions also aVect intergroup attitudes, apparently by altering the salience of relevant category representations (Doise, 1978; Turner, 1982; Worchel, 1979). Sherif (1966) suggested that attitudes and behaviors are mainly determined by the nature of the objective goals linking groups, and varying the nature of the interdependence appears to correspondingly moderate intergroup relations (see Blake, Shepard, & Mouton, 1964; Rabbie & de Brey, 1971; Sherif, Harvey, White, Hood, & Sherif, 1961). The notion that there is a positive correlation between category diVerentiation and evaluation is thus central to the common in‐group identity model: factors that promote recategorization bringing groups together, fostering mutual similarities, and blurring extant subgroup boundaries can all reduce intergroup bias. Recategorization reduces bias by increasing the attractiveness of former out‐group members, once they are included within a superordinate category. Thus, when perceivers in an initially dichotomous intergroup context are encouraged to form a common in‐group identity (i.e., representing members of both in‐groups and out‐groups as members of one overall superordinate group), in‐group favoring bias at the subgroup level will subside. According to the model the attractiveness of former out‐group members can be enhanced by factors such as those outlined above via transformation of members’ cognitive representations of group boundaries. Gaertner et al. (1989, 1990, 1993) proposed that the specific processes leading to formation of a superordinate identity (whether by cooperation or otherwise) are based on the fundamental category accentuation and assimilation eVects that we have outlined earlier in this chapter. When intergroup similarity weakens the intergroup boundary (or weakening the intergroup boundary leads to greater perceived similarity) the two groups will be perceived as a single superordinate entity. Factors that bring groups together (such as engagement in cooperative tasks) help to blur subgroup boundaries, promoting recategorization at a higher, more inclusive, level of categorization. Positive attitudes associated with the in‐group (rather than out‐group devaluation; Brewer, 1979; see also Mullen et al., 1992) can extend toward former out‐group members. This is because these previous out‐groupers have moved closer to
181
MULTIPLE SOCIAL CATEGORIZATION
the self and can be regarded as in‐group members by virtue of the inclusive common in‐group (and thus receive the benefits of in‐group favoritism). There have been numerous demonstrations of the benefits of recategorization. Worchel, Andreoli, and Folger (1977) demonstrated that groups wearing diVerent‐colored laboratory coats during intergroup interaction were more biased toward each other compared with groups dressed in same‐ colored coats. Gaertner and Dovidio (1986) and Gaertner et al. (1989, 1993) reported that integrated seating patterns (ABABAB) reduced bias relative to segregated seating (AAABBB), and promoted the perception of an inclusive single group (there was less in‐group bias in leadership votes, and lower bias on a liking measure). Gaertner et al. (1990) reduced bias by systematically varying participants’ formal identity during interaction (in the one‐group condition participants created a new single name for their group, while those in the two‐group condition maintained their previous group names). Positive mood can also reduce bias via the formation of a (inclusive) common in‐group representation (Dovidio, Gaertner, Isen, Rust, & Guerra, 1998a; Dovidio, Gaertner, Isen, & Lowrance, 1995), and a common in‐group identity has been found to facilitate subgroup relations in a variety of settings including corporate mergers (Dovidio, Gaertner, & Bachman, 2001), multiethnic high schools (Houlette et al., 2004), and stepfamilies (Banker & Gaertner, 1998). Confirming the role of categorical diVerentiation, Gaertner and colleagues (Gaertner et al., 1989, 1990) found evidence for a mediational path that explains bias reduction in common in‐group contexts via decreased diVerentiation (versus individuation) (see also Marcus‐Newhall et al., 1993). C. THE CROSSED CATEGORIZATION MODEL All good people agree and all good people say All nice people like us are we and everyone else is they: But if you cross over the sea, instead of over the way You may end up by (think of it!) looking on we as only a sort of they. Rudyard Kipling
In the crossed categorization paradigm two dimensions are made simultaneously salient for participants making group‐relevant social judgments (Brewer et al., 1987; Crisp & Hewstone, 2000a, 2001; Ensari & Miller, 1998; Hewstone et al., 1993; Miller, Urban, & Vanman, 1998; Vanbeselaere, 1987, 1991; Vescio, Hewstone, Crisp, & Rubin, 1999; for a theoretical review/ integration see Crisp et al., 2002; for a literature review Crisp & Hewstone, 1999a; for meta‐analyses Migdal et al., 1998; Urban & Miller, 1998). Take, for
182
RICHARD J. CRISP AND MILES HEWSTONE
Female (in-group)
Male (out-group)
Young (in-group)
Old (out-group)
Young (in-group)
Old (out-group)
Young female (in-group/ in-group)
Old female (out-group/ in-group)
Young male (in-group/ out-group)
Old male (out-group/ out-group)
Fig. 2. The four combined categories formed from crossing gender and age dimensions of categorization.
example, gender and age; instead of considering only females versus males or young versus old, in crossed categorization contexts perceivers attend to both of these dimensions (see Fig. 2). Then females and males can be seen to share a common category; both females and males can be young, or both females and males can be old. Deschamps and Doise (1978) argued that such crisscrossing of categorizations should lead category diVerentiation processes to work against each other. DiVerentiation on one dimension will be counteracted by assimilation on a second. Corresponding to this weakening of category distinctiveness, evaluations between composite groups that share an in‐group identity should become less diVerentiated, and bias should be reduced; this prediction has been supported by much empirical work on explicit evaluations (Crisp et al., 2001; Deschamps & Doise, 1978; Diehl, 1990; Vanbeselaere, 1987, 1991, 1996) and on implicit evaluations (Crisp et al., 2006b) and measures of stereotype threat (Rosenthal & Crisp, 2006; Rosenthal, Crisp, & Suen, in press). Work on crossed categorization clearly overlaps with that described in the previous section on the common in‐group identity model. It examines the eVects on intergroup evaluations when a shared (common) identity crosscuts an initial intergroup distinction. In Fig. 2, for instance, the first three columns of the crossed categorization design (young females versus old females or young males) can be taken as illustrative of designs used in common in‐group identity research. With respect to young females versus old females, ‘‘female’’ here is a common in‐group identity shared between young and old people. With respect to young females versus young males, ‘‘young’’ is a common in‐ group identity shared between females and males. Research that has compared evaluations among these cells in crossed categorization designs has provided strong support that the shared identity does elicit reduced intergroup bias (Crisp & Hewstone, 1999a; for meta‐analyses Migdal et al., 1998; Urban & Miller, 1998; although see Mullen, Migdal, & Hewstone, 2001, for
MULTIPLE SOCIAL CATEGORIZATION
183
mixed support, depending on the definition of in‐group bias adopted) in the same way as has been observed in research on the common in‐group identity model. Despite research on both models finding similar bias‐reducing eVects, it should be noted that research on the two models has diVered in emphasis. First, the original formulation of the common in‐group identity model focused on recategorization (that is a weakening of the salience of the initial in‐group versus out‐group identities in favor of the new common in‐group identity). Later work has, however, proposed a dual identity formulation in which the salience of the initial identities is retained alongside the common identity (Gaertner & Dovidio, 2000; but see also Gaertner et al., 1989, for some earlier discussion of these ideas). This reformulation, which we discuss in more detail later in this chapter, means that the common in‐group identity model and the crossed categorization model can be equated in terms of their emphasis on simultaneous identity salience. Second, the two models have also typically focused on, respectively, nested identities (where the shared category is superordinate to the initial in‐group/out‐group categories; e.g., British and European) versus crosscutting categories that are not more inclusive than the target dichotomy (e.g., gender and race). We (Crisp, Beck, & Hewstone, 2006a) compared paradigms used to test the common in‐group identity model and crossed categorization eVects by contrasting evaluations in a baseline condition with those in a condition in which the initial intergroup dichotomy (in‐group versus out‐group; science versus humanities students) was crosscut by a second shared identity (University of Birmingham). We measured the perceived inclusiveness of the shared identity using the item, ‘‘How much more inclusive do you think grouping people by their university is than grouping them by their subject [major]?’’ on a Likert‐type scale (1–9) and then split participants into two equal groups according to the scale median. We found that whether participants perceived this shared identity as more inclusive (i.e., a superordinate group, as typically used in common in‐group identity research) or at the same level of inclusiveness (i.e., a crosscutting group as typically used in crossed categorization research) had no impact on the bias‐ reducing eVectiveness of the intervention (see Fig. 3). For our purposes we therefore consider crosscutting categories as having the same diVerentiation‐ reducing qualities irrespective of their inclusiveness. Third, the common in‐group identity model can be construed as a kind of shifting identity model such as that discussed earlier. For instance, rather than creating a context in which subgroup and superordinate identities are both salient, a recategorization process occurs whereby perceivers come to categorize at a higher, more inclusive level in the conceptual hierarchy that includes former out‐groupers (e.g., from British—French to European,
184
RICHARD J. CRISP AND MILES HEWSTONE
In-group
Out-group
8.0 7.5 7.0
6.7
Evaluation
6.3
6.27
6.5 6.0
5.41
5.5 5.0
5.4
4.7
4.5 4.0 3.5 3.0 Baseline
Crosscutting
Common in-group
Categorization context Fig. 3. The bias‐reducing eVects of crosscutting and common in‐group (superordinate) categorizations on intergroup evaluations. Data from Crisp et al. (2006a; Experiment 3).
see Fig. 1). However, we can argue that such recategorization constitutes a context in which both original and common identities do remain salient to some degree. It may simply be that the superordinate identity is, in relative terms, more salient than the original subgroups. Given that the original subgroups remain (necessarily) the targets of the judgment measure (typically, evaluation), this cannot really be otherwise. Consistent with our general argument, that it is relative salience of original and crosscutting categories that is important (see Dovidio et al., 2006, for a similar point), some research has found that varying the salience of subgroup and superordinate identities does have predictable eVects on evaluation (see Hornsey & Hogg, 2000). For now, however, it is just important to note that while the emphasis on salience has been slightly diVerent in research on common in‐group identity and crossed categorization, for our purposes both operate according to the same diVerentiation‐based mechanism. To summarize, there is much evidence that blurring intergroup boundaries via multiple categorization can reduce intergroup bias, evidence that can be found both from research on the common in‐group identity model and the crossed categorization model. However, the above models present a rather simple conceptualization of multiple categorization, where in‐groups and out‐ groups can, under some circumstances, share an additional basis for categorization. This relationship between the salience of crosscutting categorization and category diVerentiation is illustrated in Fig. 4. We indicate crosscutting categorization here using two notations: ‘‘I’’ for in‐group aYliation, ‘‘O’’ for out‐group aYliation. Therefore, compared to a simple intergroup dichotomy
185
MULTIPLE SOCIAL CATEGORIZATION
Crosscutting categories (II versus IO)
−
Intergroup differentiation
+
Intergroup bias
Fig. 4. The relationship between divergent (crosscutting or common in‐group) categorization, diVerentiation, and bias.
(I versus O), the eVects of making an additional common in‐group salient can be indicated by II versus IO. Figure 4 summarizes the findings of the research we have reviewed so far. To the extent that multiple categorization creates a crosscutting and shared common in‐group categorization (i.e., that diverges from the basic in‐group‐out‐group dichotomy), and that the common in‐ group is equally or more salient than the original dichotomy, diVerentiation will be reduced (more salient crosscutting bases for categorization, less diVerentiation, a negative relationship). This relationship is marked with a ‘‘–’’ in Fig. 4. As diVerentiation is reduced, consistent with the category diVerentiation mechanism outlined by Doise (1978) and others, bias will correspondingly be reduced (reflecting the typically observed positive correlation between diVerentiation and bias, indicated by a ‘‘þ’’ sign in Fig. 4). This model captures the basic findings observed in research on both the common in‐group identity and the crossed categorization model. The model outlined in Fig. 4 is well established. We know, however, that multiple identities can also create convergent as well as divergent bases for group aYliation. Put another way, although the focus on shared categories is common to both the common in‐group identity and crossed categorization models, as we discussed above (and illustrated in Fig. 2), work on crossed categorization has also examined what happens when multiple categories reinforce existing boundaries. This is the case for many instances of ethnopolitical conflict. For example, in Belfast, Northern Ireland, Catholics and Protestants tend to live in diVerent places (e.g., Ardoyne versus Shankhill Road), espouse diVerent politics (Nationalist‐Republican versus Unionist‐ Loyalist), and even support diVerent football teams (e.g., Cliftonville versus Linfield); a part of the problem is precisely that there are few social categories crosscutting the religion dimension (Cairns & Mercer, 1984). Multiple converging bases for categorization can provide a stronger basis for evaluative diVerentiation (i.e., II versus OO). In Fig. 2, for instance, this would be a comparison between column 1, young females, and column 4, old males. To return to another example, a white female may not feel quite so diVerent from a black female, but may feel considerably more diVerent from a black male when gender is salient alongside race. In the latter case the salience of an additional criterion for classification has reinforced the distinctiveness of the initial dichotomy; in the former case thinking about
186
RICHARD J. CRISP AND MILES HEWSTONE
additional classifications has weakened the ‘‘us’’ versus ‘‘them’’ distinction. As such, the crossed categorization model oVers the potential to provide support for both the diVerentiation‐reducing and diVerentiation‐accentuating aspects of the supposed positive linear relationship between diVerentiation and discrimination. The way that converging versus crosscutting categorization can impact on both category diVerentiation and intergroup bias can be illustrated by a study we carried out using artificial groups (Crisp et al., 2001, Experiment 1). The groups were defined by two category dimensions, A–B and X–Y. All participants were in group A and group X (group AX was the ‘‘double in‐group’’). Evaluations were measured in the form of points allocated to members of groups AX versus BX (crosscutting condition; II versus OI) and members of groups AX versus BY (convergent categorization condition, II versus OO). Similarity was measured on a 100‐mm line (lower similarity denotes higher diVerentiation). The findings are illustrated in Fig. 5. Consistent with the idea that category diVerentiation provides a basis for category evaluation, both intergroup bias (Fig. 5A) and category diVerentiation (Fig. 5B) were higher (i.e., lower intergroup similarity) in the convergent compared to the crosscutting condition. Similar findings have been observed by Diehl (1990) and Vanbeselaere (1991, 1996). We can therefore specify in more detail the model we first outlined in Fig. 4. Category diVerentiation is positively related to intergroup bias, crosscutting categorization is negatively related to category diVerentiation, and convergent categorization is positively related to category diVerentiation (see Fig. 6). Thus, while increasing the salience of a crosscutting or common in‐group categorization reduces
A
B 99.5
80 60 38.7
40 20 II versus OO (Convergent)
II versus IO (Crosscutting)
Categorization context
Intergroup similarity
Intergroup bias
100
6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0
4.88
2.88
II versus OO (Convergent)
II versus IO (Crosscutting)
Categorization context
Fig. 5. The eVects of convergent and crosscutting categorizations on panel (A) intergroup bias (in‐group minus out‐group point allocations); panel (B) intergroup similarity (0–100 mm). Note: More similarity indicates less diVerentiation. Data from Crisp et al. (2001; Experiment 1).
187
MULTIPLE SOCIAL CATEGORIZATION
Convergent categories (II versus OO)
+ −
Intergroup differentiation
+
Intergroup bias
Crosscutting categories (II versus IO) Fig. 6. The relationship between crosscutting and convergent categorizations, diVerentiation, and bias.
category diVerentiation, and in turn reduces intergroup bias, increasing the salience of converging out‐group categorizations (II versus OO) increases category diVerentiation, which therefore increases intergroup bias (all relative to simple categorization contexts). This basic model has received considerable support from further studies which we review below. All of these studies are important because they show that there are some conditions under which the salience of crosscutting or convergent bases for categorization can be emphasized, creating complex, but definable, evaluative patterns. In the following section, we review studies that have tested what happens when there are both convergent and crosscutting bases for categorization, and, importantly, how these convergent and crosscutting bases for categorization lead to diVerent eVects depending on observed moderating conditions. The moderating conditions provide further specification to the developing model outlined in Fig. 6.
D. PATTERNS OF EVALUATION As discussed above, work on crossed categorization has considered not only crosscutting but also converging bases for categorization. For instance, crossing gender and ethnicity leads to four possible crossed category targets: black females, black males, white females, and white males. The convergent or divergent bases for intergroup similarity and category diVerentiation implied by considering complex category structures defined in this way have been the focus the work we now review. A number of distinct patterns of evaluation have been observed across the four composite groups formed from crossing two dimensions of categorization (Brewer et al., 1987; Hewstone et al., 1993). Six patterns have been identified and are listed in Table I. These patterns are valuable because they indicate the variety of ways in which convergent and crosscutting bases for categorization can form the basis for intergroup evaluation. By focusing on convergent as well as crosscutting bases for aYliation, a full test of the role of
PATTERNS
OF
TABLE I EVALUATION ACROSS CROSSED CATEGORY GROUPS Dimension 1: In‐group
Conditions
Pattern
Dimension 2: In‐group
Dimension 2: Out‐group
Dimension 2: In‐group
Dimension 2: Out‐group
2
0
0
2
Baseline
1. Additive
DiVerential importance (here dimension 1 more important)
2. Dominance
Positive aVect/in‐group prime
3. Social inclusion
ii
4. Social exclusion
ii
ii
Negative aVect/out‐group prime
Dimension 1: Out‐group
>
1 ii
io
¼
¼
1
io
>
1 ¼
io
>
io
>
oi
¼
oi
¼
oi
oo 3
>
1 ¼
oo 1
1
1
3
oi 1
1
oo 1
¼
2
oo 2
Positive aVect/diVerential importance (here dimension 1 more important)
5. Hierarchical acceptance
Negative aVect/diVerential importance (here dimension 1 more important)
6. Hierarchical rejection
ii
¼
io
>
oi
>
oo
Extreme positive aVect/low importance groups/absence of salience categorizations
7. Equivalence
ii
¼
io
¼
oi
¼
oo
4 ii
0 >
2
io
>
2
oi
¼
oo 4
0
Note: i, in‐group constituent of group; o, out‐group constituent of group. DiVerences in sign (þ/) specify where diVerences are to be expected across the four crossed category subgroups. No contrast weights can be shown for the equivalence pattern because it is a null eVect.
MULTIPLE SOCIAL CATEGORIZATION
189
diVerentiation can be carried out: while crosscutting bases for group membership should weaken category diVerentiation (an out‐group member can simultaneously be an in‐group member), convergent bases should strengthen category diVerentiation (such as when an out‐group member on one dimension is also diVerent from the perceiver on a second salient category dimension). Importantly, the variety of observed patterns illustrates that there is not a simple relationship between salient multiple aYliations and category diVerentiation (see work by Urada, Stenstrom, & Miller, in press, for an extended discussion of the diVerent feature integration strategies used in crossed categorization contexts). The additive pattern (pattern 1) documents the predicted outcome across the four crossed categorization groups when categories are of equal salience and are combined in a purely additive fashion. Thus, the double in‐group is evaluated most positively (being in‐group on both dimensions), the double out‐group least positively (being out‐group on both dimensions) and mixed in‐group/out‐group target persons fall in‐between these two extremes. The category dominance pattern (pattern 2) describes the case where a single categorization dominates and classification based on the subordinate category distinction is ignored. Examples of where this could apply are abundant, for example, occupation is largely irrelevant at a football match whereas team allegiance is clearly of paramount importance. These patterns represent simple extensions of the basic category diVerentiation model that we have discussed earlier: the more memberships a target shares with the perceiver, the more positively they are evaluated. Or, to put it another way, the more diVerent a target is, the less positively they are evaluated. This is the typical positive relationship between diVerentiation and discrimination. The category dominance pattern illustrates how salience or importance play a critical role here—the presence of convergent or divergent category labels in any given context will only have an impact on diVerentiation to the extent that they are salient or important (see Crisp et al., 2006c; Urada & Miller, 2000). Other patterns are more complex and can be thought of as involving an interactive combination of category dimensions. Social inclusion (pattern 3) describes a situation in which all groups are evaluated equally positively as long as they share membership on at least one dimension. Double out‐ groups are therefore evaluated less positively than all other groups. The social exclusion pattern (pattern 4) is the opposite of social inclusion. Here, mixed targets are evaluated as negatively as double out‐groups because they are an out‐group on at least one dimension. Hierarchical ordering, pattern 5, specifies that the eVects of one category distinction are dependent on prior categorization on the other dimension, that is in‐group (versus out‐group) classification on the first dimension will lead to greater diVerentiation on a second dimension. Hierarchical rejection, pattern 6, is a variant of this
190
RICHARD J. CRISP AND MILES HEWSTONE
hierarchical ordering pattern in that out‐group membership on the more important dimension determines diVerentiation (e.g., white supporters only derogating the black players on an opposing football team). A final pattern, equivalence, represents simply the absence of any evaluative diVerentiation across the subgroups (all subgroups are evaluated equally positively). The evidence for each of these models has been rigorously investigated in meta‐analytic (Migdal et al., 1998; Urban & Miller, 1998) and narrative (Crisp & Hewstone, 1999a) reviews. All three reviews found (among other things) that the most typically observed pattern, all other things being equal, was additive (pattern 1). For instance, of the 15 relevant studies reviewed by Crisp and Hewstone (1999a), 9 yielded support for the additive pattern. All three reviews suggested that in the majority of contexts, and across diVerent types of target groups, measures, and participants, the pattern which would most often be expected is the additive pattern where the double in‐group is evaluated most positively, the double out‐group most negatively, with the mixed groups in‐between these two extremes. This seems reasonable, based on our discussion above of basic category diVerentiation eVects. DiVerentiation between composite groups, as an additive and linear function of categorization, is a cognitive mechanism that occurs pervasively and persistently in intergroup perception. There was, however, some support for other patterns in situations that appeared to feature specifiable moderators. We observed that while the additive pattern seemed to prevail, other patterns appeared more likely in particular contexts (Crisp & Hewstone, 1999a). For instance, the social inclusion pattern was observed in 5 of the 15 studies reviewed. The variety of observed models suggests that there is not always a simple relationship between aYliation and evaluation, but that there may be varied contextual factors that strengthen or weaken the diVerentiation‐implications of category memberships in crossed category contexts. Across a subsample of studies, Urban and Miller (1998) found that mean eVect size estimates overwhelmingly supported an additive pattern. Notably, significant residual variation remained after removal of the additivity eVect (see also Migdal et al., 1998). This suggested that the variation among the observed patterns reflected the eVects of moderating variables. We proposed a process model of pattern moderation in crossed categorization contexts (Crisp et al., 2002). Here we discuss support for this model, and present an elaborated and refined account based on more recent empirical observations.
E. AFFECT AND COGNITION Urban and Miller’s (1998) meta‐analysis revealed several conditions that reliably and independently predicted deviation from the additive pattern. Combined with findings from our own reviews and theorizing (Crisp &
MULTIPLE SOCIAL CATEGORIZATION
191
Hewstone, 1999a; Migdal et al., 1998, see also Urban & Miller, 1998), as well as subsequent empirical findings, we (Crisp et al., 2002) identified two types of moderator of the pattern of evaluation expected in crossed categorization contexts, one aVective in nature and one cognitive. We outline this theoretical model and the evidence for it below. 1. AVective State Urban and Miller’s (1998) meta‐analysis revealed a possible role for positive and negative aVect in predicting patterns of crossed categorization. We (Crisp et al., 2002) elaborated on Urban and Miller’s initial theorizing on this relationship and proposed that aVect may exert its influence over crossed category evaluations via a priming mechanism. The reasoning goes as follows. In‐group and out‐group memberships carry an acquired aVective valence (Fazio, Sanbonmatsu, Powell, & Kardes, 1986; Perdue, Dovidio, Gurtman, & Tyler, 1990). In‐group categories are inherently positive in connotation, out‐group categories inherently negative.2 This is an association developed via associative learning over time (Bargh, 1997; see also Cacioppo, Marshall‐ Goodell, Tassinary, & Petty, 1992; Das & Nanda, 1963; Staats & Staats, 1958; Tajfel & Turner, 1979). The inherent positive or negative nature of in‐groups and out‐groups is important because of what we know of the impact of aVect on categorization. Positive aVect causes broader, more inclusive categorization than negative aVect (Isen & Daubman, 1984). Positive incidental aVect also increases the extent to which neutral or positive targets are associated with a positive categorization (Isen, Niedenthal, & Cantor, 1992). It cues positive material in memory (Isen, Shalker, Clark, & Karp, 1978; Teasdale & Russell, 1983), and directs attention primarily to other positive aspects in the environment (Higgins & King, 1981). Other research has shown similar eVects with regard to decision making (Isen & Means, 1983) and persuasion (Mackie & Worth, 1989). Expanding on the idea that aVect cues‐related material in memory, we can argue that activation of positive and negative aVect in contexts where judgments of in‐group and out‐group members are required might be expected to influence those judgments (see also Miller, Kenworthy, Carnales, & Stenstrom, 2006). Positive aVect will prime in‐group categories, leading to less diVerentiation due to in‐group components of crossed category 2 We note here that the associative link between in‐groups and positivity is typically stronger than between out‐groups and negativity (see Perdue et al., 1990). However, although less reliably demonstrated, the out‐group‐negative association does appear to exist in some form, and in the studies we report below we observe eVects consistent with this.
192
RICHARD J. CRISP AND MILES HEWSTONE
groups—a basis for perceiving similarity—being more powerful in determining evaluation. Put another way, positive aVect will make crosscutting shared in‐group identities more salient. Consistent with the work we have discussed previously, anything that increases the relative salience of a crosscutting categorization over an original intergroup dichotomy will have a greater diVerentiation‐reducing eVect. In contrast, negative aVect will prime out‐group categories, leading to a greater focus on out‐group constituents of crossed category composites, and greater diVerentiation from any target that possesses an out‐group constituent. Negative aVect will make convergent nonshared out‐group identities more salient, and correspondingly heighten the diVerentiation‐accentuating impact they have. The simple prediction here is that relative to the baseline additive pattern (maximal diVerentiation, ii > io ¼ oi > oo), positive mood will produce a social inclusion pattern of evaluation (ii ¼ io ¼ oi > oo), while negative mood will produce a social exclusion pattern of evaluation (ii > io ¼ oi ¼ oo), see Table I. Kenworthy, Canales, Weaver, and Miller (2003) manipulated negative incidental aVective state to examine its eVects on crossed category evaluations (incidental aVect refers to aVect that is generalized across the context, that is, not tied to a particular target; a distinction that we will return to below). In a first study, they manipulated negative mood by requiring participants to carry out a very diYcult anagram task during which they were repeatedly interrupted. Self‐reported mood confirmed the success of this manipulation. Following this manipulation, crossed category groups were evaluated. Relative to a baseline additive pattern, negative mood produced the predicted social exclusion pattern (ii > io ¼ oi ¼ oo). Consistent with the aVective priming explanation outlined above, negative aVect appeared to strengthen the influence of the out‐group constituents of the crossed category groups, shifting them toward the evaluation of the double out‐group. In further studies Kenworthy et al. (2003) showed that more specific negative moods such as sadness and anger, manipulated in a range of ways (from presentation of an aVect‐inducing short film through to autobiographical recall of negative life events), also produced the social exclusion pattern. Reported negative mood mediated the observed shift in evaluations. In a similar study we tested the moderating eVect of positive incidental aVect by manipulating performance feedback in an experiment involving crossed category subgroups defined by minimal categories (i.e., overestimators and underestimators; better at estimating small or large shapes; Crisp & Hewstone, 2000a). On the basis of previous work, we hypothesized that positive aVect would create a generalized sense of inclusiveness, reducing the extent to which the crossed category subgroups were diVerentiated from one another (for a similar study in the common in‐group identity literature
MULTIPLE SOCIAL CATEGORIZATION
193
see Dovidio, Gaertner, & Validzic, 1998b). In our study participants received either positive feedback (i.e., ‘‘You have scored 84 out of 100’’) or neutral feedback (i.e., ‘‘You have scored 51 out of 100’’) on a memory task. In the positive aVect condition, the pattern of evaluations across the four crossed categorization targets shifted from the additive pattern in the no feedback (neutral aVect) condition to a more socially inclusive pattern. In fact this shift went beyond the social inclusion pattern shown in Table I such that diVerentiation between all the crossed category groups was reduced to zero, an equivalence pattern. This pattern, while clearly representing increased inclusion, is not entirely consistent with the priming mechanism outlined above (how can the aVective congruence hypothesis explain an apparent impact on the double out‐group, which includes no in‐ group constituents?). Miller et al. (2006) have suggested that the extremely high level of inclusion observed in Crisp & Hewstone’s (2000a) study is due to the minimal nature of the groups used. Specifically, they argue that at very low levels of category importance positive aVect will override any inherent evaluative connotations associated with group memberships, leading to total inclusion. They extend this argument to suggest that category importance, in fact, interacts with aVective state. They propose that category importance increases the accessibility, and salience, of particular category dimensions and that more accessible categories become therefore more available for aVective ‘‘tagging.’’ There are several studies that support this idea. Urada and Miller (2000) tested whether positive mood would exert a priming eVect on in‐group constituents particularly for relatively more important categories. They induced positive aVect by asking participants to recall a happy life experience. Across four studies, incidental positive aVect increased preferences for mixed targets over double out‐group targets but only if they were categorized on dimensions that participants regarded as important. This produced a hierarchical acceptance pattern (see Table I) whereby only the mixed group that included an in‐group constituent on the important dimension was evaluated more positively (Ii ¼ Io > Oi > Oo, where uppercase letters indicate the dimension of higher importance). Importantly, this pattern remains consistent with the aVective priming explanation outlined above. Positive aVect primes in‐group constituents of crossed category groups, leading to their exerting a stronger eVect on subsequent evaluations (i.e., increasing positive evaluation), especially when the category is important to start with. These studies therefore support the basic mechanism we have outlined previously (Crisp et al., 2002) that positive aVect strengthens the impact of shared aYliation on diVerentiation and correspondingly weakens intergroup bias. Some further research supports the idea that importance can make crossed category constituents more or less sensitive to the moderating impact of positive mood. The studies we have discussed above all involved positive
194
RICHARD J. CRISP AND MILES HEWSTONE
or negative incidental aVect. Bodenhausen (1993) distinguished between incidental and integral aVect. Incidental aVect is external to the group context such as participants being oVered candy as a manipulation of positive aVect prior to categorization. Integral aVect is specific to an intergroup context and tied to a particular group such as performance feedback or compliments/ insults to an in‐group member from an out‐group member. Ensari and Miller (1998) examined the moderating impact of integral aVect on the pattern of crossed category evaluations observed. When positive integral aVect was induced, with compliments from the target out‐group category, the diVerence between the mixed and double in‐group target disappeared, but this only applied for the mixed category that contained the out‐group source of the compliment (i.e., if the compliment had come from member O, who was also an in‐group member simultaneously, iO). Positive integral aVect therefore improved evaluations, but only for crossed category groups that included the constituent out‐group member who had paid the compliment, producing a hierarchical rejection pattern (ii ¼ iO > oi > oo, where the uppercase ‘‘O’’ indicates the source of the positive aVect). Evaluations decreased when a negative mood was induced by an insult from an out‐ group that constituted part of the mixed group, leading to a hierarchal acceptance pattern (ii > io > Oi ¼ oo). If we think of the source of the compliment/insult in this research as increasing the relevance (and therefore, contextually, the importance) of particular categories, then these results can be seen as consistent with the research reviewed earlier that focused on incidental aVect. Positive mood enhanced the evaluation of the composite groups that included the source of the compliment; negative mood decreased the evaluation of the composite group that included the source of the insult. These findings again support the aVective priming explanation: positive mood strengthened the bias‐reducing impact of a common in‐group identity (positive aVect: ii ¼ iO), while negative mood strengthened the bias‐inducing impact of a nonshared out‐group identity (negative aVect: Oi ¼ oo). Collectively, these studies have shown that positive or negative aVect can weaken or strengthen the impact of shared aYliation on evaluations (for a detailed review of the impact of aVect on crossed categorization see Miller et al., 2006). Relating these findings to the patterns outlined in Table I, compared to the baseline additive pattern, diVerential importance will lead to category dominance, positive aVect to social inclusion (Crisp & Hewstone, 2000a), and negative aVect to social exclusion (Kenworthy et al., 2003); positive aVect and diVerential importance will lead to hierarchical rejection (Ensari & Miller, 1998); and negative aVect and diVerential importance will lead to hierarchical acceptance (Ensari & Miller, 1998; Urada & Miller, 2000).
MULTIPLE SOCIAL CATEGORIZATION
195
Crisp et al. (2002) identified not only aVective, but also cognitive moderators of crossed category evaluations. Specifically, along with positive and negative aVect, category primes (in‐group or out‐group) can also exert a significant impact on the pattern of evaluations observed in crossed category contexts. In the section below, we review evidence for category primes as moderators of crossed categorization evaluations, then outline a qualified priming mechanism that builds on our previously specified model (Crisp et al., 2002). 2. Category Primes If making a common in‐group context salient reduces diVerentiation between constituent categories (Gaertner & Dovidio, 2000), then how would this basic prediction translate in multiple category contexts? As we have discussed throughout this chapter, imposing a common in‐group categorization on an existing intergroup context reduces diVerentiation between the original in‐group and out‐group (see Gaertner et al., 1989, 1993). What happens to intragroup diVerentiation, therefore, when the existing context is already defined by two orthogonally crossed dimensions of social categorization? For instance, how would British Muslims feel about French Muslims when European as a common identity is salient? On the basis of what we know about category diVerentiation in simple and common in‐group contexts (Deschamps & Doise, 1978; Gaertner & Dovidio, 2000), Crisp et al. (2002) predicted reduced diVerentiation across the crossed category targets when a further inclusive common identity is made salient. This represents a shift along a continuum of inclusion, from the additive (maximal diVerentiation) to social inclusion (intermediate diVerentiation) to equivalence (minimal diVerentiation). We recently tested the idea that adding a further layer of categorization, a third dimension within which all multiple category members share at least one in‐group membership, will lead to social inclusion (see Urada et al., in press, who also went beyond the ‘‘two group’’ crossed categorization model). In this paradigm, we first made two dimensions of categorization salient (e.g., A–B and X–Y) and asked participants to evaluate these groups either with or without a further more inclusive common in‐group identity (i.e., all ‘‘University students’’). In three experiments, using both minimal and real groups, we (Crisp et al., 2006c) showed that priming a (common) in‐group context consistently led to a shift from the additive pattern of maximal diVerentiation (double in‐groups evaluated more positively than mixed groups who are evaluated more positively than double out‐groups) to a less diVerentiated social inclusion pattern (double in‐groups and mixed groups evaluated equally
196
RICHARD J. CRISP AND MILES HEWSTONE
and more positively than double out‐groups) (see Fig. 7).3 Consistent with the further moderating role of importance (Miller et al., 2006), the salient common in‐group identity had more or less of an impact depending on the importance of the crossed category groups. When the crossed subgroups were all of lower importance, a common in‐group identity led to an equivalence pattern (i.e., total inclusion, mirroring the eVects observed by Crisp & Hewstone, 2000b, using minimal groups under positive incidental aVect). For moderately important crossed subgroups the social inclusion pattern was observed, while for highly important subgroups the additive pattern remained resolute (see Rust, 1997, cited in Dovidio et al., 2006, for similar findings, and Urada & Miller, 2000, for identical eVects as a function of importance under positive incidental aVect instead of a common in‐group identity).
Double in-group
Mixed group
Double out-group
6 5
4.94 3.88
Evaluation
4 3
3.06
2.82
2 0.941
1 0
−0.118
−1 Baseline
In-group prime
Categorization context Fig. 7. The impact of an in‐group prime on evaluative diVerentiation between crossed category groups. Data from Crisp et al. (2006c; Experiment 1). Note that for illustrative purposes mixed groups (io and oi) have been combined.
3 It is worth noting here that the shift toward the social inclusion pattern in the common in‐group prime condition appears to be driven by a decrease in double in‐group evaluation. However, we do not believe this is inconsistent with our priming explanation, because in this study we used a within‐subjects design, and participants appeared to make their judgments globally rather than in a piecemeal way, constructing their impression of all groups in the context relative to each other.
MULTIPLE SOCIAL CATEGORIZATION
197
Crisp et al. (2006c)’s findings suggest that a context that favors assimilation will have a general diVerentiation‐reducing eVect between any subgroup included within this context. We previously argued that this diVerentiation mechanism was the same as we have described earlier in this chapter relating to the blurring of intergroup boundaries, and distinct from the aVective priming mechanism outlined in the section above (Crisp et al., 2002). However, converging evidence that we discuss below suggests that a more parsimonious explanation can account for both aVective and cognitive moderating eVects. Rather than aVective priming and category context exerting their eVects via separate routes, we here argue that they are diVerent antecedents that exert their influence via the same general priming mechanism. Both types of antecedent can increase the salience of relevant constituents of composite, crossed category, groups. While positive mood increases the salience of in‐group constituents, because both are aVectively positive, a further more inclusive common in‐group categorization—priming an in‐ group identity—also makes in‐group constituents more salient because they are both in‐groups. If this single priming explanation is correct, then not just positive aVect, but anything that primes in‐group memberships will enhance the diVerentiation‐ reducing eVects of crosscutting categorization, as Crisp et al. (2006c) found when priming a common in‐group identity. Note that we are not suggesting that aVective priming is the single mechanism here, but that a general priming mechanism per se, either positive aVect or in‐group primes, can strengthen the influence of in‐group or out‐group components of crossed category groups. Consistent with this explanation are studies we carried out that directly primed in‐group or out‐group memberships in crossed category contexts. Adapting methodology used by Perdue et al. (1990), we (Crisp & Hewstone, 2000b) used a computerized procedure to prime participants with the pronoun ‘‘we’’ or the letter string ‘‘xxxx’’ during a trait valence identification task, while concurrently thinking about one of the four combined category groups. The results were in line with the idea that priming in‐group components of composite targets will accentuate the diVerentiation (and bias)‐reducing eVects of crossed categorization. A social inclusion pattern (ii ¼ io ¼ oi > oo) was observed following an in‐group prime (‘‘we’’) relative to baseline. These findings support the idea that priming with an in‐group pronoun—which is aVectively congruent with the in‐group components of crossed category groups—can accentuate the diVerentiation‐reducing impact of crossed categorization. Further supportive evidence comes from a study by Crisp, Hewstone, Richards, and Paolini (2003). Following categorization, but before evaluations, participants were primed with out‐group pronouns. The idea here was that there would be a greater focus on out‐group components of crossed
198
RICHARD J. CRISP AND MILES HEWSTONE
category groups due to aVective congruence between the prime and target. Correspondingly, out‐group primes should enhance the diVerentiation‐ accentuating impact of out‐group constituents and lead to greater bias. Participants were asked to complete a ‘‘filler’’ proofreading task where they were required to circle all the instances of the pronoun ‘‘they,’’ or the conjunction ‘‘the,’’ depending on the experimental condition (adapted from Brewer & Gardner, 1996). We found that, compared to an additive pattern of evaluations in the baseline (neutral) prime condition, participants’ evaluations of the crossed category composite groups in the out‐group (‘‘they’’) prime condition shifted to a pattern of social exclusion (see Fig. 8). In eVect, the diVerentiation‐reducing eVects of shared categorization were negated, and the diVerentiation‐enhancing eVects of out‐group categorization were enhanced. The findings reviewed in this section support the idea that apart from positive and negative aVect, category priming can also exert a moderating eVect on the pattern of evaluations observed across crossed category groups. In‐group primes can lead to a social inclusion pattern (Crisp & Hewstone, 2000b; Crisp et al., 2006c), while out‐group primes can produce a social exclusion pattern (Crisp et al., 2003; Table I).
Double in-group
Mixed group
Double out-group
6.0 5.5
5.22
5.12
5.0 Evaluation
4.55 4.5
4.26
4.0
3.85
3.66
3.5 3.0 2.5 2.0 Baseline
Out-group prime
Categorization context Fig. 8. The impact of an out‐group prime on evaluative diVerentiation between crossed category groups. Data from Crisp et al. (2003). Note that for illustrative purposes mixed groups (io and oi) have been combined.
MULTIPLE SOCIAL CATEGORIZATION
199
F. TWO MODERATORS, ONE ROUTE The dual‐route model we have specified previously (Crisp et al., 2002) outlined two classes of moderator along with two distinct routes that lead directly to changes in evaluations. Here, we argue that the evidence points more convincingly to both moderators exerting their influence via the same route, which, notably, does not directly influence evaluations, but either strengthens or weakens the impact of in‐group or out‐group constituents of crossed category subgroups on diVerentiation. So far in our model we have argued that crosscutting categorization weakens category diVerentiation, while convergent categorization strengthens category diVerentiation. We can now add to this model the impact of the moderators discussed in this section. We have noted in our review of pattern support above that the specific pattern observed under positive or negative aVect will depend not only on the outlined priming mechanism, but the interaction of this with category importance (e.g., positive aVect þ diVerential importance leads to hierarchical rejection; Urada & Miller, 2000). While importance here plays a further moderating role, it does so in addition to the basic priming mechanism proposed here (importance simply increasing the sensitivity of relevant categories to aVective ‘‘tagging’’). What we want to do here is derive a basic underlying mechanism from work on crossed category patterns and apply this to our overall model of multiple categorization eVects. That is, we want to connect work on aVect and category primes to the broader model of multiple categorization eVects involving the impact of common in‐group and convergent identities on intergroup evaluations. What does the research reviewed in this section tell us about how aVect and category primes will aVect the relationship between categorization, diVerentiation, and bias? With this in mind our single mechanism model of aVective and cognitive moderators is as follows. Positive mood increases the salience of shared, in‐ group categorizations (Crisp & Hewstone, 2000a; Ensari & Miller, 1998; Urada & Miller, 2000)—positive mood and in‐group identities are aVectively congruent. Negative mood increases the salience of nonshared, out‐group categorizations (Ensari & Miller, 1998; Kenworthy et al., 2003)—negative mood and out‐ group identities are aVectively congruent. Similarly, adding a third layer of shared categorization (Crisp et al., 2006c) eVectively primes in‐group identities (Crisp & Hewstone, 2000b). Priming in‐group identities increases the salience of shared, in‐group categorizations. Priming out‐group identities (Crisp et al., 2003) increases the salience of nonshared, out‐group categorizations. The result of making either in‐group or out‐group components of multiple category targets salient will be an accentuation or attenuation of the eVects of crossed and convergent categorization on diVerentiation.
200
RICHARD J. CRISP AND MILES HEWSTONE
Enhancing the salience of in‐group identities should heighten the diVerentiation‐reducing impact of a common in‐group, but will have little eVect when targets are defined by two out‐group identities (i.e., convergent categorization). Enhancing the salience of out‐group identities should both reduce the impact of making a crossed or common in‐group identity salient (i.e., by enhancing the salience of the existing out‐group identity, thus counteracting any eVect of the new common in‐group), and also increase the diVerentiation‐accentuating eVects of convergent categorization. In this latter case, the target’s out‐group status would not only be reinforced by the recognition that they are diVerent from the perceiver on a second category dimension, but recognition of their out‐group status will be further enhanced by (for instance) out‐group priming caused by a negative mood. We systematically outline these eVects in Table II. In sum, both mood and direct priming of in‐group or out‐group identities enhance the diVerentiation‐reducing or diVerentiation‐enhancing eVects of multiple categorization because both eVectively make in‐group or out‐group components of crossed category groups more salient.4 We now add these moderators to our developing model of multiple categorization eVects (see Fig. 9). This general prediction, that anything that enhances the salience of crosscutting or common in‐groups (versus nonshared out‐groups) reduces diVerentiation and bias, resonates with the more general principles that are inherent to
DIFFERENTIATION‐ENHANCING Moderator
TABLE II ATTENUATING EFFECTS OUT‐GROUP IDENTITIES
OR
OF
PRIMING IN‐GROUP
AND
Process
Outcome
Positive aVect/ in‐group prime
Primes aVectively congruent in‐group categories
Enhances diVerentiation‐reducing impact of crossed categorization
Negative aVect/ out‐group prime
Primes aVectively congruent out‐group categories
Reduces the diVerentiation‐reducing impact of crossed categorization
No eVect on convergent categorization
Enhances the diVerentiation‐accentuating impact of convergent categorization
4 A further possible moderator that is consistent with our social inclusion and exclusion patterns might be regulatory focus (Higgins, 1998). Building on Sassenberg, Kessler, and Mummendey’s (2003) use of regulatory focus to explain positive–negative asymmetries in social discrimination (Mummendey et al., 1992), it might be reasonable to predict that a promotion focus (attention to positive aVect) would lead to a social inclusion pattern whereas prevention focus (attention to negative aVect) would lead to an exclusion pattern.
201
MULTIPLE SOCIAL CATEGORIZATION
Negative affect/ out-group prime −
+
Convergent categories (II versus OO)
+ −
Crosscutting categories (II versus IO)
Intergroup differentiation
+
Intergroup bias
+ Positive affect/ in-group prime
Fig. 9. The relationship between crosscutting and convergent categorizations, diVerentiation, aVective and cognitive moderators, and bias.
the model we are developing here. Recategorization (Gaertner & Dovidio, 2000) works because it makes the common in‐group more salient as a basis for evaluating others than (former) out‐group categorization. Convergent categorization, where targets are out‐group members according to two dimensions of categorization, leads to more negative evaluations than in the case of single out‐groups because the converging memberships make the out‐group identity more salient. Relatedly, positive aVect primes the common in‐group identity, making it more salient; negative aVect primes out‐group identities, making diVerences more salient (reducing, in relative terms, the salience of common in‐group identities and making such interventions less eVective). Direct priming of in‐group or out‐group constituents has the same eVects on category salience, as does inherent or acquired category importance. The basic rule here is that anything that increases the relative salience of common in‐ group identities shared with out‐groups will reduce diVerentiation, anything that increases the relative salience of targets’ out‐group identities will enhance diVerentiation. This is all important because the idea that the relative salience of shared and nonshared identities has an impact on diVerentiation will have significant implications in the next section of this chapter.
202
RICHARD J. CRISP AND MILES HEWSTONE
We now move on from characteristics of intergroup contexts that enhance or reduce the impact of multiple categorization on diVerentiation to focus more closely on the later part of our model. Underlying all the work on moderators reviewed above is the assumption that diVerentiation is positively correlated with bias. The more or less categories are diVerentiated from each other, the more or less intergroup bias will be observed. If pressures toward weakened diVerentiation are evident, such as an overarching common in‐group or positive aVect, then bias will also be reduced (there will be less evaluative diVerentiation). Conversely, if pressures toward enhanced diVerentiation are evident, such as negative aVect or anything that primes out‐ group categories, then bias will increase (there will be more evaluative diVerentiation). While all this work has specified a precise set of moderating conditions that can explain when convergent or crosscutting categorization will be more or less eVective with respect to changing intergroup evaluations, it all rests on this basic assumption: that there is a positive relationship between category diVerentiation and intergroup bias. This basic mechanism can explain the eVectiveness of imposing a common in‐group identity at reducing intergroup bias; as well as the typically greater levels of bias observed against double out‐groups, where out‐group cues toward diVerentiation converge. This model tells us how and when we should expect stronger or weaker diVerentiation—and correspondingly stronger or weaker discrimination. Since our initial statement of this model (Crisp et al., 2002) it has become increasingly apparent that neither it nor our revised model outlined in Fig. 9 capture the full breadth of multiple categorization eVects. In particular, while it accounts for the conditions under which we should expect to observe increases and decreases in bias as a result of an interaction between multiple aYliations (convergent or crosscutting) and aVective and cognitive moderators, the model only holds under conditions where there should be a positive correlation between diVerentiation and intergroup evaluation. The model outlined in Fig. 9 specifies conditions that accentuate the impact of crosscutting or convergent categorization on diVerentiation, which in turn leads to associated changes in intergroup evaluation. There are, however, a range of situations where the assumed positive relationship between category diVerentiation and evaluation does not apply, and where, in fact, we may predict the opposite, a negative relationship (e.g., Hornsey & Hogg, 2000; Vescio, Judd, & Kwan, 2004). We have carried out an extensive program of research on one moderator—in‐group identification—that has an impact on the relationship between diVerentiation and discrimination. Identification is important because it leads to eVects that diverge from all those reviewed above—eVects that are illustrative of a negative relationship between diVerentiation and evaluation. We outline the theoretical basis for this motivational component of the model below, then integrate this with cognitive and aVective
MULTIPLE SOCIAL CATEGORIZATION
203
components specified in Fig. 9, to detail a new, qualified, and elaborated model of multiple categorization eVects.
G. MOTIVATION So far, we have discussed only cognitive (diVerentiation) and aVective (mood) processes involved in multiple categorization contexts. One might, however, also expect motivational processes to play a role, as they do in the perception of social categories in ‘‘simple’’ contexts (Brown & Turner, 1979; Vanbeselaere, 1996). In this section, we review research that has indicated a role for motivational processes in understanding multiple categorization phenomena, before integrating these ideas into our developing model of multiple categorization eVects.
H. SELF‐ESTEEM At the start of this chapter we discussed how motivation can play a role in predicting when people will shift from using one basis for social categorization to another. In particular, we noted how the social identity perspective (social identity theory, Hogg & Abrams, 1988; Tajfel, 1978; Tajfel & Turner, 1979; and self‐categorization theory, Turner et al., 1987) provides a framework within which to explain if, how, and when people will disidentify with devalued identities and recategorize along more favorable criteria. Central is the idea that groups provide meaningful and important sources of self‐ esteem (Abrams & Hogg, 1998; Brown, 1984; Brown & Abrams, 1986; Ellemers, Spears, & Doosje, 2002). In particular, if groups are important due to the link between self‐evaluation and evaluation of the group, then we can make some specific predictions regarding how people will behave as a function of relative levels of self‐esteem and group status. Hogg and Abrams (1990) suggested two corollaries that specify this predicted link. Individuals, by virtue of their aYliation, will have higher self‐esteem when their in‐group is evaluated positively (Abrams & Hogg, 1998). Corollary 1 of the self‐esteem hypothesis specifies that intergroup discrimination leads to enhanced social identity and thus elevates posttest self‐ esteem. Corollary 2 states that depressed self‐esteem should lead to greater discrimination due to the desire to obtain positive self‐esteem through in‐ group enhancement. Applying these predictions to crossed category contexts, Hewstone et al. (1993) suggested that (1) posttest self‐esteem should follow the same pattern as intergroup discrimination and perceived similarity (i.e., diVerentiation) and (2) relative to simple categorization, there will be
204
RICHARD J. CRISP AND MILES HEWSTONE
a breakdown of the self‐esteem/discrimination association in responses to mixed membership groups (because categorization is divergent and not clear‐cut), and a strengthening of this relationship for double out‐groups (where cues for out‐group categorization converge). Crossing nationality and religion in Bangladesh, they found, however, no diVerences in posttest self‐esteem due to categorization condition, nor any clear evidence for a breakdown of the self‐esteem‐discrimination link. Some other studies have examined the role of self‐esteem in crossed categorization but similarly found little empirical support. Vanbeselaere (1991), for instance, found that when self‐esteem was measured after performance evaluations there were no diVerences between the simple, mixed and double out‐ group conditions. In reviewing evidence for the role of self‐esteem in multiple categorization contexts we concluded that there was, overall, little evidence for self‐esteem processes in such contexts (Crisp & Hewstone, 1999b, 2000b), which is consistent with generally equivocal support observed with respect to the self‐esteem hypothesis in general, and warnings to ensure correct implementation and interpretation of the predicted role of self‐esteem in the categorization—identification—bias relationship (see Rubin & Hewstone, 1998, for a review and suggestions for clarification of the hypothesis, also Turner, 1999; but see Smurda, Wittig, & Gokalp, 2006, for some evidence supportive of both corollaries for higher identifiers). From the studies reviewed so far it remains unclear whether the self‐ esteem implications of multiple category membership have any impact on intergroup evaluations. Motivations associated with social identities do not, however, only involve self‐esteem. Of more direct relevance to the idea that multiple categorization can change category representation, blur intergroup boundaries, and reduce intercategory diVerentiation is research that focuses on a motivational drive associated with the acquisition and maintenance of in‐group distinctiveness.
I. DISTINCTIVENESS THREAT A central proposition of social identity theory (Tajfel, 1982; Tajfel & Turner, 1979) is that group members engage in various strategies to define themselves positively in relation to other social groups; that individuals are motivated to achieve a positive and distinct self image. Distinctive social categorization clarifies and defines social situations, providing a means for predicting how out‐groupers will behave, reducing uncertainty, and providing a set of prescriptive in‐group norms to guide perceivers’ behavior (Hogg, 2000, 2001). It is therefore not just a desire for positivity that drives group members, but also a desire for distinctiveness (Brewer, 1991)—a drive
MULTIPLE SOCIAL CATEGORIZATION
205
that can manifest itself in the form of increased evaluative diVerentiation (in‐ group favoritism). When distinctiveness is compromised, being more in‐ group favoring is a clear way of reestablishing us as diVerent from them (we are better than them). There is much evidence that increased similarity can provide the motivation for intergroup discrimination. Jetten, Spears, and Manstead (2001) observed a negative and linear relationship between distinctiveness—diVerentiation—and bias (see also Jetten, Spears, & Manstead, 1997). Henderson‐King, Henderson‐ King, Zhermer, Posokhova, and Chiker (1997) found that perceived intergroup similarities led to greater bias when the out‐group was perceived as a threat to a valued in‐group identity (for a meta‐analytic review see Jetten, Spears, & Postmes, 2004). These findings diverge from the underlying model of diVerentiation–discrimination that has formed the basis for all of the research we have reviewed up until this point: the idea that diVerentiation is positively related to discrimination. This implies there must be moderating conditions under which motivational drives will or will not have an impact on intergroup bias—conditions under which we will observe either a positive or negative relationship between diVerentiation and bias. Research suggests that this moderator is in‐group identification.
J. IN‐GROUP IDENTIFICATION When identification with in‐groups is high, the more similar groups are perceived to be, the more bias is observed (Jetten et al., 2004; see also Hogg & Abrams, 1990). For instance, Jetten et al. (2001) found that for high, but not low, identifiers more bias was observed when diVerentiation was low compared to when groups were perceived as very diVerent. This finding has been replicated many times and is a highly robust characteristic of intergroup contexts involving important social identities (see Allen & Wilder, 1975; Brown & Abrams, 1986; Diehl, 1988; Jetten, Spears, & Manstead, 1997). Higher in‐group identification, contextually or chronically defined, appears to be a condition under which a negative correlation between category diVerentiation and intergroup bias will be observed. These findings are consistent with social identity theory: high identifiers (i.e., group members who feel strongly committed to a salient in‐group) appear to be motivated to diVerentiate their in‐group from similar out‐groups, on relevant dimensions of comparison, in order to restore or maintain group distinctiveness and a positive social identity. Identification therefore moderates whether category diVerentiation (intergroup distinctiveness) is positively or negatively associated with intergroup bias. For lower identifiers, identity motivations should not apply, and more diVerentiation should lead to more bias. For higher identifiers,
206
RICHARD J. CRISP AND MILES HEWSTONE
however, the opposite should be the case: less diVerentiation should lead to more bias (due to evoked motivations to retain distinctiveness).
K. MOTIVATION AND MULTIPLE CATEGORIZATION Although the theoretical basis for higher identification leading to negative reactions to intergroup similarity is well established, there were, until recently, few empirical tests of this idea in the multiple categorization literature. Surveying this literature, the applicability of the motivational model (that is, the negative, rather than positive, relationship between diVerentiation and bias) appears justified. The positive correlation between diVerentiation and intergroup bias (i.e., a reduction in bias when intergroup boundaries are blurred) has commonly been confirmed when using artificially created laboratory groups, groups that would not normally engender higher levels of identification from their members. Observations of reduced bias following crossed categorization have been provided by Vanbeselaere (1987, 1991, 1996) and Marcus‐Newhall et al. (1993), who used ad hoc categorizations. Crisp et al. (2001) and Diehl (1990) also found that bias was lower when a shared common in‐group membership cuts across an existing in‐group–out‐group diVerence, compared to when social targets were an out‐group member on both dimensions; again these studies employed artificial groups. Although there are also many real‐group studies where multiple categorization methods have successfully reduced bias, it is when groups are real that some deviations from the basic bias reduction predictions have been observed (Dovidio et al., 1998b; Hornsey & Hogg, 2000; Kessler & Mummendey, 2002). Illustrative is Hornsey and Hogg’s (2000) study. They examined relations between groups of university students who were enrolled in humanities and mathematics‐science faculties. There was greater intersubgroup bias when participants were categorized exclusively as university students than when their faculty area subgroups were made salient. Hornsey and Hogg suggested that the greater bias found in the superordinate condition may have resulted from perceived threat to the distinctiveness of a subgroup category that was important to perceivers (see also Hornsey & Hogg, 2002). The importance of identity distinctiveness represents a potentially major issue for the crossed categorization and common in‐group identity approaches (Brewer & Gaertner, 2001; Hewstone 1996). A crosscutting or common in‐group identity may only be short‐lived, or unrealistic in the face of powerful ethnic and racial categorizations. Sen (2006), recollecting his experience of Hindu–Muslim riots in his native India, referred to the ‘‘speed with which the broad human beings of January were suddenly transformed
MULTIPLE SOCIAL CATEGORIZATION
207
into the ruthless Hindus and fierce Muslims of July’’ (p. 2). In a similar manner the imposition of an overarching Yugoslavian identity in the Balkans failed, in the long term, to replace separate Serb, Croat, Macedonian, Albanian, and Muslim categories. For groups with a history of antagonism, and minorities under pressure to assimilate to a superordinate category that is dominated by a majority out‐group, the prospect of a common in‐group identity may constitute a threat (Brewer 2000; Hornsey & Hogg, 1999; van Oudenhoven, Prins, & Buunk, 1998). More generally, a superordinate identity can deprive individuals of valued social identities in smaller, less inclusive groups (Brewer, 1991, 1993). Eradicating original categorizations is unlikely to meet the needs of assimilation and diVerentiation, or of cognitive simplicity and uncertainty reduction (Brewer 2000; Hogg & Abrams, 1993). The implications of the above findings are that using common in‐groups or crosscutting categories to blur intergroup boundaries may sometimes lead to increased, rather than decreased, intergroup bias. In a statement of multiple categorization theory that focused on motivational processes, Crisp (2006) argued that an assessment of individuals’ idiosyncratic identification with, or commitment to, in‐groups might oVer not only a means for predicting when increases in bias might be expected, but also support the idea that social identity motivations have a part to play in understanding multiple categorization phenomena. It was this idea that formed the basis for a program of work we carried out testing the hypothesis that identification with in‐groups may be an important determinant of the success of multiple categorization interventions. This work provides empirical support for a modification to our existing model of multiple categorization eVects (see Fig. 9), one that incorporates motivational concerns, and which specifies when the typically observed positive correlation between diVerentiation and discrimination will be observed. In the section below, we first review the evidence that identification with in‐groups moderates reactions in multiple category contexts, before specifying our revised model of multiple categorization.
L. IDENTIFICATION AND MULTIPLE CATEGORIZATION In a set of studies van Leeuwen, van Knippenberg, and Ellemers (2003) provided evidence that weakening intergroup boundaries had diVerent eVects depending on subgroup identification. Using minimal groups and a paradigm designed to test the eVects on group perception of a merger between the in‐ group and out‐group (a design that modelled the real‐life context of corporate mergers), van Leeuwen et al. (2003) found that premerger identification predicted postmerger intergroup attitudes. Specifically, prior to the merger
208
RICHARD J. CRISP AND MILES HEWSTONE
identification did not predict intergroup bias. Following the merger, however, identification was positively correlated with bias. The higher the level of premerger identification, the higher the postmerger in‐group favoritism, supporting the idea that subgroup identification can predict diVerent reactions to recategorization. We carried out a similar program of work, but focusing on the implications of in‐group identification for the eVectiveness of intervention strategies to reduce intergroup bias. Supporting the findings of van Leeuwen et al. (2003), Crisp et al. (2006b) consistently found that identification moderated the eVects of making a superordinate context salient. In one study, the Implicit Association Test (IAT) was used as a measure of implicit intergroup bias (Greenwald, McGhee, & Schwartz, 1998). The target groups were British and French and as such participants were asked to categorize positive and negative words (e.g., ‘‘gift’’ and ‘‘war’’) alongside British and French names (e.g., ‘‘Sarah’’ and ‘‘Genevie`ve’’). The IAT bias eVect was defined as the extent to which the British participants responded faster on trials that matched ‘‘congruent’’ pairings (i.e., positive word and in‐group name) compared to ‘‘incongruent’’ pairings (i.e., positive word and out‐group name). In a control group, participants were simply asked to evaluate the British and the French. In an experimental group, participants were first asked to read a paragraph advocating closer European integration and the dissolution of member states into a ‘‘United States of Europe’’ (making a common in‐ group identity salient and threatening the distinctiveness of being British). Compared with the control group, in the experimental group the IAT eVect was higher for participants who were higher identifiers with their British category membership, and there was a trend for a decreased IAT eVect for lower identifiers (see Fig. 10). Using a similar paradigm Crisp and Beck (2005) asked participants to think about multiple characteristics that were shared between the in‐group and out‐group as a means of reducing categorical diVerentiation. Participants thought about characteristics that students at their own and a local rival university shared, for example, being studious, sporty, enthusiastic, and so on. In line with expectations, while lower identifiers were significantly less biased after completing the shared characteristics task compared to participants in the control condition, higher identifiers were more biased after completing the shared characteristics task. Crisp et al. (2006c) have further shown that identification can directly moderate the additive pattern observed in crossed categorization contexts. We observed that while imposing a superordinate, crosscutting category on an existing additive pattern of evaluation led to a weakening of evaluative diVerentiation, this was only for perceivers who regarded their in‐group subgroups as of lower importance (this was the case for both real and
MULTIPLE SOCIAL CATEGORIZATION
Lower identifiers
209
Higher identifiers 291
300
IAT effect (ms)
250 200
197
150 109 100 61 50 0 Baseline
Common in-group
Categorization context
Fig. 10. The moderating eVect of subgroup identification on implicit intergroup bias in a common in‐group context. Data from Crisp et al. (2006b; Experiment 3). Note that IAT eVect is incongruent RT minus congruent RT.
artificial groups). Higher identifiers resisted this recategorization and diVerentiation between the crossed category subgroups remained resolute. Here then, even when groups are defined by two crosscutting categorizations, identification with the composite group predicts the impact of a more inclusive common in‐group identity. The studies reviewed above illustrate that identification can moderate the typically observed positive relationship between diVerentiation and discrimination. Crisp (2006) also reports findings from an unpublished study (Bloom & Crisp, 2004) that demonstrated either a positive relationship between diVerentiation and bias, or a negative relationship between diVerentiation and bias, contingent on identification. We asked participants to generate multiple crosscutting or convergent categories. Thinking about crosscutting categories should emphasize intergroup similarity, while thinking about converging categories should reinforce diVerences and increase intergroup diVerentiation. In this experiment the target groups were ‘‘University of Birmingham [in‐group]’’ and ‘‘Aston University [out‐group]’’ students, the latter being a local rival university and obvious source of social comparison for our participants. Identification was manipulated by way of false feedback on a multiple choice task involving trivia questions about the in‐group university (e.g., ‘‘What year was the University of Birmingham founded?’’ with five options). Following false feedback informing participants that they
210
RICHARD J. CRISP AND MILES HEWSTONE
were either lower or higher identifiers with their university student group, participants completed the crosscutting or convergent categorization task which involved participants first generating, and then elaborating on, categories that either crosscut or reinforced the Birmingham–Aston intergroup distinction. Participants then completed a measure of intergroup evaluation (a Likert scale rating of team member preference in a supposed future problem‐solving task). What was striking about these findings was that participants who had received feedback that they were higher identifiers were more biased following generation of crosscutting categories compared to generation of convergent categories (see Fig. 11). In fact, there was no significant in‐group favoritism for higher identifiers in the converging categories condition, but significant bias for these participants in the crosscutting condition. This is consistent with the notion that higher identifiers are more inclined to think positively about out‐group members when they are clearly distinguishable from them. The studies reviewed above support the idea that in‐group identification can moderate the eVectiveness of diVerentiation‐reducing interventions on intergroup attitudes. Using varied multiple classification paradigms, and a range of target groups and contexts, we and others have consistently found that whereas lower identifiers respond positively to reduced intergroup diVerentiation, higher identifiers appear to react against the implied merging
Lower identifiers
Higher identifiers
2.0
Intergroup bias
1.5
1.46
1.31 0.846
1.0 0.5 0.0
−0.308
−0.5 −1.0 Convergent
Crosscutting
Categorization context Fig. 11. The impact of in‐group identification and crosscutting versus convergent categorization context on intergroup bias. Data from Bloom and Crisp (2004).
MULTIPLE SOCIAL CATEGORIZATION
211
of boundaries and become more in‐group favoring.5 That higher identifiers are likely to react negatively to blurring intergroup boundaries has significant implications for categorization models of bias reduction. As we have noted in this chapter, most research on multiple categorization has implicitly or explicitly assumed a positive correlation between category diVerentiation and intergroup bias. In other words the underlying assumption has been that to improve intergroup attitudes we need to bring groups closer together. We can do that by emphasizing crosscutting or common in‐group identities, and by creating conditions that orient people to appreciate such crosscutting ties (e.g., a positive mood). However, the work reviewed in this section shows that in specific contexts (e.g., where identification is high), everything that we do to encourage the dissolution of intergroup boundaries may, in fact, have the opposite eVect to that intended (i.e., it will increase intergroup bias). It is the recognition that creating common or crosscutting identities can sometimes threaten intergroup distinctiveness, and the need to address this issue, that prompted the research that we review in the next section.
M. DUAL IDENTIFICATION The importance of identification with real groups (versus artificial groups) was initially acknowledged in Gaertner et al.’s (1993) discussion of the Common In‐group Identity Model. They argued that a negative relationship between diVerentiation and bias would be observed when groups matter to people; when this is the case there may be a reluctance to forsake their subgroup identities in favor of a superordinate entity (see also Brown & Hewstone, 2005; Hewstone & Brown, 1986, for similar warnings in the contact hypothesis literature). To protect against loss of distinctiveness for groups involved in contact, Brown and Hewstone argued that two factors are important: (1) the salience of group boundaries should be maintained during contact to promote generalization across members of the target out‐ group and (2) each group should be distinct in terms of the expertise and experience it brings to the contact situation, creating functional complementarity between subgroup and superordinate identities. Several studies show that positive eVects of contact are more likely to generalize to the out‐group 5
Note that because identification is often skewed toward either higher or lower levels in the studies we report, it is not unusual to observe an identification categorization interaction whereby higher identifiers show increased bias, while lower identifiers do not change; or where lower identifiers show decreased bias, while higher identifiers do not change. Collectively, however, the findings from all the studies reviewed illustrate how overall lower identifiers react to blurred boundaries with decreased bias, while higher identifiers react to blurred boundaries with increased bias.
212
RICHARD J. CRISP AND MILES HEWSTONE
as a whole when the group membership of a contact partner is made salient (van Oudenhoven, Groenewoud, & Hewstone, 1996) or the partner is typical, rather than atypical, of the out‐group as a whole (Brown, Vivian, & Hewstone, 1999, Study 1). Maintaining the salience of intergroup boundaries is also important for the reasons we have set out above: for higher identifiers diVerentiation with out‐ groups is imperative, and so anything that compromises this distinctiveness will promote a reaction designed to reestablish this valued distinctiveness (i.e., in‐group favoritism). To counter this desire for distinctiveness in common in‐ group contexts, Dovidio et al. (1998) proposed the idea of a simultaneous subordinate/superordinate categorization (although for earlier indications of this idea see Gaertner et al., 1989). The idea is that by maintaining recognition of subgroup distinctiveness, while simultaneously bringing groups together on a second, crosscutting dimension, reactive tendencies expressed by higher identifiers can be curtailed. In the context of students’ university department identities, this dual identity approach was found to be eVective at reducing bias while maintaining subgroup distinctiveness (Dovidio et al., 1998; see also Hornsey & Hogg, 2000; and for convergent support related to subgroup negotiation in organizations see Eggins, Haslam, and Reynolds, 2002). The applicability of this approach to real groups balances preserved social identity at the subordinate level, while simultaneously promoting favorable intergroup attitudes at the superordinate level. Theoretically, the representation of two subgroups within one superordinate category maintains the salience of the subgroup boundary but simultaneously encourages a single inclusive identity. The idea that maintaining subgroup boundaries within a superordinate group attenuates distinctiveness threat was supported by Crisp et al. (2006b) who found that ensuring subgroup distinctiveness within a superordinate identity reduced bias specifically for higher identifiers: those group members who, in particular, view reduced distinctiveness as a threat to identity.
N. A DIFFERENTIATION–DISCRIMINATION MODEL OF MULTIPLE CATEGORIZATION EFFECTS In this section we have reviewed evidence that, under some circumstances, blurring intergroup boundaries by making common in‐group identities salient can increase, not decrease, intergroup bias. Specifically, when in‐groups are important to people, emphasizing common identities with out‐groups leads to a reactive increase in bias. This is consistent with the idea that for higher identifiers, blurring intergroup boundaries threatens valued positive distinctiveness, and, as a result, motivates such group members to show
MULTIPLE SOCIAL CATEGORIZATION
213
in‐group favoritism in an attempt to reassert their diVerences from the out‐ group. To counteract such reactions to recategorization, the salience of subgroup and superordinate group identities must be balanced. Notwithstanding the clear importance of this finding for eVorts to apply multiple categorization theorizing to real contexts of conflict, these findings present a strong basis on which to add to our existing model of multiple categorization eVects. The model we have proposed previously (Crisp et al., 2002) and refined in Fig. 9 rests on a fundamental positive relationship between intergroup diVerentiation and intergroup discrimination. This model specifies how salient crosscutting and common in‐group identities can weaken category diVerentiation, and how the impact of these shared identities on diVerentiation can be accentuated by moderators such as positive aVect and in‐ group or out‐group primes. The positive correlation between diVerentiation and discrimination means that the interactive eVect of multiple categorization and these diVerentiation‐accentuating or attenuating moderators is then reflected in intergroup bias and intergroup discrimination (Doise, 1978; Turner et al., 1987). This model is consistent with existing accounts of the eVects of common in‐group categorization (Gaertner & Dovidio, 2000) and crossed categorization (Crisp & Hewstone, 2006). The observed negative relationship between category diVerentiation and intergroup bias that we have reviewed in this section cannot, however, be accounted for by this existing framework. The model outlined in Fig. 9 specifies conditions that either strengthen or weaken category diVerentiation, which, in turn, either increases or decreases bias respectively. If applied to contexts where categories are important to perceivers then these same factors that decrease diVerentiation will not decrease, but increase intergroup bias. For instance, if a positive mood primes in‐group components of crossed category groups, making shared, common in‐group identities more salient to perceivers, then while this may reduce diVerentiation and therefore bias for group members who are not particularly committed or involved in their group memberships (Crisp & Hewstone, 2000a), the same reduction in diVerentiation should increase bias for higher identifiers. This is a novel prediction based on an integration of work underlying our previous specifications (Crisp et al., 2002) and the research that has revealed a negative relationship between decreased diVerentiation and bias (Crisp, 2006), but one that is entirely consistent with both sets of findings. Some of our work supports this integration. Crisp et al. (2006c; Experiment 3) observed a trend consistent with the idea that increasing pressures toward assimilation (creating a common in‐group context/priming in‐groups) lead to less bias for lower identifiers (less diVerentiation, less bias) but a stronger desire for diVerentiation for higher identifiers (less diVerentiation, more bias). In the light of work showing that identification can moderate the relationship between
214
RICHARD J. CRISP AND MILES HEWSTONE
diVerentiation and discrimination, we can now further qualify our model of multiple categorization eVects in Fig. 12. We should note here that although identification appears as a moderator in the later part of our model, determining the nature of the diVerentiation–bias relationship, this is not meant to imply that identification is not, implicitly or explicitly, implicated at the start of the impression formation process.6 In fact, we do not see this model as reflecting a temporal sequence, rather it specifies the consequences for intergroup evaluations of the immediate and apparent recognition of relationships between self and in‐group categories (identification), and in‐group and out‐ group categories (category diVerentiation). To summarize, so far in this chapter we have provided a refocused review of existing work on multiple categorization, by bringing to the fore the positive relationship between diVerentiation and discrimination that exemplifies much of the previous literature, and added to our previously specified model a third factor that takes perceiver motivations into account.
Negative affect/ out-group prime −
+ Low in-group identification
Convergent categories (II versus OO)
+ −
Crosscutting categories (II versus IO)
+ Intergroup bias
Intergroup differentiation − High in-group identification
+ Positive affect/ in-group prime Fig. 12. The relationship between crosscutting and convergent categorizations, diVerentiation, aVective and cognitive moderators, identity motivations, and bias. 6
We note also that, ignoring motivational implications for a moment, identification can imply category importance, and can therefore interact with positive aVect as outlined in our earlier discussion of crossed category evaluation patterns (see also Miller et al., 2006).
MULTIPLE SOCIAL CATEGORIZATION
215
However, the story does not end here. Multiple categorization is not only about emphasizing conditions that lead either to increased or decreased diVerentiation respectively, and the impact that diVerentiation has on bias depending on perceiver motivations. Considering multiple criteria for social categorization does not only either blur or reinforce intergroup diVerences. At the start of this chapter we noted that in some contexts people do not consider multiple bases for categorization simultaneously, but sometimes they shift or switch between them; they ignore some and focus on others. Related to this is a second consequence of considering multiple categorizations that does not impact on diVerentiation but rather implies a change in the way people form impressions of others. We turn now to these consequences of complexity.
O. COMPLEXITY The studies we have reviewed so far have been principally concerned with changing the perceived diVerentiation between categories—blurring intergroup boundaries, and the conditions under which this has a positive or negative impact on bias. Below we consider research that has focused more on intracategory diVerentiation—when multiple categorization leads to a breaking up of category coherence, an increase in perceived group variability and complexity. Examining how multiple categorization has been conceptualized as a means of understanding social identity complexity will, in turn, lead us to formalize a final component in our elaborated model of multiple categorization eVects.
P. SUBGROUPING AND PERCEIVED GROUP VARIABILITY Whereas the crossed categorization and common in‐group identity models have been almost exclusively concerned with intergroup judgments, work into subgrouping (Brewer et al., 1981; Maurer et al., 1995; Park, Ryan, & Judd, 1992; Richards & Hewstone, 2001) has focused on stereotyping and perceived group variability (Park & Judd, 1990; Park, Judd, & Ryan, 1991). Subgrouping describes intracategory representation as subcategories within existing higher‐order categories, reflecting the perceived variability (dispersion), and stereotypically (central tendency) of the group representation. Subgrouping refers to the process by which members of a superordinate group are sectioned into smaller meaningful clusters (Maurer et al., 1995). In contrast to ‘‘subtyping,’’ where disconfirming members of the superordinate group are psychologically ‘‘split oV’’ from the group as a whole, having
216
RICHARD J. CRISP AND MILES HEWSTONE
no impact on the overall stereotype of the group (Hewstone, 1994; Weber & Crocker, 1983), making perceivers aware of the multiple subgroup clusters that exist within a group can lead to a reduction in the tendency for out‐group members to be regarded as homogeneous, and corresponding reductions in stereotyping. Perceivers tend naturally to think of out‐groups as containing fewer subgroups than in‐groups, which statistically explains the out‐group homogeneity eVect (i.e., the tendency to see out‐groups as all alike whereas in‐ groups are seen to be varied and diverse; Jones, Wood, & Quattrone, 1981; Park et al., 1992). Perceivers who become aware that a social category (e.g., science students) can be made up of a number of subgroups (e.g., those science students who are athletic, those who like theatre, and so on) tend to think less of the overall group in terms of a central stereotype and also tend to think of the overall group as more variegated and heterogeneous. The presentation of subcategories can thus break down perceived group homogeneity and reduce the stereotypicality (central tendency) of group representation (Maurer et al., 1995; see also Pettigrew, 1981). Encouraging participants to think of the out‐ group as made up of a number of diVerent subgroups leads to perceptions of increased out‐group variability (Park et al., 1992, Study 1). Repeated use of subgroups can result in the development of a more heterogeneous group representation (Ryan, Park, & Judd, 1996). Hewstone and Hamberger (2000, Experiment 2) demonstrated that subgrouping decreased stereotyping, even when subtyping was apparent. Subgrouping not only has a moderating influence on stereotyping, but also seems the natural way people structure their intracategory group representations (Park et al., 1992). In‐group information can be organized into subgroups, which can be encoded, recalled, and applied easily and accurately, while there appears insuYcient information to organize out‐ group representations in the same way (Richards & Hewstone, 2001). Park et al. (1992, Study 2) showed that the out‐group homogeneity eVect was mediated by the tendency to see the in‐group as made up of more subgroups than the out‐group. In sum, subgrouping moderates intracategory structure such that changes reflect (or can aVect) the dispersion of group attributes around the prototype (Judd, Ryan, & Park, 1991; Linville, Fischer, & Yoon, 1996; Mullen & Hu, 1989; Park et al., 1991). Intragroup variability can be seen to reflect not only the dispersion of traits, but in addition, how complex perceptions of intracategory structure are (see Linville, 1982); that is, how diVerent elements of category representation interrelate. Recently, a qualified account of intracategory representation has been proposed that goes further and considers the crosscutting or convergent nature of the subgroups that define intracategory structure: in other words, an account of social identity complexity.
MULTIPLE SOCIAL CATEGORIZATION
217
Q. SOCIAL IDENTITY COMPLEXITY Roccas and Brewer (2002) proposed that individuals’ representations of their multiple group memberships can vary along a dimension they call complexity. Social identity complexity refers to an individual’s subjective representation of the interrelationships among his or her multiple group identities. It refers to the degree of overlap perceived to exist between groups of which a person is simultaneously a member. When one’s representation is low in complexity, it is an indication that one perceives a high overlap between both the typical characteristics of one’s various social category memberships as well as an overlap between the actual members of those same categories. In other words, low complexity is represented by high overlap between membership and characteristics. High complexity, by contrast, is the opposite. It implies that the representation of each in‐group category is distinct from the others, both in characteristics as well as membership. Brewer and Pierce (2005) have shown that higher levels of social identity complexity (more crosscutting category memberships) are associated with greater out‐group tolerance, including greater support for aYrmative action and multiculturalism. Greater complexity was also related to higher education levels, liberal political ideology, and age. This work illustrates how crosscutting categorization can have a positive impact not only at an intergroup level, but also at an intragroup level (see also research by Homan, 2006, who found that within‐category diversity represented by crosscutting versus converging identities was most productive in work teams). Related to the idea that social identity complexity changes how people define themselves and react to out‐groups is work by Mummendey and colleagues on in‐group projection (Mummendey & Wenzel, 1999). This model builds on self‐categorization theory (Turner, 1981; Turner et al., 1987) to argue that superordinate categorization creates a context of comparability between in‐group and out‐group subgroups. Mummendey and Wenzel (1999) argue that in intergroup situations where the superordinate group is salient, the in‐group prototype will be projected on to the superordinate in‐group. This in‐group projection may cause the out‐group to be perceived more negatively. If both the in‐group and out‐group are conceptualized as part of a superordinate category (a relevant comparison context), and the in‐group prototype has been projected onto the superordinate category, then the out‐group will be seen as deviating from the norm (Wenzel, Mummendey, Weber, & Waldzus, 2003). The in‐group‐projected prototype will invariably be positive and hence the out‐group will be judged less positively by comparison (Wenzel, 2001). The idea that social identity complexity can reduce intergroup bias has also been developed in the In‐group Projection Model. Waldzus, Mummendey, Wenzel, and Weber (2003) found that projection of
218
RICHARD J. CRISP AND MILES HEWSTONE
the in‐group prototype does not occur when the superordinate category is itself perceived as having a complex prototype. Although Waldzus and colleagues do not specifically define complexity as involving multiple crosscutting or converging bases for subgroup categorization, their findings are consistent with those of Brewer and Pierce (2005): categorization complexity appears to exert a debiasing eVect on intergroup relations. These studies of social identity complexity focus on the idea that how people structure their representation of in‐groups has an impact on general levels of tolerance toward out‐groups. In Section IV, we argue that the complexity of the crosscutting categories that define intergroup relations (i.e., complexity defined by both the in‐group and out‐group) also has a significant eVect on intergroup evaluations. We argue that under certain conditions the impact of multiple identities moves from the diVerentiation– discrimination model we have presented so far to one where a diVerent set of processes become relevant. We argue that multiple categorization, under conditions that create a high level of complexity, will change the strategies people adopt to form impressions of others.
IV. Decategorization In this chapter we have discussed the eVects of multiple categorization on category diVerentiation, representation, and evaluation. Blurring boundaries reduces the salience of intergroup diVerentiation and, all other things being equal, will reduce intergroup bias. But there are moderators of the impact that salient categorization has on category diVerentiation. These include positive or negative aVect or contexts that emphasize or prime crosscutting or converging bases for categorization, which can strengthen or weaken perceived intergroup diVerentiation. DiVerentiation then predicts evaluation, and we have shown how identification moderates this diVerentiation–discrimination relationship. All of this work has been concerned with changing the perceived diVerentiation between in‐groups and out‐groups. In this section we examine how thinking about multiple categories can change how we process information about others—how we approach the task of impression formation.
A. CATEGORY FUNCTION For the most part, classifying others into broad groupings—social categorization—is a time‐ and resource‐preserving mental shortcut (Fiske & Taylor, 1991; Hamilton, 1979; Macrae & Bodenhausen, 2001). As we discussed
MULTIPLE SOCIAL CATEGORIZATION
219
earlier, social categorization and category diVerentiation can be seen as an adaptive, functional mechanism adding meaning and clarity, predictability, and coherence to the social world (Doise, 1978; Turner et al., 1987), even when artificial groups are created in the laboratory (Rabbie & Horowitz, 1969; Tajfel, 1982; Tajfel et al., 1971; for a review see Bourhis, Sachdev, and Gagnon, 1994). Social categorization provides a quick understanding of what is occurring (van Knippenberg & Dijksterhuis, 2000) and it provides maximum information with the least cognitive eVort (Macrae & Bodenhausen, 2001). It reduces the need to process data by providing prescriptive group norms; it directs attention and processing resources to expectancy‐relevant information (Bodenhausen, 1992; Hastie, 1981; Macrae, Stangor, & Milne, 1994; Oakes & Turner, 1990; Stangor & Macmillan, 1992). Categorization is a simplifying mechanism, it ascribes meaning, and it is psychologically useful. The idea that we discuss in this section is whether removing the cognitive functionality of categorization will prevent its use in guiding evaluative judgments. We argue that this can be achieved by making multiple bases for social categorization simultaneously salient, providing the context is suYciently complex. This idea is consistent with some well‐established models of impression formation that provide a framework for understanding how multiple categorization can lead from a categorical mode of impression formation (where attitudes and behaviors can be guided by potentially negative stereotypes) to an individuated mode of impression formation (where evaluation is based on an assessment of a target’s idiosyncratic characteristics).
B. DUAL PROCESS MODELS OF IMPRESSION FORMATION Brewer’s (1988) dual process theory and Fiske and Neuberg’s (1990) continuum model both consider impression formation to comprise two distinct processes: categorization and individuation. The basic premise of Brewer’s (1988) dual process theory is that either a top‐down heuristic (category) or a bottom‐up systematic (individuated) approach can be adopted when forming impressions of others. Brewer and Harasty Feinstein (1999) note that this basic idea of two types of processing underlying impression formation is also reflected in Fiske and Neuberg’s continuum model. Here a continuum of impression formation is specified where one extremity is category‐based processing and the other is attribute‐based (individuated) processing. Fiske and Neuberg argue that perceivers initially search for the application of a social category to an encountered target. If this search succeeds, the target is categorized and is perceived to possess prototypic characteristics associated with the category. Following this stage, if categorization is unsuccessful,
220
RICHARD J. CRISP AND MILES HEWSTONE
there is a confirmatory process to determine how good a fit there is between the category and target. Next, there is a shift toward the individuated mode of perception by invoking an attribute‐by‐attribute approach to form an impression of the target person. If fitting the target to a category proves more diYcult, the perceiver will then look to define the target more in terms of individuated characteristics, an attentional and cognitively demanding procedure. The general idea that people adopt either a more categorical or more individuating approach to impression formation provides the basis for our predictions regarding how multiple categorization can lead to reduced bias by processes other than reduced diVerentiation. The switch in processing from using categorization to individuation can be termed decategorization. Decategorization results when initial categorizations of people as group members are replaced by a more comprehensive and diVerentiated perception of the target individual (see Brewer & Miller, 1984; Ensari & Miller, 2001; Wilder, 1986). Respected American journalist Tom Wicker described the process well when reminiscing about a 10‐day journey made in 1946 in one crowded carriage of a troop train, carrying 27 black and 3 white occupants back to the South. He wrote that the journey: ‘‘had been the time of his discovery that black people were just that—people, individuals . . . hurting, laughing, worrying in much the same fashion about much the same things . . . . No two of them were much alike in temperament or personality’’ (Wicker, 1975, p. 209). Once decategorization has occurred, the target person should be primarily defined as an individual rather than a group member, which should remove relevant category‐based bias (we can also think of this process as a contextual shift toward a more meaningful and useful way to define the social situation consistent with self‐categorization theory; Turner et al., 1987). Previous research has found that conditions that promote decategorization will encourage participants to favor the in‐group less (the judgment is then based on an appreciation of personal merit, not on the basis of preconceived, potentially negative, stereotypic expectancies; see Brewer & Miller, 1984; Krueger & Rothbart, 1988; Miller & Harrington, 1992). This change in the mode of processing should also allow the perceiver to develop a more personalized and less homogeneous perception of in‐group and out‐group members (Ensari & Miller, 2001; Fiske & Neuberg, 1989).
C. MULTIPLE CATEGORIZATION AND DECATEGORIZATION In contrast to focusing participants on two crosscutting dimensions of categorization (Deschamps & Doise, 1978; Vanbeselaere, 1987, 1991), as has typically been the case in the work reviewed so far in this chapter,
MULTIPLE SOCIAL CATEGORIZATION
221
considering categorization criteria that create a more complex context may uniquely lead to a decategorization eVect (Crisp et al., 2001). This can be achieved either through increasing the number of categories that participants need to process in forming an impression of a target person (Crisp et al., 2001; although numbers alone may not always lead to decategorization, see Urada et al., in press) or by making the nature of the relationship between those categories more complex (Hall & Crisp, 2005). The reasoning is as follows: we know that participants are cognitively able to use two crosscutting dimensions of social categorization under normal processing conditions (Vanbeselaere, 1987). This is characteristic of the work we have considered so far that forms the basis for the diVerentiation–discrimination model outlined in Fig. 12. As the number of crosscutting categorization dimensions increases, the intergroup context becomes more complex and the perceiver may no longer be able to easily determine on what basis they are similar to, or diVerent from, the target person (here the number of category dimensions to be considered defines complexity). There may be too many dimensions of membership to allow meaningful category conjunctions to form. For instance, perceivers may be able to cognitively combine categories like ‘‘single mother’’ or ‘‘working parent’’ (Hutter & Crisp, 2005, 2006), but they may find ‘‘single, black, educated, working mother’’ more diYcult, at least with respect to gaining any added value by processing information and making evaluative judgments heuristically (Fiske & Taylor, 1991; Hamilton, 1979; Macrae & Bodenhausen, 2001). There may come a point in considering successive categorical dimensions when using a particular categorization becomes ineVectual relative to other foci of impression formation (i.e., individuation). Supportive of this idea that there is a basic limit on the number of classification criteria people can eVectively use, Halford, Baker, McCredden, and Bain (2005) found that people were able to process information that can be decomposed along up to four crosscutting subtask dimensions. Beyond this, however, further classifications are not used to structure processing. We have argued that considering multiple criteria for social categorization could produce a decategorization eVect by virtue of the implied decrease in usefulness of any one dimension of social comparison for forming an impression (Crisp et al., 2001; Hall & Crisp, 2005). In eVect, it becomes less likely that any one category dimension will be used as a basis for discrimination. Importantly, we are not saying here that multiple pieces of information are not useful for making a judgment about someone (clearly the more information, the more accurate that judgment can be), rather that the more category dimensions that perceivers have to consider reduces the functionality of focusing on any one of those dimensions in forming their impression (then multiple pieces of information comprising a more individuated impression is used instead).
222
RICHARD J. CRISP AND MILES HEWSTONE
This decategorization eVect is typically not found when only two dimensions of categorization are combined (Ensari & Miller, 2001), probably because participants are quite able to consider two social categorizations simultaneously in memory and use them to guide evaluative judgments (Vanbeselaere, 1987, 1991). In such conditions where multiple categorization does not create a level of complexity that contravenes functionality, we expect the diVerentiation‐discrimination model we have outlined earlier (Fig. 12) to predict intergroup evaluations. However, when the number of categories increases and so does the complexity of the categorization context, we expect this model to be less applicable and for alternative processing tendencies to be apparent. As the number of category dimensions that need to be processed simultaneously increases it may be that the original categorization becomes no longer the only or most useful or eYcient way to define the target, therefore leading to a shift in orientation as outlined by Brewer’s (1988) and Fiske and Neuberg’s (1990) models, and more broadly self‐ categorization theory (Turner et al., 1987; i.e., an individuated impression provides a better fit than any single category). When perceivers experience conditions under which categorization becomes too complex to any longer provide a quick impression, then they may resort to a more individuated strategy. Correspondingly, they should be less biased in their evaluations of targets, as these evaluations are no longer tied to group membership.7 It is important to note here that, in part, the shift in attention from using a single dichotomous categorization to focusing on individuating characteristics involves both a quantitative and a qualitative shift in processing. On the one hand, category salience should be attenuated, and as such reliance on a diVerentiated in‐group–out‐group dichotomy will be reduced, similar to 7 Urada et al. (in press) have recently argued that under some conditions attempting to reconcile more than two bases for categorization in the crossed categorization paradigm can lead to a shift in processing orientation, consistent with the argument we present here. Where Urada et al. diVer from our proposals is that they argue for a shift from algebraic to feature‐ based integration, that is, a move away from trying to process each category membership (as we argue) toward forming a gestalt impression based on whether the target is more ‘‘in‐group‐like’’ or ‘‘out‐group‐like’’ overall (diVerent from our suggested individuation outcome). It is likely that both of these outcomes are possible in contexts where there are multiple pieces of information that require integration. In particular, while including more than two categories, Urada et al.’s targets comprised either more in‐group or more out‐group constituents; while, for instance, Hall and Crisp’s (2005) approach may have involved thinking about multiple identities that were related in complex ways, more akin to Brewer’s idea of interwoven identities that characterize social identity complexity. One possibility, therefore, is that perceivers may deal with more categories that are not related in a complex way by shifting to a feature‐based, gestalt impression formation process (that retains a simplified category‐based approach), while contexts requiring integration of multiple categories that are related in such a way as to push beyond a complexity threshold will lead to individuated strategies.
MULTIPLE SOCIAL CATEGORIZATION
223
what we expect when boundaries are blurred when a common in‐group identity is introduced. However, whereas introducing a common in‐group identity will improve intergroup relations by increasing the attractiveness of former out‐groups as they are incorporated into a new merged representation, decategorization associated with complex categorization will reduce bias by a qualitatively diVerent process: reduced use of any one categorization to define the self and others. As such, reductions in bias following decategorization should be driven by a decrease in in‐group evaluations, not an increase in out‐group evaluations. We will return to this issue when we discuss evidence for decategorization below. We have reported a number of studies that fulfill the conditions we believe necessary to demonstrate decategorization, and where the diVerentiation– discrimination model will not apply. We carried out an experiment (based on a paradigm used by Deschamps, 1977) using squares of increasing size, labeled either ‘‘A’’ or ‘‘B,’’ and colored either green or blue (Crisp & Hewstone, 1999b). In the ‘‘simple’’ condition all the squares were of the same color but the smallest were labeled ‘‘A’’ and the largest were labeled ‘‘B.’’ In the ‘‘crosscutting’’ condition, however, the color was perfectly orthogonal to the letter labeling (i.e., half the ‘‘As’’ were green and half were blue). Participants were required to then judge the length of the sides of each of the squares. Comparing the judged diVerence in side length between the largest of the As and the smallest of the Bs, the diVerence between the two classes was smaller in the crosscutting condition than in the simple condition. This represented basic support for the category diVerentiation process outlined earlier in this chapter. In this experiment, we also had, however, an ‘‘unsystematic’’ crosscutting condition. Here, the A–B classification was not systematically crossed with the green‐blue dimension (such that the smaller As and Bs were green, the larger As and Bs were blue), but the green‐blue classification of each A and B category was randomly allocated. This condition arguably created a more complex cross‐category context that no longer provided the best way to structure the context; here we observed just as much of a weakening of A–B interclass diVerentiation as in the systematic crosscutting categorization condition. This provides some evidence that category complexity, defined in a way that does not blur intergroup boundaries, can reduce the use of any single category dichotomy. In another study, Crisp et al. (2001) asked participants to consider five, rather than two, crosscutting and convergent bases for categorization. University aYliation was the baseline intergroup dichotomy (CardiV versus Bristol University students), but in the multiple categorization condition participants were required to form an impression of students at the out‐group university taking into account further, alternative, bases for categorization (age, major studied, gender). We measured both intergroup evaluations and
224
RICHARD J. CRISP AND MILES HEWSTONE
representations of the intergroup context: whether participants perceived themselves and the out‐group members as being ‘‘in two distinct groups’’ or ‘‘individuals,’’ or ‘‘in one overall common group’’ (Gaertner et al., 1989, 1990). Participants who thought about multiple alternative categories—whether crosscutting or converging—were less biased (using points allocations) than participants in the baseline condition (see Fig. 13, panel A). Although one might have expected converging categorizations to simply reinforce intergroup boundaries, the multiple converging categories varied on a number of characteristics (such as importance and relevance to the target ‘‘student’’ dichotomy) which should all increase complexity, decrease the usefulness of any single categorization for defining the context, and increase the cognitive eVort required to form an impression. Consistent with this idea, in both multiple in‐group and out‐group conditions there appeared to be a decategorization eVect. This eVect was indicated both by greater intracategory diVerentiation (indicated by the distance between the two groups marked on a 100‐mm line) in the multiple group conditions (see Fig. 13, panel B) and the decreased perception of the context as involving ‘‘two groups,’’ along with a heightened classification of former out‐groupers as ‘‘individuals’’ (see Fig. 14). Furthermore, this decategorization eVect mediated the observed reduction in bias. Brewer and Miller (1984) suggested that decategorization can reduce bias due to a shift in focus to individuating characteristics, rather than, as the diVerentiation–discrimination model specifies, by blurring intergroup boundaries. Decategorization involves two mutual and reciprocal cognitive processes: ‘‘diVerentiation’’ (distinctions are made between out‐group members—note this is within‐category diVerentiation, not between‐category diVerentiation as referred to earlier in this chapter) and ‘‘personalization’’ (out‐group members are seen in terms of their uniqueness and in relation to the self). The increase in intracategory diVerentiation observed by Crisp et al. (2001; Fig. 13, panel B) is consistent with the ‘‘diVerentiation’’ element of this definition, and the increase in individuated perceptions is consistent with the ‘‘personalization’’ element (Fig. 14). Indeed, participants in these conditions perceived the intergroup situation as less of a clearly dichotomous two‐group encounter, and it was replaced by an individuated, not a blurred (one‐group), representation—a shift in representation that mediated the reduction in intergroup bias. In a later series of studies, Hall and Crisp (2005) extended the paradigm used by Crisp et al. (2001) by using a free‐response paradigm and simply asking participants to think about alternative ways that they could classify a target group member other than the way that had been described (i.e., a student from a local rival university). On an evaluative measure of group
225
MULTIPLE SOCIAL CATEGORIZATION
A
1.75 1.50 1.28
Intergroup bias
1.25 1.00 0.737
0.75
0.5 0.50 0.25 0 Baseline
Multiple in-groups
Multiple out-groups
Categorization context B
20
Intragroup differentiation
18.11 18 16.87 16 14.42 14
12
10 Baseline
Multiple in-groups
Multiple out-groups
Categorization context Fig. 13. (A) Intergroup bias and (B) intragroup diVerentiation following consideration of five criteria for categorization. Data from Crisp et al. (2001; Experiment 2).
product (a drawing of a boat), we observed a reduction in bias when participants generated multiple alternative categories that could be used to define the out‐group member (see Fig. 15). Consistent with the idea that it is decategorization, not recategorization, that causes the reduced bias here, the
226
RICHARD J. CRISP AND MILES HEWSTONE
Two groups
Individuals
6
One group
5.59 5.23 4.9
4.78
5
5.44
4.45 Rating 1−7
4.17 4 3 3
2.66
2
1 Baseline
Multiple in-group
Multiple out-group
Categorization context Fig. 14. Representations of the social context following consideration of five criteria for categorization (on a scale ranging from 1, not at all to 7, very much so). Data from Crisp et al. (2001; Experiment 2).
In-group
Out-group
8.0 Intergroup evaluation
7.5
7.27
7.0 6.24
6.5 6.0
5.6
5.5
5.41
5.0 4.5 4.0 Baseline
Unrelated multiple alternatives
Categorization context Fig. 15. The impact of generating unrelated multiple alternative categories on intergroup evaluations. Data from Hall and Crisp (2005; Experiment 1).
MULTIPLE SOCIAL CATEGORIZATION
227
eVect was driven by a decrease in in‐group evaluations (see Gaertner et al., 1989). Although these findings are promising, we should note that some researchers (including one of us) have argued that decategorization is a limited strategy because shifting focus to perceive people as individuals rather than group members may simply evoke subtyping (Weber & Crocker, 1983; see also Hewstone, 1994; Johnston & Hewstone, 1992) without generalization of positive attitudes to the group as a whole (Hewstone & Brown, 1986). Typically, direct individual‐to‐group generalization is unlikely because the very conditions that promote personalization will sever the link between the exemplar and the category (see Scarberry et al., 1997). There are, however, reasons why we can be more optimistic about the benefits of decategorization for intergroup relations. Experimental studies have shown that within contact situations an interpersonal focus is more eVective at reducing bias than a task focus (Bettencourt, Brewer, Croak, & Miller, 1992); having out‐group friends also reduces bias (Hamberger & Hewstone 1997; Pettigrew 1997; Phinney, Ferguson & Tate, 1997), and these eVects can generalize to other members of the out‐group not involved directly in contact. Theoretically, there are also reasons to expect the improved attitudes observed toward decategorized individuals to lead to longer‐term positive eVects. Over time, the largely atypical out‐group members encountered could increase the perceived variability of the out‐group as a whole (Bettencourt et al., 1992). In addition, if perceivers become accustomed to decategorizing members of specific stigmatized groups, then they may over time develop a processing style that avoids automatic categorization and stereotyping, but instead sees such members more in terms of their individuating characteristics (or indeed, in terms of a collection of other, more positive, identities). One may speculate that multiple categorization‐ decategorization approaches could be linked in some way to the development of so‐called ‘‘chronic egalitarian goals’’ (Moskowitz, Gollwitzer, Wasel, & Schaal, 1999). We stress that this is not something that we would want people to adopt in relation to all categories—categorization is, of course, not inherently a bad thing—but in relation to some categories changing such chronic tendencies to perceive unwarranted negative stereotypes may prove highly valuable for reducing prejudice. To summarize, both Crisp et al. (2001) and Hall and Crisp (2005) support the hypothesis that when multiple categorization requires the integration of complex information to form a category‐based impression, then decategorization can occur. Below we review some related work that adds support for the idea that multiple categorization can, under a further set of circumstances, evoke a shift in the way impressions of others are formed.
228
RICHARD J. CRISP AND MILES HEWSTONE
D. CATEGORY CONJUNCTIONS As noted above, we have found that multiple categorization can be eVective at reducing bias when perceivers are required to integrate suYciently complex and disparate pieces of information to form an impression of the target. This idea that decategorization occurs only when perceivers need to seek alternative classifications gains some further support from work on integrating congruent or incongruent categorizations. For categories to be classed as either congruent or incongruent they must first fall along dimensions that are correlated. For example, if one encounters a black female then this person is categorized along two criteria—race and gender. However, these criteria are uncorrelated. Therefore, according to the diVerentiation–discrimination model outlined in the first half of this chapter, this person will be evaluated either positively or negatively as a linear additive function of the degree of shared in‐group membership with the perceiver (notwithstanding any aVective or motivational moderating factors). When the categories used are, however, correlated, then we may expect, for particular combinations, a decategorization eVect. If, for example, the target person was a female engineer (versus a male engineer), this represents an incongruent, surprising and nonnormative combination of category memberships, irrespective of self‐inclusion or aYliation. Such surprising combinations of category membership should, in the same way as category complexity, promote a shift in processing from a category‐based to an individuated mode of perception (Crisp et al., 2001; Hall & Crisp, 2005; although see Rothbart & John, 1985, who suggest one of the categories in such combinations will simply be ignored). Focusing on category combinations formed from correlated dimensions can help us to uncover further the processes associated with decategorization.
E. SURPRISING COMBINATIONS In some work on the eVects of combining nonsocial concepts, the surprising or incongruent nature of the combined concepts appears to produce emergent attributes (Hampton, 1997; Murphy, 1988; Wattenmaker, 1995; Wilkenfield & Ward, 2001). Emergent attributes are attributes ascribed to the combination that are independent of attributes associated with either of the simple constituents. Such emergent attributes have also been found when people try to combine incongruent nonsocial categories (Hastie, 1981; Hastie, Schroeder, & Weber, 1990; Kunda, Miller, & Claire, 1990). In Kunda et al.’s famous example (1990), a Harvard‐educated carpenter may be perceived as non‐ materialistic, but this is not something that springs to mind when describing
MULTIPLE SOCIAL CATEGORIZATION
229
someone who is either Harvard educated or a carpenter. It is an emergent property of the combination. The important thing about emergent attributes is what they signify. When observed this emergence suggests that impressions have been constructed in a way that draws not on schematic information stored in long‐term memory, but on an alternative process leading to the generation of nonstereotypic attributes. In other words, it is not categorization that has been used in impression formation, but individuation, consistent with decategorization eVects discussed above, and both Brewer’s (1988) and Fiske and Neuberg’s (1990) models of impression formation. In some studies, we built on this initial theorizing and empirical work to specify in more detail the processes involved in perceiving surprising category combinations (Hutter & Crisp, 2005, 2006). We developed a task in which we could quantify the emergent or constituent attributes generated when encountering either a congruent or incongruent combination of categories. This was similar to the methods used by Hastie et al. (1990) but also allowed us to test whether considering category conjunctions may also have eVects on the proportion of constituent attributes ascribed to the combination. It is possible that the conflict between constituent social categories, for surprising combinations, will not only lead to emergent attributes, but also reduced application of attributes associated with the conflicting constituents. Observing a reduction in the ascription of constituent attributes is further evidence for individuation and for the idea that complex categorization can reduce bias (in the form of reduced application of stereotypic attributes). In three studies, using varied target group combinations, we found that relative to congruent combinations, incongruent combinations (female mechanics, Oxford‐educated bricklayers) were made up of fewer constituent attributes and more emergent attributes (a qualitatively distinct representation, Hutter & Crisp, 2005). The flip side of this is that correlated categories also form unsurprising or stereotype confirming combinations such as male mechanic or female nurse. Hutter and Crisp (2006) found that participants’ impressions of these combinations were based on essentially stereotypical attributes, a finding that mirrors previous work by Eurich‐Fulcher and Schofield (1995) that found no decrease in biased evaluations when correlated categories are crossed. This suggests, as with many of the approaches described in this chapter, that considerable care must be taken in implementation of interventions that make use of correlated category dimensions. Hastie et al. (1990) conceptualized the processes described above as inconsistency resolution. It seems likely that both the emergence of new attributes and curtailing of constituents can be outcomes of such an inconsistency resolution process—an ‘‘executive function’’ (Pineda, 2000; see also Macrae,
230
RICHARD J. CRISP AND MILES HEWSTONE
Bodenhausen, Schloerscheidt, & Milne, 1999; Srull & Wyer, 1989). Again, this maps directly onto the idea that complex categorization can cause a decategorization eVect where perceivers shift to a more cognitively eVortful individuating strategy of impression formation. Consistent with this idea, Hutter and Crisp (2006) have shown that cognitive load wipes out the generation of emergent properties for surprising combinations.
F. SKILLS DEVELOPMENT From the above review of studies we have ascertained that under some conditions multiple categorization can lead to a qualitatively diVerent process from that outlined in our diVerentiation–discrimination model. In particular, when the number and/or complexity of the categories that require processing increases (where complexity is defined as the unrelated or surprising nature of orthogonally crossed categorizations), there appears to be a shift from categorical to individuated perception that is due, apparently, to the perception that categories can no longer fulfill a useful role. This shift in processing leads to reduced stereotyping and an increase in the perception of individuating or emergent attributes (Hutter & Crisp, 2005), less bias (Crisp et al., 2001; Hall & Crisp, 2005), less category use (Crisp et al., 2001), but it requires cognitive resources (Hutter & Crisp, 2006). We have argued, above, that although individuation is clearly beneficial for the specific target at hand, it will probably lead to subtyping of that individual and no generalized change in the overall stereotype of the out‐group category. However, over time, as people repeatedly think about others in terms of their individuating, rather than out‐ group, characteristics, this may eventually increase perceived out‐group heterogeneity. This will reduce stereotyping and thus debias the overall structure of particular out‐group categorizations (Bettencourt et al., 1992; Ensari & Miller, 2001). But we can also intervene actively and try to train people that single categories are not so useful in specific circumstances and that adopting a more individuated mode of processing in such situations may be more appropriate. We again stress that we are not suggesting that it is either desirable, or indeed possible, to discourage the use of categorization in all situations. Clearly, there are many situations in which categorization of the self and others is both desirable and useful. But when categorization is divisive (such as when a particular category has an unwarranted stigma associated with it), the type of multiple categorization intervention discussed here may be beneficial. In such contexts, discouraging heavy reliance on simple dichotomous categories (and instead appreciating the multiplicity of group life) may be preferable. Of course, it is unlikely, as with any intervention, that multiple category‐inspired decategorization will work immediately or in isolation,
MULTIPLE SOCIAL CATEGORIZATION
231
especially with ingrained, pervasive, and realistic conflicts. However, it may be able to play an important part of an integrated intervention strategy, perhaps providing a first step in encouraging less reliance on previously dominant categories for defining intergroup relations. Some work on the development of categorization skills adds support to this favorable prognosis for the development of intervention strategies aimed at decategorization. According to cognitive‐developmental theory, the categorization process is somewhat rigid during early childhood, particularly regarding the ability to classify objects into categories (Piaget, 1965). Young preschoolers are, for example, typically unable to sort objects consistently along a single dimension. Fortunately, developmental increases in cognitive flexibility gradually enable children to classify objects into increasingly sophisticated categories. Older preschoolers are generally able to sort consistently along one dimension, but cannot re‐sort the same objects along an alternative dimension (Bigler & Liben, 1992). By early primary school, children are usually able to perform successive object sorts, but continue to view category membership as exclusive at any given moment. Eventually, however, children transcend the view of category membership as exclusive, and can appreciate that one object may fall into two or more categories simultaneously (Bigler & Liben, 1992). Previous research (Bigler & Liben, 1993) has shown that multiple classification skills not only vary as a function of age, but that there are demonstrable individual diVerences in the ability (Averhart & Bigler, 1997). In order to try to improve this ability, Bigler and Liben (1992) created a task through which primary school children were taught to classify information along multiple dimensions. Every day, for a period of 1 week, participants were given one set of 12 pictures, of men and women engaging in stereotypically feminine (e.g., hairstylist, secretary) and stereotypically masculine (e.g., construction worker, truck driver) occupations. Children then practiced sorting these pictures, first by gender and then by occupation. Next, they were asked to sort the pictures into four piles so that matching groups of gender and occupation were formed along the matrix rows and columns. Following this stage, the participant was shown the correct 2 2 matrix (i.e., female/male manual worker/nonmanual worker) created by the experimenter, who explained the category dimensions in the matrix rows (i.e., all female or all male) and columns (i.e., all manual workers or all nonmanual workers). The experimenter then removed the stimulus from one cell of the matrix, and participants were asked to identify, from their own set of pictures, a picture that belonged in the empty quadrant of the matrix, and to justify their choice. Participants continued to practice the matrix completion task until all children had given at least one correct answer. Bigler and Liben (1992) tested their multiple classification training method on 5‐ to 10‐year‐old children attending a summer school program.
232
RICHARD J. CRISP AND MILES HEWSTONE
Participants were trained using either social stimuli, as described above, or nonsocial stimuli, learning to sort shoes and hats that diVered in color. The researchers found that, before the training, less than 3% of participants were competent at multiple classification. After training, however, 95% trained with social stimuli and 68% trained with nonsocial stimuli were able to multiply classify stimuli. Notably, participants who had acquired multiple classification skills during training with social stimuli showed significantly less gender stereotyping than participants who had not acquired multiple classification skills (with less pronounced benefits arising from training with nonsocial stimuli). Multiple classification ability can be thought of as the capacity to understand that the same person can belong to two categories simultaneously. That this ability may vary, either developmentally, as a function of training, or within individuals, may have significant implications for social perception and behavior. This is because a tendency to categorize along multiple dimensions of categorization, as opposed to a single dimension, has direct implications for social categorization and stereotyping. If, for example, a child encounters a ‘‘woman engineer,’’ because ‘‘woman’’ and ‘‘engineer’’ are typically seen as discrepant categories, a child who cannot classify an individual according to gender and occupation simultaneously may be unable to process—and incorporate into their worldview—the category combination ‘‘woman engineer.’’ Being able to think about inconsistent combinations of categories, by definition, implies less reliance on homogeneous stereotypes in forming impressions of others (see Hutter & Crisp, 2005). Consistent with this view, Bigler and Liben (1992) found that children with more advanced multiple classification skills gave less gender stereotyped responses than did those with less advanced classification skills, even when age and verbal skills were controlled. This adds to the general support for the view that interventions to reduce prejudice at school level can be eVective (Aboud & Fenwick, 1999). We should note here that, of course, the surprise someone experiences when meeting, for instance, a woman engineer, may be understandable—it reflects social reality (there are less women than men engineers). The point of training in multiple category skills is not to make people insensitive to this social reality, but to prevent the recognition of there being fewer women engineers from becoming the view that women should not be engineers. The idea is that multiple classification skills can broaden expectations about what should be, not prevent people recognizing what is. Encouraging people to learn to classify along multiple dimensions, to develop a more complex conceptualization of social identities, is an attempt to improve potential, not change how people use categorization on some fundamental level (and we would never claim that this is, indeed, even possible). Of course, basic categorization processes are inherently flexible; we all have the potential to
MULTIPLE SOCIAL CATEGORIZATION
233
classify along multiple criteria. Our argument is that context, experience, goals, and motives do not always let us see the multiple ways in which we can see ourselves and others. Especially in societies with high segmentation and convergent category boundaries, even if people recognize the multiplicity of identities only briefly, this may be the start of a process that eventually leads to dominant intergroup divides becoming less defining in conflictual intergroup relations. In fact, when we consider the model in its entirety, we can see that decategorization may be especially useful in contexts where there is a high level of hostility that defines the intergroup relations. As we noted earlier, when negative aVect is associated with the intergroup context, this is likely to prime out‐group identities, leading, for instance, to the social exclusion pattern of evaluations where even groups that shared a common in‐group identity with perceivers (io, oi) are evaluated as negatively as double out‐groups (oo). We can make a link here with research on contact that has revealed the negative impact of intergroup anxiety for intergroup relations (Islam & Hewstone, 1993). In such cases, attempting to impose shared identities will not be eVective because the dominant out‐group characterization of targets will override any possible benefits of shared categorization. Instead, decategorization may here provide the best means of at least initially encouraging social thought and interaction on the basis of criteria other than the dominant and divisive social categorization. Focusing on changing people’s cognitive style—how categorical information is processed—may be important because it avoids a potential criticism of existing models of stereotype change. That is, if categories and stereotypes are, in fact, emergent properties of context, rather than fixed and static (see Smith, 2006; Smith & DeCoster, 1998; Turner et al., 1987), then interventions that focus on changing only content may be ineVective outside of the immediately relevant laboratory context, even with repeated exposure. A focus on changing the way that people use categories, in contrast, may be far more generalizable to multiple situations and domains, and tackle those very contexts in which goals, experience, values, and motives have contrived to make categorization inflexible, and negative stereotypes associated with particular categories chronically accessible (see Bargh, 1997; Turner, 1999). The finding that multiple classification training can lead to changes in the way in which people categorize supports the eYcacy of decategorization approaches to debiasing social perception as outlined above (see Crisp & Hewstone, 2006, for a discussion of the eVects of multiple categorization on devolved category constituents). We now incorporate the idea that multiple categorization can lead to qualitatively diVerent processes, and reduced bias in diVerent ways, in a model that fully integrates the predictions made by our diVerentiation model and the decategorization predictions outlined in this section.
234
RICHARD J. CRISP AND MILES HEWSTONE
V. Conclusions A. ‘‘SOLITARIST’’ VERSUS ‘‘PLURALIST’’ APPROACHES TO HUMAN IDENTITY Amartya Sen, winner of the 1998 Nobel Prize in Economics, has recently published an elegant essay that is, essentially, in firm agreement with the position we have adopted in this chapter. Focusing especially on categorization in terms of ‘‘civilization’’ or religion, Sen argues that categorizing and identifying others simply in terms of one such dimension constitutes a ‘‘solitarist’’ approach to human identity. He is adamantly opposed to this approach and refers to ‘‘the appalling eVects of the miniaturization of people’’ (Sen, 2006, p. xvi). Moreover, Sen ties this approach to the fomenting of sectarian conflict and argues that ‘‘a major source of potential conflict in the contemporary world is the presumption that people can be uniquely categorized based on religion or culture.’’ (Sen, 2006, p. xv). One of Sen’s major targets is Samuel Huntington’s (1996) book, The Clash of Civilizations and the Remaking of World Order. According to Sen, Huntington’s argument, which focuses on religious classification (especially the contrast between Muslim versus non‐Muslim worlds) has, among others, the weaknesses that it overlooks the extent of internal diversities within civilizational categories (what we would call within‐category diVerentiation), that it overlooks the many alternative dimensions of identity relating to Muslims, and that it ignores those Muslims who happen to be non‐religious. As Sen puts it, ‘‘it is extremely odd that those who want to overcome the tensions and conflicts linked with Islamic fundamentalism also seem unable to see Muslim people in any form other than their being just Islamic . . . rather than seeing the many‐dimensional nature of diverse human beings who happen to be Muslim’’ (pp. xiv–xv). Sen’s book does a superb job of illustrating why these issues matter, and why multiple categorization should form a central plank of interventions to improve intergroup relations and promote peacebuilding. Our only point of diVerence, indeed regret, which we firmly believe is more than mere cavilling, is that Sen’s scholarly essay, which draws widely on a range of disciplines and devotes more than half a column of one index page to issues concerning ‘‘identity,’’ neither cites nor draws on a single source of social psychology. What we have tried to do in this chapter is illustrate what we know, at present, about multiple categorization, organize it theoretically, and critically review the extensive empirical evidence. Below we outline a detailed model of our current understanding of processes concerning multiple categorization and then we end with some closing remarks.
235
MULTIPLE SOCIAL CATEGORIZATION
B. A DIFFERENTIATION–DECATEGORIZATION MODEL OF MULTIPLE CATEGORIZATION EFFECTS The final, complete, version of our model in Fig. 16 argues that there are two basic consequences of multiple categorization: those associated with diVerentiation and those associated with decategorization. The model begins with a central, basic mechanism: Categorization causes category diVerentiation (e.g., black versus white, British versus French) and this diVerentiation leads to bias. Reducing diVerentiation by making crosscutting and common in‐ group identities salient (e.g., black British versus white British) can reduce intergroup bias. There are moderating factors that can accentuate or attenuate this positive relationship. Anything that increases positive mood and reduces negative aVect (e.g., intergroup anxiety) should enhance the eVectiveness of promoting common in‐group identities. However, one moderator, in‐group identification, reverses this relationship; diVerentiation becomes negatively correlated with bias in common in‐group contexts, and to avoid reactionary increases in bias it is necessary to maintain subgroup distinctiveness. For instance, emphasizing European integration for high identifying Britons is unlikely to improve attitudes toward adopting this
Negative affect/ out-group prime −
Lower
+ Low in-group identification
Convergent categories (II versus OO)
+
Crosscutting categories (II versus IO)
Intergroup bias
− High in-group identification +
Category complexity
Higher
+
Intergroup differentiation
−
Positive affect/ in-group prime Number and/or crosscutting correlated categories
− +
Intergroup bias −
Decategorization +
Skill development
Fig. 16. A diVerentiation–decategorization model of multiple categorization eVects.
236
RICHARD J. CRISP AND MILES HEWSTONE
common in‐group identity; but promoting integration while emphasizing a maintained British identity should lead to a more positive outcome. Apart from diVerentiation, multiple categorization can lead to decategorization, that is a change in the focus of impression formation such that more individuated perceptions are evoked. This tendency operates when the complexity of the categories that require processing increases (where complexity is operationalized as any combination of characteristics that decreases the usefulness of single dichotomous categorization as a means of defining the target, or increases the cognitive eVort required to integrate the available information to form an impression). When complexity is high, for instance when there are a large number of categories, and/or those categories are incongruent (e.g., woman þ mechanic, male þ midwife), there is a shift from categorical to individuated perception due to the perception that no one category can provide an eYcient and meaningful definition of the social context. This change in processing mode requires cognitive eVort, leads to more individuated and less biased perceptions, less use of constituent stereotypes to define social targets, and the use of divergent thought and emergent attributes. Furthermore, while subtyping may occur on initial training sessions, as with diVerentiation‐based methods, a number of intervention sessions will be needed to see an improvement in intergroup relations. Training in such multiple classification skills improves children’s ability to be nonbiased, but it remains to be seen whether such interventions work for adults as well. Our inclusion of a line between decategorization and skill development indicates that this is a potential outcome of decategorization, but one that is likely contingent on both time and a range of as yet unspecified conditions. The model presented in Fig. 16 provides, we believe, a comprehensive summary of what we currently know about multiple social categorization. As we have tried to illustrate, there have been a considerable range of approaches used to study how multiple identities aVect social thought and action. With this model we have attempted to integrate this diverse literature and show when and how these diVerent approaches interrelate. With this integration, however, come new questions. Which of the categorization strategies embodied within the model should be used, in which contexts, and when? We believe these are important issues for future research, and, drawing on work by others, we can oVer some suggestions. The diVerentiation–decategorization model argues that under diVerent conditions, diVerent ways of making multiple identities salient will be maximally eVective. The diVerent strategies that comprise our overall model should not, however, be taken as mutually exclusive. Such strategies can be implemented simultaneously or sequentially (see Pettigrew, 1998). As Gaertner et al. (2000) note in their reconsideration of findings from Sherif
MULTIPLE SOCIAL CATEGORIZATION
237
et al.’s (1961) classic Robbers Cave studies, the order in which diVerent strategies of bias reduction should be implemented will depend on the specific characteristics of the intergroup context. They argue, for instance, that while for more implicit forms of bias recategorization may be a good starting point, decategorization will be the best place to start when intergroup relations are characterized by high levels of overt hostility (to enable personalized interactions to set the basis for developing cross‐group friendships without inhibition associated with divisive category salience). Initially the many diVerent ways in which out‐group membership can be classified should be emphasized, especially if those diVerent classifications are surprising and disconfirm stereotypes (e.g., women can be mechanics too). This decategorization of individual group members may, over time, then generalize to perceptions of the group as a whole. After initial decategorization training one might then incorporate strategies that encourage perceivers to think about the similarities and diVerences between categories diVerently, in particular by emphasizing crosscutting or convergent bases for categorization. A positive mood and a context that favors integration (both factors that prime in‐group identities) will enhance the impact of shared common in‐ group identities. When identification with subgroups is high, however, dual identification and simultaneous categorization may be the best way to initially avoid distinctiveness threat (Brewer, 1991). Cooperative interaction will work best if it is accompanied by a common in‐group identity that does not diminish the importance assigned to subgroup categories. An illustration with clear practical implications is presented by Schofield and McGivern (1979) who argue against a color‐blind perspective in desegregated schools and for the creation of ‘‘new, supplemental group identities which cut across racial boundaries’’ (p. 107). Their observational study reports on the development of a positive group identity shared by black and white students in the form of a student council. Slavin’s (1985) review of cooperative learning indicates that such methods automatically give black and white students a ‘‘common identity’’ by assigning them to the same team, but they probably also work by retaining the salience of subgroup ethnic identities. Such issues have direct relevance for our model of multiple categorization, which involves complexity‐induced decategorization, and recategorization versus crossed categorization/simultaneous categorization as ways of reducing bias for lower and higher identifiers. Assessing the contexts within which diVerent orders of implementation will be eVective will be an important endeavor for future work. Finally, we should also note that while we believe the model we present here encompasses much of the empirical evidence from social psychological research on multiple categorization there are other ways to think about
238
RICHARD J. CRISP AND MILES HEWSTONE
multiple identities. Phinney and Alipuria (2006) point out that most social psychological research on multiple categorization has focused on simultaneous categorization in terms of two (or more) diVerent identity domains. Thus, for example, Hewstone et al. (1993) crossed religion (Hindu versus Muslim) and nationality (Bangladesh versus India), or religion, nationality, and language (Studies 1 and 2, respectively). Phinney and Alipuria, however, study a diVerent issue, that of multiple categorization within a single identity domain. This involves an analysis of individuals who are not black versus white, or Hispanic versus American, but rather black and white, or Hispanic and American. Phinney and Alipuria focus on three types of multiracial, multiethnic, or multicultural individuals: those who have parents from two diVerent ethnic or racial groups; those who have parents who both come from one and the same culture, but themselves grew up in another culture; and those who themselves may be either multiracial or monoethnic, but who incorporate into their identity a culture other than that of their parents. These individuals do not (have to) balance the importance or relevance of two distinct characteristics of the self, but rather they integrate and manage internal complexity, which may be potentially conflicting but also potentially enriching. It is possible to see some links between this work and the type of multiple categorization we have discussed in this chapter. The relationship between bicultural individuals and the host community is also analogous to the mixed in‐group in crossed categorization research, whereby the out‐ group is now a mixed in‐group/out‐group on a second dimension, the superordinate group. Thus in the case of a minority group member, the host society may be out‐group because they are white, but in‐group because the minority group members and the host group members are all inhabitants of the United Kingdom. Consequently, they may experience less prejudice from the host group than do full out‐group members. The study of biculturalism may have some important implications for how we think about multiple categorization and we believe it is an area that should be highlighted for future research (see Verkuyten, 2005, for an extended discussion of related issues).
C. EPILOGUE In this chapter, we have outlined an integrative model of multiple categorization eVects. Our model goes beyond previous attempts to bring coherence to this literature by specifying underlying processes that can apply to all areas of research within this domain. Previous theoretical statements, which have focused only on specific types of multiple categorization (Crisp, 2006; Crisp et al., 2002; Hewstone et al., 1993; Urban & Miller, 1998), have been incorporated here within our overall model, and we have proposed
MULTIPLE SOCIAL CATEGORIZATION
239
a distinction between diVerentiation and decategorization processes in multiple categorization depending on category complexity. In an address (November 30, 2006, Keele University) Trevor Phillips, Chair of the Commission for Racial Equality in the United Kingdom, highlighted the slow rate of progress in making British society more equal. He argued that extrapolating from the present rate of progress, it would take until 2080 to elect a culturally representative House of Commons, 2085 to close the gender pay gap, and until 2105 to close the ethnic employment gap. We argue that an increasing recognition of the utility of multiple social categorization can begin to inform policy and practice relating to improving intergroup relations, reducing social exclusion, and promoting egalitarianism. More research is essential, but the integration of existing work we have presented may, we hope, help us to develop improved interventions, and achieve such goals as outlined by Trevor Phillips considerably more quickly than current projections might suggest.
Acknowledgments The publication of this chapter gives us an opportunity to bring together our respective independent and joint research programs over many years. We gratefully acknowledge various sources of funding, including the Economic and Social Research Council (R000239382) (to both authors); British Academy grants SG‐31041 and SG‐34045, and Leverhulme Trust grant (F/00094/H) (to Richard Crisp); and grants from the Community Relations Unit (Northern Ireland), the Templeton Foundation, and the Russell Sage Foundation (to Miles Hewstone). We also thank the graduate students, research staV, and other collaborators who have made major contributions to this endeavor; these are made explicit in authorship of primary studies, but we especially thank the late Brian Mullen, Ed Cairns, Mir Rabiul Islam, Chick Judd, Michael Migdal, Jared Kenworthy, Mark Rubin, and Terri Vescio, as well as Laura Bache, Rowena Cocker, Matthew Farr, Natalie Hall, Russell Hutter, Graham MoYt, Harriet Rosenthal, Sofia Stathi, Catriona Stone, Mein‐woei Suen, and Rhiannon Turner. Finally, we thank all of the following for commenting constructively on an earlier draft: Marilynn Brewer, Sam Gaertner, Jolanda Jetten, Jared Kenworthy, Norman Miller, Mark Rubin, and Douglas Stenstrom. This chapter received an honorable mention for the Gordon Allport Intergroup Relations Prize (2006) of the Society for the Psychological Study of Social Issues (SPSSI).
References Aboud, F. E., & Fenwick, V. (1999). Exploring and evaluating school‐based interventions to reduce prejudice. Journal of Social Issues, 55, 767–786. Abrams, D., & Hogg, M. A. (1998). Prospects for research in group processes and intergroup relations. Group Processes and Intergroup Relations, 1, 7–20.
240
RICHARD J. CRISP AND MILES HEWSTONE
Allen, V. L., & Wilder, D. A. (1975). Categorization, belief similarity, and intergroup discrimination. Journal of Personality and Social Psychology, 32, 971–977. Allen, V. L., Wilder, D. A., & Atkinson, M. L. (1983). Multiple group membership and social identity. In T. R. Sarbin & K. E. Scheibe (Eds.), Studies in social identity (pp. 92–115). New York: Praeger. Allport, G. W. (1954). The nature of prejudice. Garden City, NY: Doubleday. Arcuri, L. (1982). Three patterns of social categorization in attribution memory. European Journal of Social Psychology, 12, 271–282. Averhart, C. J., & Bigler, R. S. (1997). Shades of meaning: Skin tone, racial attitudes, and constructive memory in African American children. Journal of Experimental Child Psychology, 67, 363–388. Banker, B.S, & Gaertner, S. L. (1998). Achieving stepfamily harmony: An intergroup‐relations approach. Journal of Family Psychology, 12, 310–325. Bargh, J. A. (1997). The automaticity of everyday life. In R. Wyer, Jr. (Ed.), The automaticity of everyday life: Advances in social cognition (Vol. 10, pp. 1–61). Mahwah, NJ: Erlbaum. Bargh, J. A., & Pratto, F. (1986). Individual construct accessibility and perceptual selection. Journal of Experimental Social Psychology, 22, 211–293. Beach, S. R. H., & Tesser, A. (1995). Self‐esteem and the extended self‐evaluation maintenance model: The self in social context. In M. Kernis (Ed.), EYcacy, agency, and self‐esteem (pp. 145–170). NewYork: Plenum. Bettencourt, B. A., Brewer, M. B., Croak, M., & Miller, N. (1992). Cooperation and the reduction of intergroup bias: The role of reward structure and social orientation. Journal of Experimental Social Psychology, 28, 301–319. Bigler, R. S., & Liben, L. S. (1992). Cognitive mechanisms in children’s gender stereotyping: Theoretical and educational implications of a cognitive‐based intervention. Child Development, 63, 1351–1363. Bigler, R. S., & Liben, L. S. (1993). A cognitive‐developmental approach to racial stereotyping and reconstructive memory in Euro‐American children. Child Development, 64, 1507–1518. Billig, M., & Tajfel, H. (1973). Social categorization and similarity in intergroup behaviour. European Journal of Social Psychology, 3, 27–52. Blake, R. R., Shepard, H. A., & Mouton, J. S. (1964). Managing intergroup conflict in industry. Texas: Gulf Publishing Co. Bloom, Z., & Crisp, R. J. (2004). Social identification and intergroup bias in multiple membership contexts. Unpublished Dataset, University of Birmingham, UK. Bodenhausen, G. V. (1992). Information processing functions of generic knowledge structures and their role in context eVects in social judgment. In N. Schwarz & S. Sudman (Eds.), Context eVects in social and psychological research (pp. 267–278). New York: Springer‐Verlag. Bodenhausen, G. V. (1993). Emotions, arousal, and stereotypic judgments: A heuristic model of aVect and stereotyping. In D. M. Mackie & D. L. Hamilton (Eds.), AVect, cognition, and stereotyping: Interactive processes in group perception (pp. 13–37). San Diego, CA: Academic Press. Bourhis, R. Y., Sachdev, I., & Gagnon, A. (1994). Intergroup research with the Tajfel matrices: Methodological notes. In M. P. Zanna & J. M. Olson (Eds.), The psychology of prejudice: The Ontario symposium (Vol. 7, pp. 209–232). Hillsdale, NJ: Erlbaum. Brewer, M. B. (1979). Ingroup bias in the minimal intergroup situation: A cognitive‐ motivational analysis. Psychological Bulletin, 86, 307–324. Brewer, M. B. (1988). A dual process model of impression formation. In R. Wyer & T. Srull (Eds.), Advances in social cognition (Vol. 1, pp. 1–36). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
MULTIPLE SOCIAL CATEGORIZATION
241
Brewer, M. B. (1991). The social self: On being the same and diVerent at the same time. Personality and Social Psychology Bulletin, 17, 475–482. Brewer, M. B. (1993). Social identity, distinctiveness, and ingroup homogeneity. Social Cognition, 11, 150–164. Brewer, M. B. (2000). Superordinate goals versus superordinate identity as bases of intergroup cooperation. In D. Capozza & R. Brown (Eds.), Social identity processes: Trends in theory and research (pp. 117–132). Thousand Oaks, CA: Sage. Brewer, M. B. (2001). Ingroup identification and intergroup conflict: When does ingroup love become outgroup hate? In R. Ashmore, L. Jussim, & D. Wilder (Eds.), Social identity, intergroup conflict, and conflict reduction (pp. 17–41). New York: Oxford University Press. Brewer, M. B., & Brown, R. J. (1998). Intergroup relations. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 2, pp. 554–594). New York: McGraw‐Hill. Brewer, M. B., & Campbell, D. T. (1976). Ethnocentrism and intergroup attitudes: East African evidence. New York: Halstead Press. Brewer, M. B., Dull, L., & Lui, L. (1981). Perceptions of the elderly: Stereotypes as prototypes. Journal of Personality and Social Psychology, 41, 656–670. Brewer, M. B., & Gaertner, S. L. (2001). Toward reduction of prejudice: Intergroup contact and social categorization. In R. Brown & S. L. Gaertner (Eds.), Blackwell handbook of social psychology: Intergroup processes (pp. 451–474). Oxford: Blackwell. Brewer, M. B., & Gardner, W. (1996). Who is this ‘‘We’’? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71, 83–93. Brewer, M. B., & Harasty Feinstein, A. S. (1999). Dual processes in the cognitive representation of persons and social categories. In S. Chaiken & Y. Trope (Eds.), Dual process theories in social psychology (pp. 255–270). New York: Guilford. Brewer, M. B., Ho, H., Lee, J., & Miller, N. (1987). Social identity and social distance among Hong Kong school children. Personality and Social Psychology Bulletin, 13, 156–165. Brewer, M. B., & Miller, N. (1984). Beyond the contact hypothesis: Theoretical perspectives on desegregation. In N. Miller & M. B. Brewer (Eds.), Groups in contact: The psychology of desegregation (pp. 281–302). San Diego, CA: Academic Press. Brewer, M. B., & Pierce, K. P. (2005). Social identity complexity and outgroup tolerance. Personality and Social Psychology Bulletin, 31, 428–437. Brown, R. J. (1984). The role of similarity in intergroup relations. In H. Tajfel (Ed.), The social dimension: European developments in social psychology (pp. 603–623). Cambridge: Cambridge University Press. Brown, R. J., & Abrams, D. (1986). The eVects of intergroup similarity and goal interdependence on intergroup attitudes and task performance. Journal of Experimental Social Psychology, 22, 78–92. Brown, R. J., & Hewstone, M. (2005). An integrative theory of intergroup contact. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 37, pp. 255–343). Orlando, FL: Academic Press. Brown, R. J., & Turner, J. C. (1979). The criss‐cross categorization eVect in intergroup discrimination. British Journal of Social and Clinical Psychology, 18, 371–383. Brown, R. J., & Turner, J. C. (1981). Interpersonal and intergroup behaviour. In J. C. Turner & H. Giles (Eds.), Intergroup behaviour (pp. 33–65). Oxford: Blackwell. Brown, R. J., Vivian, J., & Hewstone, M. (1999). Changing attitudes through intergroup contact: The eVects of group membership salience. European Journal of Social Psychology, 29, 741–764.
242
RICHARD J. CRISP AND MILES HEWSTONE
Brown, R. J., & Wade, G. (1987). Superordinate goals and intergroup behavior—the eVect of role ambiguity and status on intergroup attitudes and task‐performance. European Journal of Social Psychology, 17, 131–142. Cacioppo, J. T., Marshall‐Goodell, B. S., Tassinary, L. G., & Petty, R. E. (1992). Rudimentary determinants of attitudes: Classical conditioning is more eVective when prior knowledge about the attitude stimulus is low than high. Journal of Experimental Social Psychology, 28, 207–233. Cadinu, M. R., & Rothbart, M. (1996). Self‐anchoring and diVerentiation processes in the minimal group setting. Journal of Personality and Social Psychology, 70, 661–677. Cairns, E., & Mercer, G. W. (1984). Social identity in Northern Ireland. Human Relations, 37, 1095–1102. Campbell, D. T. (1956). Enhancement of contrast as a composite habit. Journal of Abnormal and Social Psychology, 53, 350–355. Carter, N. A. (2006). Political institutions and multiple social identities. In R. J. Crisp and M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications (pp. 239–268). Hove, E. Sussex: Psychology Press (Taylor & Francis). Chaiken, S. & Trope, Y. (Eds.) (1999). Dual process theories in social psychology. New York: Guilford. Coser, L. A. (1956). The function of social conflict. New York: Free Press. Crisp, R. J. (2002). Social categorization: Blending the boundaries. The Psychologist, 15, 612–615. Crisp, R. J. (2006). Commitment and categorization in common ingroup contexts. In R. J. Crisp & M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications (pp. 90–111). Hove, E. Sussex: Psychology Press (Taylor & Francis). Crisp, R. J., & Beck, S. R. (2005). Reducing intergroup bias: The moderating role of ingroup identification. Group Processes and Intergroup Relations, 8, 173–186. Crisp, R. J., Beck, S. R., & Hewstone, M. (2006a). Inclusiveness and crossed categorization. Unpublished manuscript, University of Birmingham, UK. Crisp, R. J., Ensari, N., Hewstone, M., & Miller, N. (2002). A dual‐route model of crossed categorization eVects. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 13, pp. 35–74). Hove, E. Sussex: Psychology Press (Taylor & Francis). Crisp, R. J., & Hewstone, M. (1999a). DiVerential evaluation of crossed category groups: Patterns, processes, and reducing intergroup bias. Group Processes and Intergroup Relations, 2, 303–333. Crisp, R. J., & Hewstone, M. (1999b). Subcategorization of physical stimuli: Category diVerentiation and decategorization processes. European Journal of Social Psychology, 29, 665–671. Crisp, R. J., & Hewstone, M. (2000a). Crossed categorization and intergroup bias: The moderating roles of intergroup and aVective context. Journal of Experimental Social Psychology, 36, 357–383. Crisp, R. J., & Hewstone, M. (2000b). Multiple categorization and social identity. In D. Capozza & R. Brown (Eds.), Social identity processes: Trends in theory and research (pp. 149–166). Beverly Hills: Sage. Crisp, R. J., & Hewstone, M. (2001). Multiple categorization and implicit intergroup bias: DiVerential category dominance and the positive‐negative asymmetry eVect. European Journal of Social Psychology, 31, 45–62. Crisp, R. J., & Hewstone, M. (2006). Multiple social categorization: Context, process, and social consequences. In R. J. Crisp & M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications. Hove, E. Sussex: Psychology Press (Taylor & Francis).
MULTIPLE SOCIAL CATEGORIZATION
243
Crisp, R. J., Hewstone, M., & Cairns, E. (2001). Multiple identities in Northern Ireland: Hierarchical ordering in the representation of group membership. British Journal of Social Psychology, 40, 501–514. Crisp, R. J., Hewstone, M, Richards, Z., & Paolini, S. (2003). Inclusiveness and crossed categorization: EVects on co‐joined category evaluations of ingroup and outgroup primes. British Journal of Social Psychology, 42, 25–38. Crisp, R. J., Hewstone, M., & Rubin, M. (2001). Does multiple categorization reduce intergroup bias? Personality & Social Psychology Bulletin, 27, 76–89. Crisp, R. J., Stone, C. H., & Hall, N. R. (2006b). Recategorization and subgroup identification: Predicting and preventing threats from common ingroups. Personality and Social Psychology Bulletin, 32, 230–243. Crisp, R. J., Walsh, J., & Hewstone, M. (2006c). Crossed categorization in common ingroup contexts. Personality and Social Psychology Bulletin, 32, 1204–1218. Daniel, G. R. (2002). More than Black? Multiracial identity and the new racial order. Philadelphia: Temple University Press. Das, J. P., & Nanda, P. C. (1963). Mediated transfer of attitudes. Journal of Abnormal and Social Psychology, 66, 12–16. Deaux, K. (1996). Social identification. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 777–798). New York: Guilford. Deschamps, J. C. (1977). EVects of crossing category membership on quantitative judgment. European Journal of Social Psychology, 7, 517–521. Deschamps, J. C., & Brown, R. J. (1983). Superordinate goals and intergroup conflict. British Journal of Social Psychology, 22, 189–195. Deschamps, J. C., & Doise, W. (1978). Crossed category memberships in intergroup relations. In H. Tajfel (Ed.), DiVerentiation between social groups (pp. 141–158). Cambridge: Cambridge University Press. Deutsch, M. (1973). The resolution of conflict. New Haven, CO: Yale University Press. Diehl, M. (1988). Social identity and minimal groups: The eVects of interpersonal and intergroup attitudinal similarity on intergroup discrimination. British Journal of Social Psychology, 27, 289–300. Diehl, M. (1990). The minimal group paradigm: Theoretical explanations and empirical findings. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 1, pp. 263–292). Chichester: Wiley. Dijksterhuis, A., & van Knippenberg, A. (1998). Inhibition, Aberdeen, and other cloudy subjects. In R. S. Wyer, Jr. (Ed.), Advances in social cognition (Vol. 11, pp. 83–96). Hillsdale, NJ: Lawrence Erlbaum Associates. Doise, W. (1978). Groups and individuals. Explanations in social psychology. Cambridge: Cambridge University Press. Doise, W., Deschamps, J. C., & Meyer, G. (1978). The accentuation of intra‐category similarities. In H. Tajfel (Ed.), DiVerentiation between social groups (pp. 159–168). London: Academic Press. Doosje, B., Spears, R., & Koomen, W. (1995). When bad isn’t all bad: The strategic use of sample information in generalization and stereotyping. Journal of Personality and Social Psychology, 69, 642–655. Dovidio, J. F., Gaertner, S.L, & Bachman, B. A. (2001). Racial bias in organizations: The role of group processes in its causes and cures. In M. E. Turner (Ed.), Groups at work: Theory and research (pp. 415–444). Mahwah, NJ: Lawrence Erlbaum Associates. Dovidio, J. F., Gaertner, S. L., Hodson, G., Riek, M. B., Johnson, K. M., & Houlette, M. (2006). Recategorization and crossed categorization: The implications of group salience and representations for reducing bias. In R. J. Crisp & M. Hewstone (Eds.), Multiple social
244
RICHARD J. CRISP AND MILES HEWSTONE
categorization: Processes, models and applications (pp. 65–89). Hove, E. Sussex: Psychology Press (Taylor & Francis). Dovidio, J. F., Gaertner, S. L., Isen, A. M., & Lowrance, R. (1995). Group representations and intergroup bias: Positive aVect, similarity, and group size. Personality and Social Psychology Bulletin, 21, 856–865. Dovidio, J. F., Gaertner, S. L., Isen, A. M., Rust, M., & Guerra, P. (1998a). Positive aVect, cognition, and the reduction of intergroup bias. In C. Sedikides, J. Schopler, & C. A. Insko (Eds.), Intergroup cognition and intergroup behavior (pp. 337–366). Mahwah, NJ: Erlbaum. Dovidio, J. F., Gaertner, S. L., & Validzic, A. (1998b). Intergroup bias: Status, diVerentiation and a common in‐group identity. Journal of Personality and Social Psychology, 75, 109–120. Eggins, R. A., Haslam, S. A., & Reynolds, K. J. (2002). Social identity and negotiation: Subgroup representation and superordinate consensus. Personality and Social Psychology Bulletin, 28, 887–899. Eiser, R. J., van der Pligt, J., & Gossop, M. R. (1979). Categorization, attitude and memory for the source of attitude statements. European Journal of Social Psychology, 9, 243–251. Ellemers, N., Spears, R., & Doosje, B. (2002). Self and social identity. Annual Review of Psychology, 53, 161–186. Ensari, N., & Miller, N. (1998). EVect of aVective reactions by the outgroup on preferences for crossed categorization discussion partners. Journal of Personality and Social Psychology, 75, 1503–1527. Ensari, N., & Miller, N. (2001). Decategorization and the reduction of bias in the crossed categorization paradigm. European Journal of Social Psychology, 31, 193–216. Esses, V. M., Jackson, L. M., & Armstrong, T. L. (1998). Intergroup competition and attitudes toward immigrants and immigration: An instrumental model of group conflict. Journal of Social Issues, 54, 699–724. Eurich‐Fulcher, R., & Schofield, J. W. (1995). Correlated versus uncorrelated social categorizations: The eVect on intergroup bias. Personality and Social Psychology Bulletin, 21, 149–159. Evans‐Pritchard, E. E. (1940). The Nuer. London: Oxford University Press. Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic activation of attitudes. Journal of Personality and Social Psychology, 50, 229–238. Fein, S., & Spencer, S. J. (1997). Prejudice as self‐image maintenance: AYrming the self through derogating others. Journal of Personality and Social Psychology, 73, 31–44. Fiske, S. T. (2000). Stereotyping, prejudice, and discrimination at the seam between the centuries: Evolution, culture, mind, and brain. European Journal of Social Psychology, 30, 299–322. Fiske, S. T., & Neuberg, S. L. (1989). Category‐based and individuating processes as a function of information and motivation: Evidence from our laboratory. In D. Bar‐Tal, C. F. Graumann, A. W. Kruglanski, & W. Stroebe (Eds.), Stereotyping and prejudice: Changing conceptions (pp. 83–104). New York: Springer‐Verlag. Fiske, S. T., & Neuberg, S. L. (1990). A continuum model of impression formation from category based to individuating process: Influences of information and motivation on attention an interpretation. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1–74). New York: Academic Press. Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York: McGraw‐Hill, Inc. Gaertner, S. L., & Dovidio, J. F. (1986). Prejudice, discrimination, and racism: Problems, progress and promise. In J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, discrimination, and racism (pp. 315–332). Orlando, FL: Academic Press. Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology Press/Taylor & Francis.
MULTIPLE SOCIAL CATEGORIZATION
245
Gaertner, S. L., Dovidio, J. F., Anastasio, P. A., Bachman, B. A., & Rust, M. C. (1993). The common ingroup identity model: Recategorization and the reduction of intergroup bias. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 2, pp. 247–278). Chichester: John Wiley. Gaertner, S. L., Dovidio, J. F., Banker, B. S., Houlette, M., Johnson, K. M., & McGlynn, E. A. (2000). Reducing intergroup conflict: From superordinate goals to decategorization, recategorization, and mutual diVerentiation. Group Dynamics: Theory, Research, and Practice, 4, 98–114. Gaertner, S. L., Dovidio, J. F., Guerra, R., Rebelo, M., Monteiro, M. B., Riek, M. B., et al. (in press). The common ingroup identity model: Applications to adults and children. In S. R. Levy, & M. Killen (Eds.), Intergroup relations: An integrative developmental and social psychological perspective. Oxford: Oxford University Press. Gaertner, S. L., Mann, J. A., Dovidio, J. F., Murrell, A. J., & Pomare, M. (1990). How does cooperation reduce intergroup bias? Journal of Personality and Social Psychology, 59, 692–704. Gaertner, S. L., Mann, J. A., Murrell, A. J., & Dovidio, J. F. (1989). Reducing intergroup bias: The benefits of recategorization. Journal of Personality and Social Psychology, 57, 239–249. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual diVerences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464–1480. Guetzkow, H. (1955). Multiple loyalties. Princeton, NJ: Princeton University Press. Halford, G. S., Baker, R., McCredden, J. E., & Bain, J. D. (2005). How many variables can humans process? Psychological Science, 16, 70–76. Hall, N. R., & Crisp, R. J. (2005). Considering multiple criteria for social categorization can reduce intergroup bias. Personality and Social Psychology Bulletin, 31, 1435–1444. Hamberger, J., & Hewstone, M. (1997). Inter‐ethnic contact as a predictor of blatant and subtle prejudice: Tests of a model in four West European nations. British Journal of Social Psychology, 36, 173–190. Hamilton, D. L. (1979). A cognitive attributional analysis of stereotyping. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12, pp. 53–84). New York: Academic Press. Hampton, J. A. (1997). Emergent attributes in combined concepts. In T. B. Ward, S. M. Smith, & J. Vaid (Eds.), Creative thought: An investigation of conceptual structures and processes (pp. 83–111). Washington, DC: APA books. Haslam, S. A., Oakes, P. J., Turner, J. C., & McGarty, C. (1995). Social categorization and group homogeneity: Changes in the perceived applicability of stereotype content as a function of comparative context and trait favorableness. British Journal of Social Psychology, 34, 139–160. Hastie, R. (1981). Schematic principles in human memory. In E. T. Higgins, C. P. Herman, & M. P. Zanna (Eds.), Social cognition: The Ontario symposium (Vol. 1, pp. 39–88). Hillsdale, NJ: Erlbaum. Hastie, R., Schroeder, C., & Weber, R. (1990). Creating complex social conjunction categories from simple categories. Bulletin of the Psychonomic Society, 28, 242–247. Henderson‐King, E., Henderson‐King, D., Zhermer, N., Posokhova, S., & Chiker, V. (1997). Ingroup favouritism and perceived similarity: A look at Russians’ perceptions in the post‐ Soviet era. Personality and Social Psychology Bulletin, 34, 1013–1021. Hewstone, M. (1994). Revision and change of stereotypic beliefs: In search of the elusive subtyping model. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 5, pp. 69–109). Chichester: J. Wiley.
246
RICHARD J. CRISP AND MILES HEWSTONE
Hewstone, M. (1996). Contact and categorization: Social psychological interventions to change intergroup relations. In N. Macrae, C. Stangor, & M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 323–368). New York: Guildford. Hewstone, M., & Brown, R. (1986). Contact is not enough: An intergroup perspective on the ‘‘Contact Hypothesis.’’ In M. Hewstone & R. Brown (Eds.), Contact and conflict in intergroup encounters (pp. 1–44). Oxford: Blackwell. Hewstone, M., & Hamberger, J. (2000). Perceived variability and stereotype change. Journal of Experimental Social Psychology, 36, 103–124. Hewstone, M., Islam, M. R., & Judd, C. M. (1993). Models of crossed categorization and intergroup relations. Journal of Personality and Social Psychology, 64, 779–793. Hewstone, M., Rubin, M., & Willis, H. (2002). Intergroup bias. Annual Review of Psychology, 53, 575–604. Hewstone, M., Turner, R. N., Kenworthy, J. B., & Crisp, R. J. (2006). Multiple social categorization: Integrative themes and future research priorities. In R. J. Crisp & M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications. Hove, E. Sussex: Psychology Press (Taylor & Francis). Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 1–46). New York: Academic Press. Higgins, E. T., & King, G. A. (1981). Accessibility of social constructs: Information processing consequences of individual and contextual variability. In N. Cantor & J. F. Kihlstrom (Eds.), Personality, cognition, and social interaction (pp. 69–122). Hillsdale, NJ: Erlbaum. Hogg, M. A. (2000). Subjective uncertainty reduction through self‐categorization: A motivational theory of social identity processes. European Review of Social Psychology, 11, 223–255. Hogg, M. A. (2001). Self‐categorization and subjective uncertainty resolution: Cognitive and motivational facets of social identity and group membership. In J. P. Forgas, K. D. Williams, & L. Wheeler (Eds.), The social mind: Cognitive and motivational aspects of interpersonal behaviour (pp. 323–349). New York: Cambridge University Press. Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of intergroup relations and group processes. London: Routledge. Hogg, M. A., & Abrams, D. (1990). Social motivation, self‐esteem and social identity. In D. Abrams & M. A. Hogg (Eds.), Social identity theory: Constructive and critical advances (pp. 28–47). New York: Harvester Wheatsheaf. Hogg, M. A., & Abrams, D. (1993). Towards a single‐process uncertainty‐reduction model of social motivation in groups. In M. A. Hogg & D. Abrams (Eds.), Group motivation: Social psychological perspectives (pp. 173–190). London: Harvester Wheatsheaf. Homan, A. C. (2006). Harvesting the value in diversity: Examining the eVects of diversity beliefs, cross‐categorization, and superordinate identities on the functioning of diverse work groups. Academic Dissertation, University of Amsterdam. Hornsey, M. J., & Hogg, M. A. (1999). Subgroup diVerentiation as a response to an overly‐ inclusive group: A test of optimal distinctiveness theory. European Journal of Social Psychology, 29, 543–550. Hornsey, M. J., & Hogg, M. A. (2000). Subgroup relations: A comparison of mutual intergroup diVerentiation and common ingroup identity models of prejudice reduction. Personality and Social Psychology Bulletin, 26, 242–256. Hornsey, M. J., & Hogg, M. A. (2002). The eVects of status on subgroup relations. British Journal of Social Psychology, 41, 203–218. Horowitz, D. L. (1985). Ethnic groups in conflict. Berkeley, CA: University of California Press.
MULTIPLE SOCIAL CATEGORIZATION
247
Houlette, M. A., Gaertner, S. L., Johnson, K. M., Banker, B. S., Riek, B. M., & Dovidio, J. F. (2004). Developing a more inclusive social identity: An elementary school intervention. Journal of Social Issues, 60, 35–55. Hugenberg, K., & Bodenhausen, G. V. (2002). Category membership moderates the inhibition of social identities. Journal of Experimental Social Psychology, 40, 233–238. Huntington, S. P. (1996). The clash of civilizations and the remaking of world order. New York, NY: Simon & Schuster. Hutter, R. R. H., & Crisp, R. J. (2005). The composition of category conjunctions. Personality and Social Psychology Bulletin, 31, 647–657. Hutter, R. R. H., & Crisp, R. J. (2006). Implications of cognitive busyness for the perception of category conjunctions. Journal of Social Psychology, 146, 253–256. Isen, A. M., & Daubman, K. A. (1984). The influence of aVect on categorization. Journal of Personality and Social Psychology, 47, 1206–1217. Isen, A. M., & Means, B. (1983). Positive aVect as a variable in decision making. Social Cognition, 2, 18–31. Isen, A. M., Niedenthal, P. M., & Cantor, N. (1992). An influence of positive aVect on social categorization. Motivation and Emotion, 16, 65–78. Isen, A. M., Shalker, T., Clark, M. S., & Karp, L. (1978). AVect, accessibility of material and behavior: A cognitive loop? Journal of Personality and Social Psychology, 36, 1–2. Islam, M. R., & Hewstone, M. (1993). Dimensions of contact as predictors of intergroup anxiety, perceived outgroup variability, and outgroup attitude: An integrative model. Personality and Social Psychology Bulletin, 19, 700–710. Jetten, J., Spears, R., & Manstead, A. S. R. (1997). Distinctiveness threat and prototypicality: Combined eVects on intergroup discrimination and collective self‐esteem. European Journal of Social Psychology, 27, 635–657. Jetten, J., Spears, R., & Manstead, A. S. R. (2001). Similarity as a source of discrimination: The role of group identification. European Journal of Social Psychology, 31, 621–640. Jetten, J., Spears, R., & Postmes, T. (2004). Intergroup distinctiveness and diVerentiation: A meta‐analytic integration. Journal of Personality and Social Psychology, 86, 862–879. Johnston, L. A., & Hewstone, M. (1992). Cognitive models of stereotype change: (3) Subtyping and the perceived typicality of disconfirming group members. Journal of Experimental Social Psychology, 28, 360–386. Jones, E. E., & Goethals, G. R. (1972). Order eVects in impression formation: Attribution context and the nature of the entity. In E. E. Jones, D. E. Kanouse, H. H. Kelly, R. E. Nisbett, S. Valins, & B. Weiner (Eds.), Attribution: Perceiving the causes of behavior (pp. 27–46). New Jersey: General Learning Press. Jones, E. E., Wood, G. C., & Quattrone, G. A. (1981). Perceived variability of personal characteristics in ingroups and outgroups: The role of knowledge and evaluation. Personality and Social Psychology Bulletin, 7, 523–528. Judd, C. M., Ryan, C. S., & Park, B. (1991). Accuracy in the judgment of ingroup and outgroup variability. Journal of Personality and Social Psychology, 61, 366–379. Kenworthy, J. B., Canales, C. J., Weaver, K. D., & Miller, N. (2003). Negative incidental aVect and mood congruency in crossed categorization. Journal of Experimental Social Psychology, 39, 195–219. Kessler, T., & Mummendey, A. (2002). Sequential or parallel processes? A longitudinal field study concerning determinants of identity management strategies. Journal of Personality and Social Psychology, 82, 75–88. Krueger, J., & Rothbart, M. (1988). Use of categorical and individuating information in making inferences about personality. Journal of Personality and Social Psychology, 55, 187–195.
248
RICHARD J. CRISP AND MILES HEWSTONE
Kunda, Z., Miller, D. T., & Claire, T. (1990). Combining social concepts: The role of causal reasoning. Cognitive Science, 14, 551–577. Levin, S., Sinclair, S., Veniegas, R., & Taylor, P. (2002). Perceived discrimination in the context of multiple social identities. Psychological Science, 13, 557–560. LeVine, R. A., & Campbell, D. T. (1972). Ethnocentrism: Theories of conflict, ethnic attitudes and group behaviour. New York: Wiley. Linville, P. W. (1982). The complexity‐extremity eVect and age‐based stereotyping. Journal of Personality and Social Psychology, 42, 193–211. Linville, P. W., Fischer, G. W., & Yoon, C. (1996). Perceived covariation among the features of ingroup and outgroup members: The outgroup covariation eVect. Journal of Personality and Social Psychology, 70, 421–436. Lipset, S. M. (1960). Political man. Garden City, NY: Doubleday. Mackie, D. M., & Worth, L. T. (1989). Processing deficits and the mediation of positive aVect in persuasion. Journal of Personality and Social Psychology, 57, 27–40. Macrae, C. N., & Bodenhausen, G. V. (2001). Social cognition: Categorical person perception. British Journal of Psychology, 92, 239–255. Macrae, C. N., Bodenhausen, G. V., & Milne, A. B. (1995). The dissection of selection in person perception: Inhibitory processes in social stereotyping. Journal of Personality and Social Psychology, 69, 397–407. Macrae, C. N., Bodenhausen, G. V., Schloerscheidt, A. M., & Milne, A. B. (1999). Tales of the unexpected: Executive function and person perception. Journal of Personality and Social Psychology, 76, 200–213. Macrae, C. N., Stangor, C., & Milne, A. B. (1994). Activating social stereotypes: A functional analysis. Journal of Experimental Social Psychology, 30, 370–389. Marcus‐Newhall, A., Miller, N., Holtz, R., & Brewer, M. B. (1993). Cross‐cutting category membership with role assignment: A means of reducing intergroup bias. British Journal of Social Psychology, 32, 125–146. Markman, E. M., & Callanan, M. A. (1984). An analysis of hierarchical classification. In R. J. Sternberg (Ed.), Advances in the psychology of human intelligence (pp. 325–365). Hillsdale, NJ: Erlbaum. Maurer, K. L., Park, B., & Rothbart, M. (1995). Subtyping versus subgrouping processes in stereotype representation. Journal of Personality and Social Psychology, 69, 812–824. McGarty, C. (1999). Categorization in social psychology. London: Sage. McGarty, C. (2006). Hierarchies and groups: The roles of salience, overlap, and background knowledge in selecting meaningful social categorizations from multiple alternatives. In R. J. Crisp & M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications (pp. 25–49). Hove, E. Sussex: Psychology Press (Taylor & Francis). McGarty, C., & Penny, R. E. C. (1988). Categorization, accentuation, and social judgement. British Journal of Social Psychology, 27, 147–157. Messick, D. M., & Mackie, D. M. (1989). Intergroup relations. Annual Review of Psychology, 33, 45–81. Migdal, M., Hewstone, M., & Mullen, B. (1998). The eVects of crossed categorization in intergroup evaluations: A meta‐analysis. British Journal of Social Psychology, 37, 303–324. Miller, N., & Harrington, H. J. (1992). Social categorization and intergroup acceptance: Principles for the design and development of cooperative learning teams. In R. Hertz‐ Lazarowitz & N. Miller (Eds.), Interaction in cooperative groups: the theoretical anatomy of group learning (pp. 203–227). New York: Cambridge University Press. Miller, N., Kenworthy, J., Carnales, C. J., & Stenstrom, D. M. (2006). Explaining the eVects of crossed categorization on ethnocentric bias. In R. J. Crisp & M. Hewstone (Eds.), Multiple
MULTIPLE SOCIAL CATEGORIZATION
249
social categorization: Processes, models and applications (pp. 160–188). Hove, E. Sussex: Psychology Press (Taylor & Francis). Miller, N., Urban, L. M., & Vanman, E. J. (1998). A theoretical analysis of crossed social categorization eVects. In C. Sedikides, J. Schopler, & C. Insko (Eds.), Intergroup cognition and intergroup behavior (pp. 313–419). Hillsdale, NJ: Erlbaum. Moskowitz, G. B., Gollwitzer, P. M., Wasel, W., & Schaal, B. (1999). Preconscious control of stereotype activation through chronic egalitarian goals. Journal of Personality and Social Psychology, 77, 167–184. Mullen, B., Brown, R. J., & Smith, C. (1992). Ingroup bias as a function of salience, relevance and status: An integration. European Journal of Social Psychology, 22, 103–122. Mullen, B., & Hu, L. (1989). Perceptions of ingroup and outgroup variability: A meta‐analytic integration. Basic and Applied Social Psychology, 10, 233–252. Mullen, B., Migdal, M. J., & Hewstone, M. (2001). Crossed categorization versus simple categorization and intergroup evaluations: A meta‐analysis. European Journal of Social Psychology, 31, 721–736. Mummendey, A., Simon, B., Dietze, C., Grunert, M., Haeger, G., Kessler, S., et al. (1992). Categorization is no enough: Intergroup discrimination in negative outcome allocations. Journal of Experimental Social Psychology, 28, 125–144. Mummendey, A., & Wenzel, M. (1999). Social discrimination and tolerance in intergroup relations: Reactions to intergroup diVerence. Personality and Social Psychology Review, 3, 158–174. Murphy, G. L. (1988). Comprehending complex concepts. Cognitive Science, 12, 529–562. Murphy, R. F. (1957). Intergroup hostility and social cohesion. American Anthropologist, 59, 1018–1035. Mussweiler, T., Gabriel, S., & Bodenhausen, G. (2000). Shifting social identities as a strategy for deflecting threatening social comparisons. Journal of Personality and Social Psychology, 79, 398–409. Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality. Oxford: Blackwell. Oakes, P. J., & Turner, J. C. (1990). Is limited information processing capacity the cause of social stereotyping? In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 1, pp. 111–135). Chichester: John Wiley. Oakes, P. J., Turner, J. C., & Haslam, S. A. (1991). Perceiving people as group members: The role of fit in the salience of social categorizations. British Journal of Social Psychology, 30, 125–144. Park, B., & Judd, C. M. (1990). Measures and models of perceived group variability. Journal of Personality and Social Psychology, 59, 173–191. Park, B., Judd, C. M., & Ryan, C. S. (1991). Social categorization and the representation of variability information. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 2, pp. 211–245). Chichester: John Wiley. Park, B., & Rothbart, M. (1982). Perceptions of outgroup homogeneity and levels of social categorization: Memory for the subordinate attributes of ingroup and outgroup members. Journal of Personality and Social Psychology, 42, 1051–1068. Park, B., Ryan, C. S., & Judd, C. M. (1992). The role of meaningful subgroups in explaining diVerences in perceived variability of ingroups and outgroup. Journal of Personality and Social Psychology, 63, 553–567. Perdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990). ‘‘Us’’ and ‘‘Them’’: Social categorization and the process of intergroup bias. Journal of Personality and Social Psychology, 59, 475–486.
250
RICHARD J. CRISP AND MILES HEWSTONE
Pettigrew, T. F. (1981). Extending the stereotype concept. In D. L. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behaviour (pp. 303–333). Hillsdale, NJ: Lawrence Erlbaum Associates. Pettigrew, T. F. (1997). Generalized intergroup contact eVects on prejudice. Personality and Social Psychology Bulletin, 23, 173–185. Pettigrew, T. F. (1998). Intergroup contact theory. Annual Review of Psychology, 49, 65–85. Phinney, J. S., & Alipuria, L. L. (2006). Multiple social categorization and identity among multiracial, multiethnic, and multicultural individuals: Processes and implications. In R. J. Crisp and M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications (pp. 211–238). Hove, E. Sussex: Psychology Press (Taylor & Francis). Phinney, J. S., Ferguson, D. L., & Tate, J. D. (1997). Intergroup attitudes among ethnic minority adolescents: A causal model. Child Development, 68, 955–969. Piaget, J. (1965). The moral judgment of the child. New York: The Free Press. Piaget, J., & Weil, A.‐M. (1951). The development in children of the idea of the homeland and of relations with other countries. International Social Science Bulletin, 3, 561–578. Pineda, D. A. (2000). Executive function and its disorders. Revista de Neurologia, 30, 764–768. Rabbie, J. M., & de Brey, J. H. C. (1971). The anticipation of intergroup cooperation and competition under private and public conditions. International Journal of Group Tensions, 1, 230–251. Rabbie, J. M., & Horowitz, M. (1969). Arousal of ingroup‐outgroup bias by chance win or loss. Journal of Personality and Social Psychology, 13, 269–277. Rae, D. W., & Taylor, M. (1970). The analysis of political cleavages. New Haven, CN: Yale University Press. Reynolds, K. J., Turner, J. C., & Haslam, S. A. (2000). When are we better than them and they worse than us? A closer look at social discrimination in positive and negative domains. Journal of Personality and Social Psychology, 78, 64–80. Richards, Z., & Hewstone, M. (2001). Subtyping and subgrouping: Processes for the prevention and promotion of stereotype change. Personality and Social Psychology Review, 5, 52–73. Roccas, S., & Brewer, M. B. (2002). Social identity complexity. Personality and Social Psychology Review, 6, 88–106. Rosch, E., Mervis, C. B., Gray, W., Johnson, D., & Boyes‐Braem, P. (1976). Basic objects in natural categories. Cognitive Psychology, 8, 382–439. Rosch, E. H. (1978). Principles of categorization. In E. Rosch & B. Lloyd (Eds.), Cognition and categorization (pp. 27–48). Hillsdale, NJ: Erlbaum Associates. Rosenthal, H. E. S., & Crisp, R. J. (2006). Reducing stereotype threat by blurring intergroup boundaries. Personality and Social Psychology Bulletin, 32, 501–511. Rosenthal, H. E. S., Crisp, R. J., & Suen, M.‐W. (in press). Improving performance expectancies in stereotypic domains: Task relevance and the reduction of stereotype threat. European Journal of Social Psychology. Ross, M. H. (1985). Internal and external conflict and violence: Cross‐cultural evidence and a new analysis. Journal of Conflict Resolution, 29, 547–579. Rothbart, M., & John, O. P. (1985). Social categorization and behavioral episodes: A cognitive analysis of the eVects of intergroup contact. Journal of Social Issues, 41, 81–104. Rubin, J. M., & Hewstone, M. (1998). Social identity theory’s self‐esteem hypothesis: A review and some suggestions for clarification. Personality and Social Psychology Review, 2, 40–62. Rust, M. (1997). Social identity and social categorization. Unpublished Ph.D. dissertation, Department of Psychology, University of Delaware, Newark, DE. Cited in Dovidio et al. (2006).
MULTIPLE SOCIAL CATEGORIZATION
251
Ryan, C. S., Park, B., & Judd, C. M. (1996). Assessing stereotype accuracy: Implications for understanding the stereotype process. In C. N. Macrae, C. Stangor, & M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 121–157). New York: Guildford. Said, E. (1999). Out of place: A memoir. New York, NY: Vintage Books/Random House. Sassenberg, K., Kessler, T., & Mummendey, A. (2003). Less negative ¼ more positive? Social discrimination as avoidance or approach. Journal of Experimental Social Psychology, 39, 48–58. Schaller, M., & Maass, A. (1989). Illusory correlation and social categorization: Toward an integration of motivational and cognitive factors in stereotype formation. Journal of Personality and Social Psychology, 56, 709–721. Schofield, J. W., & McGivern, E. (1979). Changing interracial bonds in a desegrated school. In R. G. Blumberg & W. J. Roye (Eds.), Interracial bonds (pp. 106–119). New York: General Hall. Sen, A. (2006). Identity and violence: The illusion of destiny. New York, NY: W. W. Norton & Co. Sherif, M. (1966). Group conflict and cooperation. London: Routledge Kegan Paul. Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif, C. W. (1961). Intergroup conflict and cooperation. The Robbers’ Cave experiment. Norman, OK: University of Oklahoma Book Exchange. Sherif, M., & Sherif, C. W. (1953). Groups in harmony and tension. New York: Harper. Sherif, M., White, B. J., & Harvey, O. J. (1955). Status in experimentally produced groups. American Journal of Sociology, 60, 370–379. Shih, M., Pittinsky, T. L., & Ambady, N. (1999). Stereotype susceptibility: Identity salience and shifts in quantitative performance. Psychological Science, 10, 80–83. Simmel, G. (1950). The sociology of Georg Simmel. (Kurt WolV, Trans.). New York, NY: Free Press. (Original work published 1908). Sinclair, L., & Kunda, Z. (1999). Reactions to a black professional: Motivated inhibition and activation of conflicting stereotypes. Journal of Personality and Social Psychology, 77, 885–904. Sinclair, S., Hardin, C. D., & Lowery, B. S. (2006). Self‐stereotyping in the context of multiple social identities. Journal of Personality and Social Psychology, 90, 529–542. Slavin, R. E. (1985). Cooperative learning: Applying contact theory in desegregated schools. Journal of Social Issues, 41, 45–62. Smith, E. R. (2006). Multiply categorizable social objects: Representational models and some potential determinants of category use. In R. J. Crisp & M. Hewstone (Eds.), Multiple social categorization: Processes, models and applications. Hove, E. Sussex: Psychology Press (Taylor & Francis). Smith, E. R., & DeCoster, J. (1998). Knowledge acquisition, accessibility, and use in person perception and stereotyping: Simulation with a recurrent connectionist network. Journal of Personality and Social Psychology, 74, 21–35. Smurda, J. D., Wittig, M. A., & Gokalp, G. (2006). EVects of threat to a valued social identity on implicit self‐esteem and discrimination. Group Processes and Intergroup Relations, 9, 181–197. Srull, T. K., & Wyer, R. S. (1989). Person memory and judgement. Psychological Review, 96, 58–83. Staats, A. W., & Staats, C. K. (1958). Attitudes established by classical conditioning. Journal of Abnormal and Social Psychology, 57, 37–40. Stangor, C., Lynch, L., Duan, C., & Glass, B. (1992). Categorization of individuals on the basis of multiple social features. Journal of Personality and Social Psychology, 62, 207–218.
252
RICHARD J. CRISP AND MILES HEWSTONE
Stangor, C., & Macmillan, D. (1992). Memory for expectancy‐congruent and expectancy‐ incongruent information: A review of the social and social developmental literatures. Psychological Bulletin, 111, 42–61. Steele, C. M. (1997). A threat in the air: How stereotypes shape the intellectual identities and performance of women and African‐Americans. American Psychologist, 52, 613–629. Steele, C. M., Spencer, S., & Aronson, J. (2002). Contending with group image: The psychology of stereotype and social identity threat. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 34, pp. 379–440). San Diego: Academic Press. Stryker, S., & Statham, A. (1985). Symbolic interaction and role theory. In C. Lindzey & E. Aronson (Eds.), Handbook of social psychology (3rd ed., pp. 311–378). New York, NY: Random House. Tajfel, H. (1959). Quantitative judgment in social perception. British Journal of Psychology, 50, 16–29. Tajfel, H. (1978). Social categorization, social identity and social comparison. In H. Tajfel (Ed.), DiVerentiation between social groups: Studies in the social psychology of intergroup relations (pp. 61–76). London: Academic Press. Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33, 1–39. Tajfel, H., Flament, C., Billig, M., & Bundy, R. F. (1971). Social categorization and intergroup behaviour. European Journal of Social Psychology, 1, 149–178. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). California: Brooks/Cole. Tajfel, H., & Wilkes, A. L. (1963). Classification and quantitative judgment. British Journal of Psychology, 54, 101–113. Taylor, S. E., Fiske, S. T., EtkoV, N. L., & Ruderman, A. J. (1978). Categorical and contextual bases of person memory and stereotyping. Journal of Personality and Social Psychology, 36, 778–793. Teasdale, J. D., & Russell, M. L. (1983). DiVerential aspects of induced mood on the recall of positive, negative, and neutral words. British Journal of Clinical Psychology, 22, 163–171. Tesser, A. (1988). Toward a self‐evaluation maintenance model of social behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 181–227). New York: Academic Press. Turner, J. C. (1981). The experimental social psychology of intergroup behaviour. In J. C. Turner & H. Giles (Eds.), Intergroup behaviour (pp. 66–101). Oxford: Basil Blackwell. Turner, J. C. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (Ed.), Social identity and intergroup relations (pp. 15–40). Cambridge: Cambridge University Press. Turner, J. C. (1999). Some current issues in research on social identity and self‐categorization theories. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity (pp. 6–34). Malden, MA: Blackwell. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self‐categorization theory. Oxford: Basil Blackwell. Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. (1994). Self and collective: Cognition and social context. Personality and Social Psychology Bulletin, 202, 454–463. Urada, D. I., & Miller, N. (2000). The impact of positive mood and category importance on crossed categorization eVects. Journal of Personality and Social Psychology, 78, 417–433. Urada, D. I., Stenstrom, D., & Miller, N. (in press). Crossed categorization beyond the two‐ group model. Journal of Personality and Social Psychology. Urban, L. M., & Miller, N. M. (1998). A theoretical analysis of crossed categorization eVects: A meta‐analysis. Journal of Personality and Social Psychology, 74, 894–908.
MULTIPLE SOCIAL CATEGORIZATION
253
van Knippenberg, A. V., & Dijksterhuis, A. (2000). Social categorization and stereotyping: A functional perspective. In W. Stroebe & M. Hewstone (Eds.), European Review of Social Psychology (Vol. 11, pp. 105–144). Chichester: John Wiley & Sons Ltd. van Knippenberg, A., van Twuyer, M., & Pepels, J. (1994). Factors aVecting social categorization processes in memory. British Journal of Social Psychology, 33, 419–431. van Leeuwen, E., van Knippenberg, D., & Ellemers, N. (2003). Continuing and changing group identities: The eVects of merging on social identification and ingroup bias. Personality and Social Psychology Bulletin, 29, 679–690. van Oudenhoven, J. P., Groenewoud, J. T., & Hewstone, M. (1996). Cooperation, ethnic salience and generalization of interethnic attitudes. European Journal of Social Psychology, 26, 649–661. van Oudenhoven, J. P., Prins, K. S., & Buunk, B. P. (1998). Attitudes of minority and majority members towards adaptation of immigrants. European Journal of Social Psychology, 28, 995–1013. van Rijswijk, W., & Ellemers, N. (2002). Context eVects on the application of stereotype content to multiple categorizable targets. Personality and Social Psychology Bulletin, 28, 90–101. van Twuyver, M., & van Knippenberg, A. (1995). Social categorization as a function of priming. European Journal of Social Psychology, 25, 695–702. van Twuyver, M., & van Knippenberg, A. (1999). Social categorization as a function of relative group size. British Journal of Social Psychology, 38, 135–156. Vanbeselaere, N. (1987). The eVect of dichotomous and crossed social categorizations upon intergroup discrimination. European Journal of Social Psychology, 17, 143–156. Vanbeselaere, N. (1991). The diVerent eVects of simple and crossed categorizations: A result of the category diVerentiation process or of diVerential category salience? In W. Stroebe & M. Hewstone (Eds.), European Review of Social Psychology (Vol. 2, pp. 247–278). Chichester: John Wiley. Vanbeselaere, N. (1996). The impact of diVerentially valued overlapping categorizations upon the diVerentiation between positively, negatively and neutrally evaluated social groups. European Journal of Social Psychology, 26, 75–96. Verkuyten, M. (2005). The social psychology of ethnic identity. Hove: Psychology Press. Vescio, T. K., Hewstone, M., Crisp, R. J., & Rubin, M. (1999). Perceiving and responding to multiply categorizable individuals: Cognitive processes and aVective intergroup bias. In D. Abrams & M. A. Hogg (Eds.), Social identity and social cognition (pp. 111–140). Oxford: Blackwell. Vescio, T. K., Judd, C. M., & Kwan, V. S. Y. (2004). Categorization and intergroup bias in crossed contexts: Evidence of reductions in the strength of categorization, but not intergroup bias. Journal of Experimental Social Psychology, 40, 478–496. Waldzus, S., Mummendey, A., Wenzel, M., & Weber, U. (2003). Towards tolerance: Representations of superordinate categories and perceived ingroup prototypicality. Journal of Experimental Social Psychology, 39, 31–47. Wallace, M. D. (1973). Alliance polarization, cross‐cutting, and international war. 1815–1964. Journal of Conflict Resolution, 17, 575–604. Wattenmaker, W. D. (1995). Knowledge structures and linear separability: Integrating information in object and social categorization. Cognitive Psychology, 28, 274–328. Weber, R., & Crocker, J. (1983). Cognitive processes in the revision of stereotypic beliefs. Journal of Personality and Social Psychology, 45, 961–977. Wenzel, M. (2001). A social categorization approach to distributive justice: Social identity as the link between relevance of inputs and need for justice. British Journal of Social Psychology, 40, 315–335.
254
RICHARD J. CRISP AND MILES HEWSTONE
Wenzel, M., Mummendey, A., Weber, U., & Waldzus, S. (2003). The ingroup as pars pro toto: Projection from the ingroup onto the inclusive category as a precursor to social discrimination. Personality and Social Psychology Bulletin, 29, 461–473. Wicker, T. (1975). A time to die. Harmondsworth: Penguine Books. Wilder, D. A. (1986). Social categorization: Implications for creation and reduction of intergroup bias. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19, pp. 293–355). New York: Academic Press. Wilkenfield, M. J., & Ward, T. B. (2001). Similarity and emergence in conceptual combination. Journal of Memory and Language, 45, 21–38. Worchel, P. (1979). Trust and distrust. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 174–187). Belmont, CA: Wadsworth. Worchel, S., Andreoli, V. A., & Folger, R. (1977). Intergroup cooperation and intergroup attraction: The eVect of previous interaction and outcome of combined eVort. Journal of Experimental Social Psychology, 13, 131–140.
ON THE PARAMETERS OF HUMAN JUDGMENT
Arie W. Kruglanski Antonio Pierro Lucia Mannetti Hans‐Peter Erb Woo Young Chun
This chapter reviews research concerning several continuous parameters whose combinations determine the judgmental impact of the information given. The present framework oVers an integration of prior judgmental models in various domains by conceptually distinguishing between qualitatively distinct contents of information and the quantitative dimensions on which judgmental contexts may vary. Evidence for the present formulation includes a broad array of findings across diVerent areas of psychology, and the discussion considers the comprehensive perspective it oVers on human judgment.
I. Introduction The judgment of persons, objects, and events is a ubiquitous human activity inexorably involved in people’s everyday aVairs. From the moment we open our eyes, if not before, we continually pass judgments on a variety of issues. Is it time to get up, brush our teeth, get dressed? What should one wear? What is the weather like? How should one’s day unfold? Is it safe to cross the street? Should one buy or sell? Should one ‘‘order in’’ tonight? And so continually, until the moment we shut our eyes and lull ourselves to sleep. Small wonder then that the 255 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39005-3
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
256
ARIE W. KRUGLANSKI et al.
mysteries of human judgment have fascinated theorists and researchers across the diVerent areas of psychology including its social, cognitive, perceptual, clinical, organizational, and personality subfields. The domain of human judgments is exceedingly diverse and multifaceted. First, judgments vary a great deal in the domain of content to which they belong. ‘‘Causal attributions’’ constitute judgments, but so do ‘‘category assignments,’’ ‘‘stereotypic characterizations,’’ ‘‘behavioral identifications,’’ ‘‘likelihood estimates,’’ ‘‘personality impressions,’’ ‘‘attitudes,’’ and ‘‘beliefs.’’ Secondly, judgments vary in their speed and immediacy. Some are based on painstaking deliberations and a laborious processing of available information. Other ‘‘pop out’’ quite eVortlessly from the stimulus field confronting the perceiver. They come to mind instantly and intuitively. They feel objective and inevitable, as if simply impelled by the external reality. Judgments vary also on the process‐awareness dimension. Some (e.g., of attorneys in a court of law, or of scholars in an academic treatise) are highly conscious and transparent; they often come in a written form wherein the chain of reasoning is explicated in an elaborate detail. Others arise mysteriously, as if out of nowhere, and as an end result of an opaque process refractory to conscious discovery (Nisbett & Wilson, 1977; Wilson, Dunn, Kraft, & Lisle, 1989). Possibly driven by this multifarious variability, a plethora of models and theories has been advanced to explain diVerent judgmental phenomena, some highlighting judgmental content domains (e.g., models of stereotyping, of attribution, or of attitudes), others drawing distinctions along the amount of processing continuum (Fiske & Neuberg, 1990; Petty & Cacioppo, 1986), yet other stressing the diVerences between ‘‘automatic’’ and ‘‘deliberative’’ ways of judging, including aspects of eYciency as well as awareness (Bargh, 1996; Devine, 1989). As a consequence of these varied distinctions, the field of human judgment appears highly fragmented these days, and the diVerent bodies of judgmental research hardly interact with each other. The question, therefore, is whether an integrative assembly of these various ‘‘puzzle pieces’’ is at all possible and ‘‘how is it all put back together?’’ (Anderson et al., 2004, p. 1036). Clearly, a comprehensive synthesis of the general judgmental domain would be desirable in light of the growing heterogeneity of concepts and models that lend this field of inquiry a complex and unwieldy character. It is along these lines that Newell (1990, p. 18) argued for ‘‘the necessity of a theory that provides the total picture and explains the role of the parts and why they exist.’’ In the present chapter, we oVer an outline of such theory. Our key notion is that of judgmental parameters. By these we mean several dimensional variables intersecting at some of their values in each judgmental instance. These variables have been yielded by decades of research on human
ON THE PARAMETERS OF HUMAN JUDGMENT
257
judgment and are well known and noncontroversial. Our proposal highlights the role of their combinations in determining the degree to which a specific, situationally given, information aVects judgments (i.e., the degree to which it is persuasive, convincing, and impactful). Moreover, whereas each judgmental situation involves some informational contents, it is presently assumed that apart from the parametric values with which they happen to be associated, these contents as such are immaterial to the information’s impact. As will be shown, the conceptual framework resulting from these assumptions aVords a novel way of viewing prior findings, and it enables the generation of novel predictions in a variety of domains. This chapter is divided into four sequential parts. In Part 1, we describe the judgmental function of rule following and contrast it with its various proposed alternatives such as associative learning, pattern recognition, or attribute matching. As a preview of what is to come, we conclude that, in general, judgments are mediated by rules, broadly conceived, and that the diVerences between rule following and other putative processes are semantic rather than substantive. In Part 2, we identify several continuous parameters whose combinations at various values determine whether and to what degree the information given would aVect judgments. In Part 3, we present our theory of the judgmental process and the supporting empirical evidence from the realm of complex social judgments. Part 4 recapitulates our model’s contributions and considers their implications.
II. The Role of Rules in Judgment Formation We are assuming that judgments are based on information the knower treats as ‘‘evidence.’’ For instance, in forming an attitudinal judgment one may consider information about the positive and negative consequences mediated by an attitude object (Albarracin, Johnson, & Zanna, 2005); such presumed consequences serve as evidence for the attitude object’s goodness (or badness). In forming a causal attribution, one may consider information about the covariation of an eVect with an entity as evidence that the entity is cause of that eVect (Kelley, 1967, 1971). In forming an impression of a person, prejudiced individuals may treat that individual’s category membership (e.g., her or his gender, profession, race, age, or religion) as evidence for various characteristics stereotypically attached to that category. In forecasting one’s future aVective state, given a present traumatic event (e.g., denial of tenure, bankruptcy), one’s evidence might be one’s presumed immediate feeling, plus a subjective estimate of its duration (Wilson, Centerbar, Kermer, & Gilbert, 2005), and so on.
258
ARIE W. KRUGLANSKI et al.
A. THE IMPLICATIONAL STRUCTURE OF REASONING From the present perspective, to fulfill the evidential function information has to fit an inference rule of an ‘‘IF THEN’’ type (Erb et al., 2003; Kruglanski & Thompson, 1999a,b; Pierro, Mannetti, Kruglanski, & Sleeth‐ Keppler, 2004). For example, information that an actor succeeded on a diYcult cognitive task could serve as evidence that he/she is intelligent, given the prior assumption that ‘‘success at diYcult cognitive tasks betokens intelligence.’’ More formally speaking, the reasoning from evidence to conclusions is syllogistic.1 It includes a major premise, the ‘‘IF X THEN Y’’ conditional rule mentioned above, and a minor premise that instantiates the antecedent term of the major premise (X) for a given entity (event, and so on) P, asserting that P is X, hence that Y is to be expected. For instance, a person may subscribe to the stereotypic belief ‘‘if rocket scientist then intelligent.’’ On an instance of encountering P, a rocket scientist (minor premise instantiating the antecedent term of the major premise), the knower would be subjectively justified in inferring that he or she is, therefore, intelligent. B. CONDITIONING PHENOMENA The foregoing, schematic, depiction of syllogistic reasoning may sound highly ‘‘rational,’’ ‘‘conscious,’’ and ‘‘explicit’’ and in this sense distinct from associatively mediated judgments, assumed to be ‘‘intuitive,’’ ‘‘implicit,’’ or ‘‘automatic.’’ Yet, from the present perspective, the distinction between the two merits a closer look. An insight into the processes involved comes from the many decades of research on conditioning phenomena. These have been generally viewed as the paradigmatic example of associative learning; nonetheless, there has been increasing tendency to accept the conclusion (Holyoak, Koh, & Nisbett, 1989; Rescorla, 1985; Rescorla & Holland, 1982; Rescorla & Wagner, 1972) that they are fundamentally rule based, in fact. Originally, associative learning was assumed to constitute learning by contiguity, and repeated pairing of a conditioned stimulus (CS) and an unconditioned stimulus (US). Yet, evidence from the animal‐learning literature suggests that neither is necessary nor suYcient for conditioning to occur. This is attested by the fact that conditioning can occur on a single co‐occurrence of stimuli, and when an interval of minutes or even hours elapses between 1 It is not meant here that people’s reasoning generally follows the rules of formal logic, or is capable of drawing the various logical implications of conditional statements, a state of aVairs bellied by decades of research on the Wason’s four card task (Wason & Johnson‐Laird, 1972) among others.
ON THE PARAMETERS OF HUMAN JUDGMENT
259
the stimuli. Thus, if a rat ingested a novel substance and was made ill minutes or hours later, it will form a strong aversion to that substance (Garcia, McGowan, Ervin, & Koelling, 1968). As Holyoak et al. (1989, p. 316) noted ‘‘taste aversion . . . is only the extreme of a continuum of successful conditioning that occurs despite lags between CS and US presentation. There are many demonstrations of classical and instrumental conditioning in which the delay between events is on the order of many seconds or minutes.’’ Blocking experiments too establish that temporal contiguity between a CS and a US is not suYcient for conditioning. In this research, a normally eVective CS, placed in close temporal relation to a reinforcer, seems largely unable to produce conditioning if paired with another CS that had been previously established as a signal for that reinforcer. That temporal contiguity is not suYcient for conditioning has been further demonstrated by (1) Rescorla’s (1968) results that if the probability of the reinforcer is the same in the absence of the CS as in its presence no appreciable conditioning will take place, (2) by conditioned inhibition eVects whereby a higher order stimulus (CS1) paired on some trials with another CS stimulus that on other trials was paired with some US (say, a shock) comes actually to inhibit the reaction associated with the US (e.g., crouching) even though a strict associative interpretation (of second order conditioning) would suggest that this stimulus (i.e., the CS1) should evoke that very reaction. It thus appears that an animal, rather than responding mechanistically to contiguous pairings of stimuli over repeated occasions (representing the associative account), is attempting to learn environmental contingencies in which the occurrence of one event (e.g., shock) is conditional on the occurrence of another (e.g., noise). This formulation is reminiscent of Tolman’s (1932) classic sign‐learning theory whereby what is learned is an expectancy, or a conditional probability, that a given environmental ‘‘sign’’ (e.g., the appearance of a light) presages a given ‘‘significate’’ (e.g., food). According to Holyoak et al. (1989, p. 320) specifically, Rules drive the system’s behavior by means of a recognize‐act cycle. On each cycle, the conditions of rules are matched against representations of active declarative information, which we . . . term messages; rules with conditions that are satisfied by current messages become candidates for execution. For example, if a message representing the recent occurrence of a tone is active, the conditions of the above rule will be matched and the actions it specifies may be taken.
It is noteworthy that the ‘‘rule’’ assumed by Holyoak et al. (1989) is analogous to the major premise, and the ‘‘message’’ is analogous to the minor premise, that is, an instantiation of the antecedent term in the major premise, warranting the inference of the consequent term. Thus, ‘‘the rat’s knowledge about the relation between tones and shocks might be informally represented
260
ARIE W. KRUGLANSKI et al.
by a rule such as ‘if a tone sounds in the chamber then a shock will occur, so stop other activities and crouch’’’ (Holyoak et al., 1989). The foregoing represents the major premise, and the sounding of a tone in a specific instance represents the minor premise, jointly aVording the inference that ‘‘crouching’’ was indicated. The rules underlying conditioning phenomena may be applied speedily and eYciently. It has been long known that ‘‘automatic’’ phenomena involve a routinization of IF THEN sequences. Research in this domain has demonstrated that social judgments (Smith & Branscombe, 1988; Smith, Branscombe, & Bormann, 1988) represent a special case of procedural learning (Anderson, 1983), based on processes (such as practice) that strengthen the IF THEN relation resulting in an increased ‘‘automaticity’’ of rule‐based responses (cf. Bargh, 1996; Neal, Wood, & Quinn, 2006). 1. Pavlovian Versus Evaluative Conditioning Gawronski and Bodenhausen (2006) suggested that whereas classical or Pavlovian conditioning is, in fact, propositional, or rule‐based, truly associative processes are based on evaluative learning which is not. A paradigmatic case of evaluative conditioning is the pairing of a valenced US (say a positively valenced smiling face) with a neutral CS (say an expressionless face) and finding that the CS then acquired the valence of the US (i.e., became positively valenced). Such paradigm is admittedly diVerent from the Pavlovian conditioning paradigm in which the CS constitutes a signal for the occurrence of the US. The processes mediating evaluative conditioning are not well understood at this time (Walther, Nagengast, & Trasselli, 2005). Thus, it seems premature to conclude that evaluative conditioning is not propositional in nature. For example, Fazio (personal communication, 2005) has proposed that evaluative conditioning may represent the misattribution to the CS of aVect evoked by the US. Such a process is inferential (as are attributional processes generally) and hence propositional, in the IF THEN sense. Specifically, one might infer that if positive aVect was experienced in presence of the CS then it may have been caused by the CS, warranting a reexperience of the aVect on subsequent CS presentations. Alternatively, the CS and the US may form a Gestalt or a group of likable or unlikable individuals (Walther et al., 2005). Once a CS was categorized as member of that group, its subsequent appearances may evoke the aVect/ evaluation accorded to that group. That would also explain why in evaluative conditioning situations, presentations of the CS alone do not lead to extinction unlike such presentations in a Pavlovian conditioning situations (Baeyens, Crombez, Van den Bergh, & Eelen, 1988). In the latter situation, an appearance of a CS without a subsequent US is inconsistent with the IF
ON THE PARAMETERS OF HUMAN JUDGMENT
261
THEN rule whereby if the CS appears, the US will follow. By contrast, in an evaluative conditioning situation, presentation of the CS alone does not invalidate the rule: Once the CS was categorized as member of a group, its appearance without alternative members of the group does not undermine its group membership. Imagine that X was classified as a ‘‘terrorist’’ because of her or his associative connection with a known Al Qaeda member (say Osama Bin Laden). Encountering X subsequently without the presence of Bin Laden should not undermine her or his membership in the terrorist category, because the relevant proposition ‘‘X is a terrorist’’ does not imply that X would always, or even often, be in Bin Laden’s company.
C. PERCEPTION AND PSYCHOPHYSICS The idea that even the most basic perceptual judgments are rule‐based receives support from research in psychophysics (Pizlo, 2001). Whereas Fechner (1860/1966) posited that the ‘‘percept’’ is a result of a causal chain of events emanating from the object, and giving rise to corresponding sensory data and brain processes to end with the precept of the distal stimulus, subsequent approaches, including Helmholtzian, Structural, and Gestalt perspectives, provided evidence that the percept is not directly elicited by the stimulus, but rather that it involves an unconscious inference (Helmholtz, 1910/2000) from an associated bundle of sensations. An approach developed within the computer vision community (Pizlo, 2001) treats perception as solution of an inverse2 problem that depends critically on innate constraints, or rules, for interpreting proximal stimuli (e.g., the retinal images). According to this view, ‘‘perception is about inferring the properties of the distal stimulus X given the proximal stimulus Y (Pizlo, 2001, p. 3146, emphasis added). Similar views have been articulated by Rock (1983) who discussed perceptual phenomena as inferences from premises (p. 3). Thus, for example, ‘‘Assuming the perceptual system has available in some form ‘knowledge’ about the principles of geometrical optics, objective size could be inferred or computed by taking account of distance. In following the same rules and making inferences from them, various anomalies, such as . . . the perceived motion of the afterimage, become perfectly predictable’’ (p. 37). And in the realm of vision Gregory (1997, p. 11) asserted that ‘‘Seeing objects involves general rules, and knowledge of objects from previous experience, derived largely from active hands‐on exploration.’’ Therefore, ‘‘Hypotheses of perception . . . are risky, as they are 2 Inverse in the sense that the proximal stimulus (e.g., the retinal image) originally produced by the distal stimulus is now used to decode such stimulus, going backward as it were.
262
ARIE W. KRUGLANSKI et al.
predictive and they go beyond sensed evidence to hidden properties and to the future . . .’’ (Gregory, 1997, p. 10). Finally, an Annual Review chapter ‘‘treats object perception as a visual inference problem’’ (Kersten, Mammassian, & Yuille, 2004, p. 272, emphasis added), and proposes that ‘‘the visual system resolves ambiguity through built in knowledge of how retinal images are formed and uses this knowledge to automatically and unconsciously infer the properties of objects . . .’’ (Kersten et al., 2004, p. 273, emphasis added).
D. AWARENESS Unconscious inferences are not unique to the realm of perceptional phenomena. Routinized cognitive rules as well become ‘‘eYcient,’’ requiring lesser attentional resources for the execution of judgments. As William James (1890, p. 496) aptly put it ‘‘consciousness deserts all processes when it can no longer be of use.’’ James’ parsimony principle asserts that routinization removes the need for conscious control of the process, rendering awareness of the process superfluous. Logan (1992) similarly argued that automatization of activities eVects an attentional shift to higher organizational levels (see also Vallacher & Wegner, 1985). A champion tennis player, for instance, does not need to pay attention to the correct grip, knee bending, or the following through on her strokes. Instead, she can concentrate on strategic aspects of the game, such as the type of stroke to execute and its court placement. Indeed, various judgmental eVects based on routinized IF THEN rules take place outside of conscious awareness. For instance, social cognitive work on spontaneous trait inferences (Newman & Uleman, 1989; Uleman, 1987) suggests that lawful (i.e., rule following) inferences presumably can occur without explicit inferential intentions, and without conscious awareness of the inference process (Newman & Uleman, 1989, p. 156): The spontaneous trait inference that ‘‘Mary is intelligent’’ from information that she ‘‘solved the mystery half way through the novel’’ requires the inference rule ‘‘if solves a mystery quickly then intelligent.’’ Someone who did not subscribe to that inference rule would be unlikely to draw that specific conclusion. Schwarz and Clore’s (1996) ‘‘feelings as information’’ model oVers another instance of unconsciously mediated inferences in the realm of social judgments. For instance, a transient mood state may be mistakenly attributed to one’s general life satisfaction (Schwarz & Clore, 1983; Schwarz, Servay, & Kumpf, 1985) based on an IF THEN rule linking one’s feeling state with an overall satisfaction. Or consider the well‐known use of ‘‘ease of retrieval’’ as evidence for a trait. Schwarz et al. (1991) asked participants to recall either 6 or 12 examples of situations in which they either behaved assertively and felt at ease, or behaved unassertively and felt insecure. Recalling 6 examples was experienced as easy, while recalling 12 examples
ON THE PARAMETERS OF HUMAN JUDGMENT
263
was experienced as diYcult. Subsequent self‐ratings of assertiveness indicated that participants rated themselves as less assertive after recalling 12 rather than 6 examples of assertive behavior. Apparently, the diYculty of recalling 12 instances implied to participants (in an IF THEN fashion) that they must not be very assertive. Jacoby, Kelley, Brown, and Jasechko (1989) had participants read a list of names pertaining to nonfamous individuals. Immediately afterward or following 1 day’s delay participants were given a diVerent list of names, including some of the previously presented ones, as well as other nonfamous or famous names. Especially after the delay, participants tended to misidentify the nonfamous names from the original list as famous, based on an inference from their feeling of familiarity or fluency. In most of these cases, participants were unaware of the basis of their inferences leading Schwarz and Clore (1996, p. 437) to conclude that ‘‘reliance on experiences generally does not involve conscious attribution’’ (see also, Schwarz, 2004). The foregoing review suggests that the distinction between associative versus rule‐based judgments is problematic: (1) conditioning eVects originally viewed as quintessentially ‘‘associative’’ have been generally recognized as rule‐based, that is, as a form of signal or expectancy learning (Baeyens & De Houwer, 1995; De Houwer, Thomas, & Baeyens, 2001; Tolman, 1932); (2) IF THEN judgmental rules can be routinized or ‘‘automatized,’’ and hence, occur with considerable eYciency (Bargh, 1996); and (3) they can operate outside of conscious awareness, giving rise to inferences of which true basis one may not be cognizant (Schwarz & Clore, 1996), including perceptual judgments believed to be based on hard wired inference rules (Gregory, 1997; Kersten et al., 2004; Pizlo, 2001; Rock, 1983). In a recent paper, Moors and De Houwer (2006, p. 13) wondered ‘‘whether rule‐based processes and associative ones can be distinguished empirically.’’ They further stated, and we agree that ‘‘If no functional diVerence can be found between the two processes, or if no agreement can be obtained about what this diVerence should be, rule‐based [and] associative models remain empirically indistinguishable theories’’ (Moors and De Houwer, 2006).
E. PATTERN RECOGNITION The phenomenon of ‘‘pattern recognition’’ has been occasionally juxtaposed to rule‐based process (Lieberman, Gaunt, Gilbert, & Trope, 2002). But are the two actually incompatible? Pattern recognition, after all, does constitute an inference from a given cue‐constellation (e.g., a given set of facial features or a given set of pathological symptoms) to an implied, perceptual, or conceptual judgment (e.g., that the seen face belongs to an acquaintance, or that the assembly of pathological symptoms represent a given illness). In animal research, a similar concept of cue‐configuration is explicitly treated as an
264
ARIE W. KRUGLANSKI et al.
antecedent term of an ‘‘IF THEN’’ rule. Holyoak et al. (1989, p. 319) for instance, suggested that ‘‘configural cues are not identified with perceptual units; rather they emerge during learning as multiple‐element conditions of rules.’’ In fact, Lieberman et al. (2002, p. 221, emphasis added) allow that ‘‘products of the X system’’ (assumed to operate on the basis of ‘‘pattern recognition’’) ‘‘can also be described as the result of executing ‘if then’ statements.’’ In theoretical depictions, pattern recognition is often said to depend on constraint satisfaction, that is, on a relative fit between (1) external input stimuli and (2) a preexisting structure of associations in memory. The activation of a concept is said to occur whenever such fit obtains. Although the language here may appear to diVer from the syllogistic terminology of IF THEN rules, the contents are remarkably similar. The ‘‘preexisting structure of associations in memory’’ represents a compound X that if aYrmed in a given instance by the ‘‘external input stimuli’’ indicates Y, that is, a given inference or conclusion. For instance, a conjunctive presence of ‘‘elegant attire,’’ ‘‘interest in politics,’’ and ‘‘high degree of articulateness’’ may be assumed to indicate a ‘‘lawyer’’ to an individual holding the appropriate IF THEN rule. If a newly encountered individual presented this particular ‘‘association’’ of characteristics, this could be regarded as an ‘‘external input stimulus’’ that fits the antecedent term of the rule, warranting the ‘‘lawyer’’ inference. The issue of ‘‘pattern recognition’’ recalls the question whether in a conditioning context the animal responds to the situation as a whole, that is, to an entire ‘‘Gestalt’’ versus singling out particular features of the situation in its ‘‘hypotheses’’ (Krechevsky, 1932) about the correct rule. Conditioning theorists have opted for the latter view. As Holyoak et al. (1989) put it: ‘‘A rat may receive a shock while listening to an unfamiliar tone, scratching itself, looking left, and smelling food pellets . . . the rule ‘if tone, then expect shock’ will be more likely be generated in this situation than the rule ‘If looking left, scratching, and smelling pellets, then expect shock’ ’’ (Holyoak et al., 1989, p. 320; see also, Holland, Holyoak, Nisbett, & Thaggard, 1986). A diVerent way to think about the problem of ‘‘pattern recognition’’ is in terms of the kindred notion of ‘‘attribute matching,’’ and the possibility of it serving as a basis for categorization. According to Murphy and Medin (1985), however, ‘‘Instead of attribute matching, categorization may be based on an inference process’’ (Murphy & Medin, 1985, p. 295, emphasis added). A major problem with the attribute matching view is that it views categories as mere sum total of their attributes. This ‘‘ignores the problem of how one decides what is to count as an attribute (and it involves) failing to view concepts in terms of the relations between exemplar properties and the categorization system: Human . . . theories are ignored’’ (Murphy & Medin, 1985, p. 295). Thus, somewhat analogously to a conditioning situation wherein the animal ‘‘decides’’ what features of the situation should figure in its ‘‘hypothesis’’ about the appropriate rule (Holyoak et al., 1989;
ON THE PARAMETERS OF HUMAN JUDGMENT
265
Krechevsky, 1932) so in classifying objects individuals decide, on the basis of their prior knowledge, that is, their lay theories and the specific context of application what counts as an instance of a category. In an example given by Murphy and Medin (1985, p. 295, emphasis added), the same behavior ‘‘jumping into a swimming pool with one’s clothes on . . . could imply drunkenness in one context and heroism in another (e.g., jumping into the pool to save someone from drowning).’’ In present terms, one’s lay theory or inference rule may indicate that if a pattern was encountered wherein ‘‘someone jumped into a swimming pool with one’s clothes on’’ and ‘‘someone else was drowning’’ this indicated heroism, whereas if the pattern included ‘‘drinking,’’ ‘‘jumping into a swimming pool with one’s clothes on,’’ and ‘‘no one drowning’’ this indicated drunkenness. In summary, a variety of evidence and theoretical considerations across diVerent domains of psychology converge on the notion that judgments (whether assessed directly or through behavioral manifestations) are ruled‐ based. To make a judgment is to go beyond the ‘‘information given’’ (Bartlett, 1932; Bruner, 1973) by using it as testimony for a conclusion in accordance with an ‘‘IF THEN’’ statement to which the individual subscribes. Such implicational structure appears to characterize explicit human inferences (Anderson, 1983), implicit conclusion‐drawing (Schwarz & Clore, 1996), conditioning responses in animal learning (Holyoak et al., 1989; Rescorla & Wagner, 1972), and perceptual judgments of everyday objects (Gregory, 1997; Pizlo, 2001; Rock, 1983). The basic IF THEN form appears essential to all such inferences, whether conscious or nonconscious, instantaneous or retarded, innate or learned. It is a fundamental building block out of which all epistemic edifices seem to be constructed. Thus, though the terminology of ‘‘constraint satisfaction,’’ ‘‘associative patterns,’’ ‘‘attribute matching,’’ and so on may seem rather diVerent from the rule‐ like language of IF THEN premises, the underlying structure of inference seems common to all instances of judgment. III. The Parameters of Human Judgment The foregoing assumption implies a fundamentally unitary approach to human judgment. Yet our emphasis on commonality is quite compatible with variability of judgment types and of conditions wherein contextually given information (contained in message arguments, heuristic cues, statistical data, and so on) may aVect judgments. In what follows we seek to understand the essence of such variability. Toward that aim, we now identify several parameters of the judgmental process. These are assumed to constitute continuous dimensions present in any judgmental situation. We propose that diVerent judgment types vary in their parameter values. Moreover, variability in the
266
ARIE W. KRUGLANSKI et al.
judgmental impact of the information given depends on the particular combination of parameter values characterizing a specific judgmental context.
A. INFORMATIONAL RELEVANCE Our key parameter is the degree of informational relevance. It is intimately tied to the notion of inference rules discussed above. Specifically, degree of relevance that information X has for judgment Y is defined as the extent to which the individual subscribes3 to the IF X THEN Y proposition, or the major premise of a syllogism. We assume that the relevance of a given X to a given Y may vary widely across persons as well as (for a given person) across times, representing rule‐learning. In some instances, the implication may be experienced as quite strong so that aYrming X (the minor premise) would create a strong sense that Y too is the case. Strong inferences of this sort may be aVorded by the way our perceptual system is hard wired (Pizlo, 2001). Nonetheless, perceptual learning of some sort does take place (cf. Gregory, 1997; Rock, 1983), for instance, at the level of fine discriminations. This is exemplified by a variety of perceptual expertise individuals may acquire with practice. Such discrimination learning (e.g., in the realm of sound discrimination; Wright & Fitzgerald, 2003) may involve procedural training wherein the rules of inference (our notion of ‘‘major premises’’) from a given stimulus array may be appropriately augmented, and/or an improved processing may take place of the stimulus (our notion of ‘‘minor premises’’) fitting the rules in question. As Bruner (1958, pp. 90–91) observed: ‘‘We learn the probabilistic texture of the world, conserve this learning, use it as a guide to tuning our perceptual readiness to what is most likely next. It is this that permits us to go beyond the information given.’’ Some inferential rules may be highly routinized via sustained practice (Schneider & ShiVrin, 1977), others may derive from a unique personal experience (Garcia et al., 1968), or be based on pronouncement of a trusted ‘‘epistemic authority’’ (Kruglanski et al., 2005). With lesser degree of routinization, a less impactful experience or a less trusted epistemic authority, the X to Y implication may be weaker and more tenuous. In those instances, the confidence in Y given X would be correspondingly feeble. Prior judgmental research (reviewed subsequently) often confounded diVerent informational types or contents (e.g., ‘‘heuristic’’ versus ‘‘statistical’’ information or ‘‘cue’’ versus ‘‘message’’ information) with diVerent degrees of their 3
By ‘‘subscription to’’ we mean the degree of the X–Y contingency represented in the individual’s memory, such that the subsequent activation of X in some context will tend to evoke the expectancy of Y.
ON THE PARAMETERS OF HUMAN JUDGMENT
267
potential relevance to research participants (Erb et al., 2003; Kruglanski & Thompson, 1999a,b). If so, claims about qualitative diVerences in the judgmental process may have stemmed in part from focusing on the qualitative diVerences in informational contents, rather than on the quantitative diVerences in parameter values that these contents exhibited. For example, the IF THEN rule lending relevance to a given content of information, say about the communicator’s expertise (the ‘‘expertise heuristic’’) may be believed in less strongly than an IF THEN rule lending relevance to a given message argument (i.e., a diVerent content). If the recipient had suYcient motivation and capacity to consider both, the latter information may override in its impact the former information (Pierro et al., 2004). We revisit this point at a later juncture. 1. Potential and Perceived Relevance It is useful to distinguish between potential relevance that a given item of information (X) has with respect to a given judgment (Y), and its perceived relevance in a specific situation. Potential relevance denotes the assignment of relevance under optimal processing conditions of attentional focus. It represents the knower’s degree of belief in the IF X THEN Y proposition if inquired about it directly. Often, however, conditions are suboptimal. The individual may be unable to focus on X or detect it in the informational array, or the IF X THEN Y rule may not be strongly activated in this person’s mind. In such circumstances, the information’s perceived relevance may diVer from its potential relevance, and the information’s actual impact may not be commensurate with its potential impact. In the classic study by Petty, Wells, and Brock (1976), high‐quality (hence highly potentially relevant) arguments exerted lesser persuasive impact under distraction (versus no distraction) conditions and low‐quality arguments exerted greater persuasive impact under distraction (versus no distraction) conditions. In present terms, both cases reflected a discrepancy between perceived and potential relevance under distraction (hence, reduced focus) conditions.
B. GLEANING THE RELEVANCE OF THE INFORMATION GIVEN Confronted with a judgmental question requiring an answer, an individual typically attempts to glean the relevance of the information given to the judgmental task, that is, determines its ‘‘true’’ (potential) relevance for the judgment at hand. How close he or she may come to divining such relevance may depend on two factors: (1) task diYculty or ‘‘demandingness’’ and (2) processing resources. These two factors represent our next two judgmental parameters.
268
ARIE W. KRUGLANSKI et al.
Task demands constitute an external factor analogous to the weight of a barbell. Processing resources are internal and analogous to a weight lifter’s muscular strength and determination, aVecting whether he or she would lift the weight to a required height.
C. TASK DEMANDS The task at hand may determine how easy or diYcult it is to detect the potential relevance to the judgmental question of the information given. Consistent with our syllogistic analysis, it is possible to distinguish two separate sources of task diYculty: (1) diYculty of aYrming the minor premise of the syllogism, that is, identifying that X is the case, and (2) diYculty of activating the (IF X THEN Y) inference rule (i.e., the major premise) that X may instantiate. Various circumstances may obscure the identity of the critical information X, serving as the minor premise in an inference. The stimulus package in which such information is embedded may be highly complex and lengthy. It may contain considerable noise, and the relevant signal (i.e., items of relevant evidence) may be faint, or insuYciently salient to attract the knower’s attention. The placement of the relevant informational items in the sequence presented to the knower may also matter. A front end placement may render the items easier to process, whereas a later placement may make their processing more diYcult due to the depletion of cognitive resources that the early items might have eVected. For instance, in typical persuasion studies peripheral or heuristic cues have been presented up front, and the message arguments have come subsequently. Moreover, the message arguments have been typically lengthier and more complex than the cues. In a review of the relevant literature, Pierro, Mannetti, Erb, Spiegel, and Kruglanski (2005) found this to be the case in the preponderance of instances. Either length and/or complexity and/or the order of presentation could have rendered the message arguments more diYcult to process than the heuristic cues. In Trope and Gaunt’s (2000) attributional research, saliency of information was manipulated via the modality in which it was presented. The auditory modality rendered the critical information (about contextual pressures applied on an actor) more salient, and hence easier to process, than information presented in a visual modality (the written text). In work on judgment under uncertainty (Kahneman, 2003; Tversky & Kahneman, 1974), the base rate information was often presented briefly and up front, whereas the heuristic, ‘‘representativeness’’ information typically appeared later in a relatively lengthy vignette. This might have rendered the ‘‘representativeness’’ information more challenging to process than the base rate information (for reviews
ON THE PARAMETERS OF HUMAN JUDGMENT
269
see Chun & Kruglanski, 2006; Erb et al., 2003). In person perception research (Brewer, 1988; Fiske & Neuberg, 1990), the category information was often presented briefly and early and the individuating information about the target subsequently and/or more extensively (Neuberg & Fiske, 1987). This, again, suggests that the two information types (i.e., ‘‘category’’ and ‘‘individuating’’ information) might have diVered in processing diYculty. In the realm of visual perception, correct identification of objects can be easier or more diYcult (Dosher, Liu, Blair, & Lu, 2004; Posner, 1980), depending on luminosity, external noise, or whether the target region was cued in advance of the stimulus presentation (Santhi & Reeves, 2004), and so on. The inference rules lending the ‘‘information given’’ relevance to a judgment may also be more likely activated in some contexts than in others. Thus, some environmental stimuli more than others may prime specific inference rules (Higgins, 1996) or serve as retrieval cues for such rules. Too, diVerent framings of a problem may prime diVerent rules leading to the utilization of diVerent types of information (e.g., of statistical versus ‘‘psychological’’ information) (Hilton, 1995; Schwarz et al., 1991). Contexts that prime or increase the retrieval likelihood of given inference rules increase the ease with which such rules will be applied, hence lessen the diYculty of the judgmental task involved. In summary, some judgmental tasks are more demanding than others. The reason that this matters is because in much judgmental research diVerent contents of information inadvertently diVered in their processing diYculty. Because diVerence in content is qualitative by definition, it is possible that claims for qualitatively diVerent processes of judgment often rested on the confounding of informational contents with task demands in which the latter rather than the former was actually responsible for judgmental outcomes. We revisit this issue in a subsequent section, describing our empirical work in the present conceptual paradigm. D. COGNITIVE RESOURCES Successful gleaning of informational relevance may be importantly determined by individuals’ own capabilities in a given context. Two major classes of capability factors may be distinguished related to: (1) rule accessibility and (2) attentional capacity. 1. Rule Accessibility As noted earlier, potential relevance of information given concerns the strength with which the individual subscribes to an IF THEN rule‐linking conceptual categories to each other. But individuals’ readiness to apply a
270
ARIE W. KRUGLANSKI et al.
given rule (Bruner, 1957) might be low in which case the judgmental relevance of the information given might go unrecognized. The readiness of rule application depends on its accessibility from chronic or acute sources (Chaiken, Liberman, & Eagly, 1989; Higgins, 1996; Wood, Kallgren, & Preisler, 1985). For instance, in a study by Chaiken et al. (1988) ‘‘subjects classified as chronic users of the length‐strength rule agreed more with a message that ostensibly contained many (versus few) arguments, especially after exposure to a task that primed this heuristic. In contrast, chronic nonusers of this rule were no more persuaded by the long message than by the short one, regardless of whether the length‐strength heuristic had or had not been primed’’ (Chaiken et al., 1989, pp. 217–218). In research on judgment under uncertainty, Trope and Ginossar (1988, p. 227) reviewed studies suggesting that ‘‘the influence of statistical rules, like the influence of any rule depends on their prior activation (accessibility) and their concurrent activation.’’ Presumably, that is so because the more accessible a given (in principle available) rule, the greater one’s readiness to apply it, whereas the application of less accessible rules may require laborious retrieval work. 2. Attentional Capacity An individual whose attentional capacity is taxed (e.g., by cognitive busyness with other matters) may be less able to thoroughly process and hence to adequately glean the relevance of the information given under high‐ demand conditions. This may reduce the diVerence in impact of highly judgmentally relevant versus less judgmentally relevant information. Research by Petty et al. (1976) referred to earlier demonstrated that under distraction (versus no distraction) conditions, individuals were less sensitive to quality of the message arguments. Cognitive capacity may depend also on circadian rhythm (Bodenhausen, 1990), one’s degree of mental fatigue (Webster, Richter, & Kruglanski, 1996), alcoholic intoxication (Steele & Josephs, 1990), and the degree of depletion by prior activities (Baumeister, Muraven, & Tice, 2000). The foregoing factors may reduce individuals’ processing capacity, hence increase the diYculty of ‘‘gleaning,’’ the potential relevance of the information given to the judgment at hand. Whereas the factor of rule accessibility, considered earlier, is related to the major premise of a syllogism, attentional capacity pertains both to the individual’s ability to aYrm the minor premise (by appropriately identifying the information given as the antecedent term of the rule) and to considering the minor and the major premises jointly, that is, carrying out the reasoning process from evidence to conclusion.
ON THE PARAMETERS OF HUMAN JUDGMENT
271
E. MOTIVATION: NONDIRECTIONAL AND DIRECTIONAL Whether stemming from external task demands or from internal processing limitations, gleaning diYculty may be compensated for by one’s motivation to make sense of the information. Motivation determines (1) the extent of eVort put into information processing and (2) the weights assigned to diVerent informational items as function of their compatibility with the individuals’ various wishes and desires. The former eVect is a function of one’s nondirectional motivation and the latter of one’s directional motivation (see Kruglanski, 1989, 1996, 2004 for reviews). Individuals’ extent of nondirectional motivation to thoroughly process information is determined by their various information‐processing goals, such as the goal of accuracy (Chaiken et al., 1989; Petty & Cacioppo, 1986), accountability (Tetlock, 1985), cognitive activity (Cacioppo & Petty, 1982), evaluation (Jarvis & Petty, 1996), or closure (Kruglanski, 2004; Kruglanski & Webster, 1996; Kruglanski, Pierro, Mannetti, & De Grada, 2006; Webster & Kruglanski, 1998). For instance, the higher the magnitude of the goals of accuracy or cognitive activity, the greater the degree of the processing motivation. By contrast, the higher the magnitude of the goal of closure, the lesser the degree of processing motivation. Directional motivation reflects the degree to which given contents of judgment are desired by the individual. DiVerential weight given to informational items as function of their congruence with a given directional motivation may result in bias aimed at bringing judgments in line with the individual’s wishes and desires. For instance, an individual informed that coVee drinking is unhealthy (or healthy) may appropriately distort her or his recollection of coVee drinking instances (Kunda & Sanitioso, 1989) so as to reach a desirable conclusion (see also Dunning, 1999; Kruglanski, 1999; Kunda & Sinclair, 1999). Finally, a recent review (Kersten et al., 2004, p. 284) remarked that: ‘‘object perception shows biases consistent with preferred views.’’ The notion that perceptual judgments are susceptible to motivational influences is compatible with the functional understanding of perception. In this vein, ProYtt and colleagues have found that the perception of slants of hills and of distances is aVected by whether the perceiver is wearing a heavy backpack (ProYtt, Stefanucci, Banton, & Epstein, 2003), is old or young (Bhalla & ProYtt, 1999), is fatigued (ProYtt, Bhalla, Gossweiler, & Midgett, 1995), or contemplates action goals (Witt, ProYtt, & Epstein, 2004). Such perceptual biases may guide the mobilization of resources needed for motivated action, and in this sense are functional. In summary then, directional motivational eVects have been observed at diVerent levels of psychological phenomena, including the very basic ones in the realm of perception.
272
ARIE W. KRUGLANSKI et al.
As the judgmental activity unfolds, magnitude of processing motivation may be determined by the desirability of initially formed beliefs. Should these be desirable‐the individual would be disinclined to engage in further information processing, lest current conclusions be undermined by further data. On the other hand, should the initial beliefs be undesirable‐the individual would be inclined to process further information that hopefully would serve to sweep away the initial, unpalatable notions (Ditto and Lopez, 1992). In other words, directional motivation may determine the degree of work individuals are prepared to invest in information processing en route to a judgment. F. PROPERTIES OF THE JUDGMENTAL PARAMETERS 1. Multiple Determination of Parameter Values As presently conceived, the judgmental parameters constitute quantitative continua whose specific values are determined by diverse factors: A given informational stimulus may aVord a strong inference because its perceived relevance was innately ‘‘wired into’’ the human perceptual system, because such relevance was learned over repeated experience (Neal et al., 2006), or because it was derived from highly regarded ‘‘epistemic authority’’ (Kruglanski et al., 2005), and so on. Similarly, task demands could be multiply determined by informational complexity, signal to noise ratio, ordinal position, or perceptual modality. Cognitive capacity could be determined by rule accessibility, in turn aVected by the recency or frequency of its activation (Higgins, 1996), and/ or by cognitive capacity determined by cognitive load, fatigue, and depletion occasioned by prior pursuits (Baumeister et al., 2000). Motivation could be determined by expectancies and values attached to a variety of judgmental outcomes and processes, for example to the cognitive activity itself (Cacioppo & Petty, 1982), to cognitive closure (Kruglanski, 2004; Kruglanski & Webster, 1996), accuracy (Funder, 1987; Kruglanski, 1989), accountability (Tetlock, 1985), impression management (Chaiken et al., 1989), ego enhancement (Kunda, 1990), and so on. This heterogeneity notwithstanding, we are assuming that as far as information’s impact is concerned, diverse determinants of a given parameter’s values are functionally equivalent. From this perspective, it matters not why a given information is subjectively relevant to a given judgment, why a given judgmental task is demanding for an individual, why an individual’s cognitive resources at a given moment are ample or sparse, why an individual is motivated or unmotivated to expend eVorts on the processing of given judgmentally relevant information, it matters only that the above parameter values occur at given magnitudes.
ON THE PARAMETERS OF HUMAN JUDGMENT
273
2. Orthogonality of the Parameters We assume further that the judgmental parameters are orthogonal and that their values derive from largely independent determinants. Thus, subjective‐ relevance of information may derive from a prior forging of conditional IF THEN links between informational categories, the magnitude of processing motivation may derive from various goals that persons might have, task demands may depend on nature of the problem posed, and the stimulus context, and cognitive resources may depend on rule accessibility and cognitive busyness, all representing very diVerent concerns. Nonetheless, under some conditions the parameters may exert influence on one another. For instance, highly relevant information might be used more frequently than less relevant information, resulting in its greater accessibility, in turn elevating the level of the cognitive resource parameter. Conversely, high accessibility of information might increase its perceived relevance in some contexts, for example by lending it experiential ‘‘fluency’’ (Jacoby et al., 1989; Schwarz & Clore, 1996). A given bit of information may be perceived as more relevant to a judgment the more congruent it is with the knower’s wishes and desires (Lord, Ross, & Lepper, 1979). Thus, in order to justify their ‘‘freezing’’ on early information persons under high need for closure might perceive it as more relevant to a judgment at hand than persons under low need for closure (Webster & Kruglanski, 1998). By contrast, persons with a high need for cognition (Cacioppo & Petty, 1982) may perceive the early information as less relevant, in order that they may carry on with their information‐processing activity. Finally, limited cognitive capacity may reduce processing motivation or induce a need for cognitive closure (cf. Kruglanski & Webster, 1996), and so on. Despite these interrelations, however, the judgmental parameters are relatively independent because for the most part their determinants are unique or nonoverlapping.
3. Independence of Informational Contents Inevitably, some parametric values characterize some informational contents. For instance, low‐task demands and/or subjective relevance may have characterized in prior research ‘‘peripheral’’ or ‘‘heuristic’’ cues, whereas higher task demands and subjective relevance may have characterized ‘‘message or issue information’’ (Chaiken et al., 1989; Petty & Cacioppo, 1986). For many individuals (lacking statistical know how), low degrees of perceived relevance may characterize statistical information (e.g., base rates) (Erb et al., 2003; Hilton, 1995). High saliency and hence low‐processing
274
ARIE W. KRUGLANSKI et al.
demands may characterize social category information in some cases (Brewer, 1988; Fiske & Neuberg, 1990), and so on. From the present perspective, it is essential to conceptually distinguish between parametric values and the contents to which such values may be attached in some contexts. Whereas some parametric values must characterize any informational contents, the very same values (e.g., a given degree of subjective relevance, task demands, and motivational significance) may characterize numerous alternative contents as well. Moreover, it is the parametric values rather than the contents that determine the judgmental impact of information. In other words, a given combination of parameter values is assumed to have the same degree of impact across diVerent possible contents having those same values. By contrast, the same informational contents, if characterized by diVerent parameter values (e.g., the same information diVering in subjective relevance or degree of motivational significance to diVerent knowers), are assumed to diVer in their judgmental impact.
G. MULTIDIMENSIONAL PARAMETRIC SPACE It is possible to conceptualize the constellations of parametric values characterizing diVerent judgmental contexts as points in a multidimensional space defined by the present quasi‐orthogonal parameters. Each such point is assumed to diVer from all the others in the impact the associated informational contents will create. Whereas in theory each of our continuous parameters allows an open ended amount of fine gradations, in practice we have identified some relatively coarse parametric regions and compared them in empirical research (e.g., by comparing the impact of a highly demanding information conjoined to high‐ (versus low‐) processing motivation and/or cognitive resources). In summary, we have listed five general parameters whose joint operation may determine whether and to what extent the information given would aVect judgments. These are: (1) the subjective relevance of the information, (2) task demands, (3) cognitive resources, (4) nondirectional, and (5) directional motivation. Each parameter is assumed (1) to constitute a quantitative continuum, (2) to be represented at some value in each judgmental situation, (3) to be (largely) orthogonal to the remaining parameters, (4) to be multiply determined, and (5) conceptually independent of informational contents to which it happens to be associated. In the next section, we outline a set of specific postulates and derivations concerning the role our parameters play in determining the judgmental impact of the information given.
ON THE PARAMETERS OF HUMAN JUDGMENT
275
IV. A Parametric Model of Social Judgment A. THE ROLE OF SUBJECTIVE RELEVANCE Postulate 1: Judgmental impact of information is a positive function of its degree of perceived relevance in a given context. Derivation 1 (from Postulate 1): The impact of the information given will depend on the likelihood of its relevance being adequately perceived. In turn, such likelihood will be determined by the degree to which the information is accessible in memory and instantiated with respect to a given object of judgment.4 Evidence for this derivation comes from three studies completed by Kopetz and Kruglanski (2006), investigating an eVect originally reported by Pavelchak (1989), and oVered in support of a dual‐mode model of impression formation. In the first session of Pavelchak’s (1989) experiment, participants rated the likeability of 35 academic majors and 50 personality traits. In the second session carried out 10–14 days later, participants were presented with six stimulus persons, each portrayed via four traits. In the category condition, participants first guessed the targets’ academic majors, then rated their likeability. In the piecemeal condition, participants rated the targets’ likeability after exposure to their traits but before guessing their majors. It was found that participants in the category condition (those who categorized the targets prior to rating their likeability) made likeability ratings that were more consistent with the categories’ likeability than with the likeability of the traits. Participants, in the piecemeal condition, made likeability ratings that were more congruent with the likeability of the traits than with the likeability of the categories. Pavelchak (1989, p. 361) viewed these results as ‘‘clear evidence that there are two distinct modes of person evaluation: one computed from attribute evaluations and one based on category evaluations.’’ It is possible, however, that the findings were due to diVerential accessibility of the category and the trait information in Pavelchak’s (1989) two conditions. Because in the category condition, participants made their likeability ratings immediately after exposure to the targets’ guessed categories, category information may have been more recent and hence more accessible in their memory than the trait information. Similarly, because in the piecemeal 4
By instantiation is meant applicability to a given object of judgment, thus if the inference rule states that ‘‘if X then Y,’’ instantiation with respect to an object A would amount to aYrming that A is X (and hence that Y is expected).
276
ARIE W. KRUGLANSKI et al.
condition participants made their likeability ratings immediately after exposure to the targets’ presented traits, trait information may have been more recent and hence more accessible in their memories than category information. In other words, Pavelchak’s (1989) results could have been due to diVerential accessibility rather than to qualitative diVerences in processing category and trait information as implied by the dual‐mode approach. To test this possibility, Kopetz and Kruglanski’s (2006) first study carried out an extended replication of Pavelchak’s (1989) two sessions experiment. During the first session, participants evaluated the likeability of 48 academic majors and 66 personality traits. In the second session, taking place 10–14 days later, participants in a replication condition were presented with six stimulus persons characterized by two personality traits and were asked either (1) to guess the target person’s likely academic major and then to rate the person’s likeability or (2) to rate the target’s likeability and then guess this individual’s likely major. In two novel conditions, the targets were characterized by two academic majors and asked (3) to guess the target’s most likely trait, and then rate this individual’s likeability, or (4) to rate the target’s likeability and then to guess this individual’s most likely trait. The replication condition obtained results similar to those of Pavelchak (1989). Participants who guessed the target’s category prior to rating this person’s likeability exhibited smaller deviation of rated likeability from liking for the category than liking for the traits, whereas those who guessed the target’s category (this person’s major) after exposure to the traits but before guessing the category exhibited smaller deviation of the target’s rated liking from the traits’ (versus the category’s) likeability. A reversal of these findings obtained in Kopetz and Kruglanski’s (2006) novel conditions: Where participants guessed the target’s trait prior to rating her or his likeability, there was smaller deviation of rated likeability from liking for the trait (versus liking for the categories). However, where participants rated the target’s likeability prior to guessing her or his likely trait, there obtained a smaller deviation of the target’s rated likeability from liking for the categories (versus liking for the trait). These findings are illustrated in Fig. 1. In Kopetz and Kruglanski’s (2006) subsequent study, participants rated in the first session the likeability of various traits and majors. In the same session, participants were presented with six target persons each depicted via four personality traits and were asked to guess each person’s academic major. In a second session (10–14 days later), participants were presented with the same target persons described in terms of the same traits and categories (in counterbalanced order). Whereas in the prior study we manipulated the presentation order of specific categories or specific traits pertaining to the targets, in the present study we primed the general constructs of ‘‘categories,’’ or ‘‘traits’ ’’ by having participants, prior to rating their liking for the targets
ON THE PARAMETERS OF HUMAN JUDGMENT
277
1.2
Degree of deviation
1.0 0.8 0.6 0.4 0.2 0.0 Trait-categoryTraitCategory-trait- Categoryjudgement judgement- judgement judgement-trait category Category deviation
Piecemeal deviation
Fig. 1. Deviation of six targets’ rated likeability from the likeability of the target’s categories (college majors) and descriptive traits (Kopetz & Kruglanski, 2006, Study 1).
1.2
Degree of deviation
1.0 0.8 0.6 0.4 0.2 0.0 Category prime Category deviation
Trait prime Piecemeal deviation
Fig. 2. Deviation of six targets’ rated likeability from the likeability of the target’s categories (college majors) and descriptive traits as function of priming the concepts of ‘‘category’’ versus ‘‘traits’’ (Kopetz & Kruglanski, 2006, Study 2).
278
ARIE W. KRUGLANSKI et al.
either categorize 20 diVerent objects into groups or describe 20 diVerent objects in terms of their traits. The results (see Fig. 2) showed that when the concept of ‘‘category’’ was primed, participants liking of the target persons was less deviant from their liking for the target’s category (i.e., her or his major) than from their liking for the target’s traits, whereas when the concept of ‘‘trait’’ was primed, participants liking of the target persons was less deviant from their liking for the targets’ traits than from their liking for the target’s category. In the third and final study using a similar two session paradigm, Kopetz and Kruglanski (2006) investigated the notion that accessibility of information is likely to increase its impact only to the extent that the information appears to apply to, or is instantiated for, the specific target of judgment: After having rated in the first session, the likeability of various traits and major categories in the second session participants evaluated six stimulus persons each depicted via two academic majors or two personality traits. To assess credibility, participants also rated the most likely and the most unlikely academic major of the six stimulus persons depicted via the traits, and the most likely traits for the six persons depicted via the academic majors. At that point, participants were shown a picture of each target person and asked to rate her or his likeability. Concomitantly with the presentation of the picture, participants were subliminally primed with either the most likely or the most unlikely major for targets depicted via the traits, or the most likely or the most unlikely trait for targets depicted via the majors. It was found (see Fig. 3A) that when participants were presented with targets depicted by traits and then primed with their academic majors, participants liking for the targets was less deviant from liking for the primed categories than the depicted traits, but only when the categories were instantiated for the specific target, that is were consistent with the target’s alleged traits. Similarly, when participants were presented with targets depicted by the categories (see Fig. 3B), their liking for those targets was less deviant from liking for the primed traits than liking for the targets’ categories but, again, only when the primed traits were instantiated for the specific target, that is, when they were consistent with the categories to which each target person was said to belong. In summary, the Kopetz and Kruglanski (2006) studies support the notion that Pavelchak’s (1989) results may have been due to the diVerential accessibility of category and trait information rather than due to a fundamental diVerence in the way in which these two information types are processed. Our experiments suggest that both types of information have greater impact when they are perceived to instantiate the minor premise of the syllogism that lends them relevance. In turn, the likelihood of such a perception depends on the accessibility of the information and its consistency with what is already known about a given target of judgment.
ON THE PARAMETERS OF HUMAN JUDGMENT
279
A 1.2 Degree of deviation
1.0 0.8 0.6 0.4 0.2 0.0 Traits-consistent category
Traits-inconsistent category
Category deviation
Piecemeal deviation
B 1.4
Degree of deviation
1.2 1.0 0.8 0.6 0.4 0.2 0.0 Category-consistent traits Category deviation
Category-inconsistent traits Piecemeal deviation
Fig. 3. (A) Deviation of six targets’ rated likeability from the likeability of the target’s traits‐ consistent versus traits‐inconsistent primed categories. (B) Deviation of six targets’ rated likeability from the likeability of the targets’ category‐consistent versus category‐inconsistent primed traits (Kopetz & Kruglanski, 2006, Study 3).
Derivation 2 (from Postulate 1): Assuming that the (potential) relevance of information is adequately perceived, more subjectively relevant information will exert greater judgmental impact than less subjectively relevant information.
280
ARIE W. KRUGLANSKI et al.
Though seemingly straightforward and simple, this derivation illuminates some intriguing findings in social cognition. These have to do with the shifting conditions under which the same information will exert greater or lesser impact. 1. Teaching Relevance We assume that the potential relevance of information is not fixed and that it can be altered through learning. Pertinent in this connection is work by Nisbett, Fong, Lehman, & Cheng (1987) and by Sedlmeier (1999) on the successful teaching of statistical reasoning. From the present perspective, such teaching imparts statistical rules to individuals hence it strengthens the ‘‘IF THEN’’ connections between statistical concepts (e.g., the base rates of some event’s occurrence) and likelihood judgments. Indeed, research has demonstrated that teaching statistical reasoning results in an increased use of statistical information. As Sedlmeier (1999, p. 190) expressed it: ‘‘The pessimistic outlook of the heuristics and biases approach cannot be maintained . . . Training about statistical reasoning can be eVective.’’ 2. Framing EVects Contextual relevance of diVerent types of information can be inferred from the framing of the situation, that is, from the way the situation is apprehended by the individual. From this perspective, neglect of statistical information (e.g., the base rates) and the tendency to rely on individuating profile (representativeness) information in early studies (Tversky & Kahneman, 1974) might have stemmed from a ‘‘psychological’’ framing of the problem. Consistent with this possibility, work by Hilton and Slugoski (2001), Schwarz et al. (1991), and by Zukier and Pepitone (1984) established that framing the same problems as ‘‘statistical’’ or ‘‘scientific’’ appreciably reduces the neglect of statistical information. The key role of subjective relevance in determining the judgmental impact of statistical information hardly escaped the attention of researchers (cf. Borgida & Brekke, 1981). For instance, Bar‐Hillel (1990, p. 201) concluded that base‐rates are by and large neglected if and when they are considered to be irrelevant to the prediction at hand furthermore in the tasks that dominate laboratory studies of base rate neglect, base rates provide only a general informational background on which other information, which typically pertains more directly or specifically to the target case, is added. Such information tends to render the arbitrary base rates subjectively irrelevant . . . (emphasis added).
From the present perspective, the subjective relevance parameter applies as much to statistical information as it does to ‘‘heuristics’’ to which ‘‘statistics’’
ON THE PARAMETERS OF HUMAN JUDGMENT
281
are often juxtaposed. Whichever of these two information types would appear more relevant to the judgment at hand is likely to be used as evidence for the judgment. Ginossar and Trope (1980) discovered, for example, that when the individuating vignette (the ‘‘representativeness’’ information) was rendered nondiagnostic, and hence subjectively irrelevant to determining the target’s profession, participants did utilize the base rates. We revisit at a later juncture the issue of relative (subjective) relevance of diVerent information types. 3. Trait Centrality in Impression Formation The phenomenon of trait centrality (Asch, 1946), long regarded as a major psychological insight into person perception, turns out to be a matter of subjective relevance as well. In other words, a given stimulus trait (e.g., ‘‘warm’’ or ‘‘cold’’) impacts personality impressions (e.g., that the target is friendly or cooperative) to the extent that it implies the impressions in question (Wishner, 1960; Zanna & Hamilton, 1972). In this sense, the well‐known ‘‘warm‐cold’’ eVect is nonunique. For instance, whereas ‘‘warmth’’ but not ‘‘speed’’ may generally imply ‘‘friendliness,’’ ‘‘speed’’ but not ‘‘warmth’’ may generally imply ‘‘athleticism.’’ In this vein, Fishbach, Kruglanski, and Chun (2005, Study 1) found that informing participants that a target is (among other characteristics) ‘‘warm’’ versus ‘‘cold’’ led them to conclude that she is also ‘‘friendly,’’ ‘‘likeable,’’ and so on. Including in the same trait list the information that the target is ‘‘fast’’ versus ‘‘slow,’’ however, had no eVect on these personality dimensions. By contrast, including in the same trait list the information that Mike, a candidate for a basketball team, is ‘‘fast’’ versus ‘‘slow’’ led participants to infer that he is a good prospect for the team, and is characterized by other athletic traits, whereas information that she is ‘‘warm’’ versus ‘‘cold’’ had no eVect on this and related judgments (see Fig. 4). 4. Subjective Relevance and Lay Theories As the term suggests, subjective relevance may vary across individuals in accordance with their lay theories about implicational IF THEN relations among categories. In this vein, Fishbach et al. (2005, Study 2) found that people’s lay theories about the degree to which ‘‘warmth’’ implies ‘‘friendship,’’ or ‘‘fastness’’ implies ‘‘basketball ability’’ mediated the relation between a list including the ‘‘warm’’ versus ‘‘cold’’ items and judgments of friendship, and between a list including the ‘‘fast’’ versus ‘‘slow’’ items and judgments of basketball potential (see Fig. 5). Ginossar and Trope (1980) assessed the degree to which participants possessed the sampling rule that links base rates to probability judgments. ‘‘It turned out that most subjects based their estimates on the base rate frequencies, but a sizable
282
ARIE W. KRUGLANSKI et al.
Evaluation
6
5
4
3 Warm
Cold Fast Trait information Basketball player
Slow
Friend
Fig. 4. Evaluation of target as a function of trait information and impression task (Fishbach et al., 2006, Study 1).
Trait: fast versus warm
.11
(.38*)
Interest in recruiting
(.64**)
.71**
.73** Perceived predictive value Trait: warm versus fast
.08
(.54**)
Interest in friendship
(.61**)
.84** .54* Perceived predictive value
Fig. 5. Mediation of impressions by lay theories (Fishbach et al., 2006, Study 2). *p < .05; **p < .01. Note: numbers in parentheses are zero‐order standardized s.
ON THE PARAMETERS OF HUMAN JUDGMENT
283
minority did not’’ (Trope & Ginossar, 1988, p. 215). Consistent with the present analysis, participants who possessed the sampling rule exhibited greater utilization of the base rates than ones who did not possess the rule. These results demonstrate, again, that the subjective relevance of information (in this case base rate information) may vary across people, and that it determines their judgments in response to the information given. B. THE ROLE OF TASK DEMANDS AND PROCESSING RESOURCES Postulate 2: The likelihood of recognizing the potential relevance of the information given is a direct function of one’s cognitive resources and one’s processing motivation and is an inverse function of task demands. Definition: Let the combination of cognitive resources and processing motivation defined an individual’s processing potential Derivation 3 (from Postulates 1 and 2): The higher the task demands, the greater the processing potential needed for the information given to exert judgmental impact correspondent with its potential relevance. Corollary to Derivation 3: Where the processing potential is insuYcient given the level of task demands, the information given will fail to exert judgmental impact commensurate with its potential relevance.
1. Persuasion Research A pervasive finding in persuasion research has been that ‘‘peripheral’’ or ‘‘heuristic’’ cues exert judgmental impact (i.e., eVect change in recipients’ attitudes or opinions) under conditions of low‐processing resources, for example where recipients’ interest in the task is low, when they are cognitively busy or distracted, when their need for cognition is low, and so on. By contrast, ‘‘message arguments’’ have been found to exert their eVects typically under conditions of high‐processing potential (e.g., high interest in the task, or ample cognitive capacity). In reviews of these studies (Erb et al., 2003; Kruglanski & Thompson, 1999a,b; Kruglanski, Thompson, & Spiegel, 1999; Kruglanski, Chen et al., 2006; Pierro et al., 2005), it became apparent, however, that often in persuasion research the type of the information (i.e., ‘‘peripheral’’ or ‘‘heuristic’’ cues versus message arguments) was confounded with task demands. Because the message arguments were typically lengthier, more complex, and were placed later in the informational sequence, their processing may have imposed higher processing demands than the processing of ‘‘cues’’ that
284
ARIE W. KRUGLANSKI et al.
were typically brief, simple, and presented up front. When these confoundings were experimentally removed, the previously found diVerences between conditions under which the ‘‘cues’’ versus the ‘‘message arguments’’ (or vice versa) exerted their persuasive eVects were eliminated. In one study (Kruglanski and Thompson, 1999a, Study 4), brief expertise information conveyed by the communicator’s status (professor in a high‐ versus low‐prestige university) was followed by a lengthy expertise information presented via the speaker’s curriculum vitae. Under cognitive load that limited recipients’ processing resources, the brief expertise information aVected judgments, whereas the lengthy expertise information did not. In the absence of cognitive load, by contrast, it was the lengthy expertise information that impacted judgments but not the brief expertise information. In another study (Pierro et al., 2005, Study 1), it was found that brief message arguments impacted judgments under low‐motivational involvement in the issue (hence, under low‐processing resources), whereas lengthy subsequent arguments impacted judgments under high‐motivational involvement (hence, high‐processing resources). Other studies obtained similar results (for a review see Kruglanski et al., 2006). In a recent study, Pierro, Mannetti, and Kruglanski (2006) extended this work to examine the eVects of length (and hence diYculty) of processing information about an expert or an inexpert source and the degree of motivational involvement on (1) attitude stability over time and (2) attitude behavior relations. In the first phase of the study, participants received an appeal from a person introduced via a brief or a lengthy curriculum vitae that implied him to be an expert or an inexpert in psychological science (a full professor in cognitive psychology at the University of Milan versus an instructor in the psychology of tourism at a technical institute). The appeal introduced a novel proposal to institute a compulsory participation in psychology experiments for psychology majors at Italian universities. In the high‐involvement condition, participants were led to believe that the new program is about to begin next year, and hence that they themselves would be aVected. In the low‐involvement condition, participants were told that the program will commence 5 years hence, and hence that it would not apply to them. Participants were also invited to participate in an experiment said to take place a month later. In the first phase of the study, following the presentation of the persuasive appeal (the same in all conditions) participants’ attitudes and behavioral intentions to participate in the experiment were measured. The attitude results replicated and extended those of Pierro et al. (2005). Participants presented with the brief expertise information showed more positive attitudes toward the proposal and corresponding behavioral intentions, as a function of source expertise under low involvement but not under high involvement, whereas participants presented
ON THE PARAMETERS OF HUMAN JUDGMENT
285
with the lengthy expertise information showed more positive attitudes and behavioral intentions as a function of source expertise under high involvement but not under low involvement. Of greater interest were the attitude and intention data collected at phase 2, conducted 3 weeks later. As shown in Figs. 6 and 7, the attitudes and behavioral intentions of the brief information participants showed no eVect of expertise in either the low‐ or the high‐involvement conditions attesting to low attitude stability (as predicted by the ELM). By contrast, the attitudes and intentions of the lengthy information participants exhibited an eVect of expertise in the high‐ but not in the low‐involvement condition. Finally, as shown in Fig. 8, the actual behavior assessed a month following the persuasive manipulation revealed an eVect of source expertise only for high‐involvement participants who received the lengthy source information. Appropriate multiple regression results yielded similar results suggesting together that attitude stability and the consistency between attitudes and behavior depend on the relation between task demands and individuals’ processing potential (operationalized in this study via motivational involvement). The extensive elaboration that occurs when the task demands are considerable and when the processing potential is high increases the robustness of the attitudes formed as manifested in their stability over time and their ability to predict behavior. Thus, the eVects predicted by the ELM for message arguments (Petty & Cacioppo, 1986) seem to also hold for source information of 9 8 Attitude change
7 6 5 4 3 2 1 0 Brief-low involvement
Brief-high involvement Inexpert
Long-low involvement
Long-high involvement
Expert
Fig. 6. Attitude change 3 weeks after the exposure to the persuasive message as a function of length of the message, source expertise, and involvement (Pierro et al., 2006).
286
ARIE W. KRUGLANSKI et al.
9 Behavioral intentions
8 7 6 5 4 3 2 1 0 Brief-low involvement
Brief-high involvement Inexpert
Long-low involvement
Long-high involvement
Expert
Fig. 7. Behavioral intentions 3 weeks after the exposure to the persuasive message as a function of length of the message, source expertise, and involvement (Pierro et al., 2006).
80 70
Behavior
60 50 40 30 20 10 0 Brief-low involvement
Brief-high involvement Inexpert
Long-low involvement
Long-high involvement
Expert
Fig. 8. Behavior 1 month after the exposure to the persuasive message as a function of length of the message, source expertise, and involvement (Pierro et al., 2006).
comparable processing diYculty. Considered collectively with the bulk of prior persuasion research (Chaiken, Wood, & Eagly, 1996), these findings support the hypothesis that appreciating the (subjective) relevance of the information given (e.g., the diVerence between communication from an
ON THE PARAMETERS OF HUMAN JUDGMENT
287
expert versus a nonexpert) hence enabling it to exert its eVects on attitudes and behaviors depends on the match between recipients’ (cognitive and motivational) processing resources and the demands of the information processing task. 2. Dispositional Attributions A major question posed by attribution researchers concerned the process whereby a given behavior emitted by an actor is causally ascribed to the situational context, or to the actor’s disposition. In this vein, Trope (1986) reviewed evidence that ambiguous behaviors tend to be disambiguated by an assimilation to the context in which they are taking place. For instance, an ambiguous facial expression is likely to be perceived as sad if the context is sad as well (e.g., a funeral) and as happy if the context was happy (e.g., a party). Once the behavior had been identified, however, and the question of its causal origin was pondered, the context should play a subtractive (rather an assimilative) role in determining the behavior’s causal attribution. Specifically, the role of the context is subtracted to determine the role of the actor’s disposition in producing the behavior. For instance, if the context was sad, an individual’s sad expression would tend not to be attributed to the actor’s dispositional sadness because other persons in the same situation would probably seem sad as well. Of present interest, Trope and Alfieri (1997) found that the assimilative process of behavior identification was independent of cognitive load, whereas the subtractive process of dispositional attribution was undermined by load. These investigators also found that invalidating the contextual information did not manage to erase its eVect on the behavioral identification, whereas it did erase it on the dispositional attribution. Two alternative explanations may account for these results: (1) that the two processes are qualitatively distinct and (2) that for some reason the behavior identification task in Trope and Alfieri’s (1997) studies was less demanding than the dispositional attribution task, hence that it was less sensitive to load, and perhaps carried out more automatically and hence less impacted by subsequent (invalidating) information. Consistent with the latter interpretation, Trope and Gaunt (2000) discovered that when demands associated with the dispositional attribution task were lowered (e.g., by increasing the salience of the information given), the subtraction of context from dispositional attributions was no longer aVected by load. Furthermore, Chun, Spiegel, and Kruglanski (2002) found that when the behavior identification task was made more diYcult (e.g., by decreasing the salience of the information given) it too was undermined by load. Moreover, under those conditions invalidating the information on which the behavioral identifications were based managed to undo those identifications.
288
ARIE W. KRUGLANSKI et al.
These findings are consistent with the notion that when a judgmental task (e.g., of ‘‘behavior identification’’ or of ‘‘dispositional attribution’’) is suYciently demanding, its adequate performance requires cognitive resources and can be undermined by load. Furthermore, addressing such a task can be a conscious, deliberative process registered in awareness. Consequently, invalidating the informational input into this process is likely to be taken into account, resulting in appropriate adjustments to the judgments rendered. When the task is substantially less demanding, however, it requires correspondingly less resources, possibly to the point of immunity from interference by (some degrees of ) load. Furthermore, under such conditions the process may occur so quickly and subconsciously that its details are not fully encoded. Hence, invalidating the informational input into this process may not occasion corrective adjustments to the pertinent judgments. 3. Base‐Rate Neglect We have suggested that the judgmental impact of information depends on appreciating its (subjective) relevance to the question at stake, and that such appreciation, in turn, depends on the relation between task demands and processing resources. Jointly, these notions are capable of casting a new light on the problem of base rate neglect and on conditions under which statistical versus ‘‘heuristic’’ information may impact individuals’ judgments. In the original demonstrations of base rate neglect (Kahneman & Tversky, 1973), the base rate information was typically presented briefly, via a single sentence, and up front. By contrast, the individuating (representativeness) information was presented subsequently via a relatively lengthy vignette. If one assumes that participants in such studies had suYcient motivation and cognitive capacity to wade through the entire informational package with which they were presented, they might have managed to fully process the later, lengthier, and hence more demanding vignette information, to have paid it ample attention, and consequently to have given it considerable weight in the ultimate judgment. This is analogous to the finding in persuasion studies that the lengthier, later appearing, message argument information but not the brief, up front appearing, ‘‘cue’’ information typically had impact under ample processing resources (e.g., of high‐processing motivation and cognitive capacity). If the above is true, we should be able to ‘‘move’’ base rate neglect around by reversing the relative length and ordinal position of the base rate and the individuating (representativeness) information. A series of studies by Chun and Kruglanski (2006) attempted just that. In one condition of their first study, the typical lawyer–engineer paradigm (Kahneman & Tversky, 1973) was replicated via a presentation of brief and up‐front base rate information followed by lengthier individuating information.
ON THE PARAMETERS OF HUMAN JUDGMENT
289
In another condition, these relations were reversed by presenting brief individuating information first, followed by lengthier and more complex base rate information (in which the overall base rate of lawyers and engineers was decomposed into base rates of the various subcategories of lawyers and engineers). As predicted, the former condition replicated the typical finding of base rate neglect, whereas the latter condition evinced considerable base rate utilization. A subsequent study added a manipulation of cognitive load. The former results were now replicated in the low‐load condition, but were reversed in the high‐load condition. Regardless of information type, under load the brief up‐front information was utilized more than the lengthy subsequent information, whereas in the absence of load the lengthy and subsequent information was utilized more. The next study used two types of individuating information, presented in two sequences. In one sequence, brief information consistent with the engineer stereotype was followed by lengthy information consistent with the lawyer stereotype. In the alternative sequence, brief information consistent with the lawyer stereotype was followed by lengthy information consistent with the engineer stereotype. It was found that heightened cognitive load led to reliance on the brief and up‐front stereotype, whereas the absence of load prompted reliance on the lengthier and subsequent stereotype. Finally, the last study presented participants with information about two samples. In one condition, the first sample consisting of 30% engineers and 70% lawyers (the 30/70 sample) was followed by a second sample consisting of 70% engineers and 30% lawyers (the 70/30 sample). In a second condition, the first sample consisted of 70% engineers and 30% lawyers whereas the second sample consisted of 30% engineers and 70% lawyers. In addition, we manipulated load. It was found that the first sample was relied more under load, and the second under no load. Hence in the 30/70; 70?30 condition cognitive load decreased the perceived likelihood of the target being an engineer, whereas in the 70/30; 30/70 condition cognitive load increased the perceived likelihood of the target being an engineer, whereas in the 70/30 sample such likelihood was higher under load (versus no load). To summarize then, evidence across domains (i.e., of persuasion, attribution, and judgment under uncertainty) supports Derivation 2 that the higher the task demands, the greater should be the processing resources if the information given is to exert judgmental impact commensurate with its potential relevance.
C. RELATIVE INFORMATIONAL IMPACT In much judgmental research, participants are presented with several (typically two) types of information and the question posed is which of the two has the greater judgmental impact and under what conditions. For instance,
290
ARIE W. KRUGLANSKI et al.
persuasion researchers wondered when do peripheral or heuristic cues have greater impact than message arguments and when does the opposite hold true (Chaiken et al., 1989; Petty & Cacioppo, 1986). Workers in the domain of biases and heuristics (Borgida & Brekke, 1981; Ginossar & Trope, 1980; Hilton & Slugosky, 2001; Kahneman, 2003; Trope & Ginossar, 1988; Tversky & Kahneman, 1974) inquired when statistical information (e.g., base rates) is neglected in favor of simplistic rules of thumb (or heuristics) and when it is taken into account. Workers in the area of impression formation (Brewer, 1988; Fiske & Neuberg, 1990; Fiske, Lin, & Neuberg, 1999) asked when individuating information about the target is taken into account, and when is it neglected in favor of social category information, and so on. Often, the diVerent types of information presented to participants have (inadvertently) diVered in their subjective relevance to these persons. For instance, in the domain of persuasion Pierro et al. (2004) carried out an extensive content analysis of experimental materials in persuasion studies and concluded that, typically, the cues presented to participants were judged as less relevant to the judgmental (attitudinal) topic than were the message arguments. Note that in much persuasion research the cues but not the message arguments exerted judgmental impact under low‐processing resources, whereas the message arguments did so under high‐processing resources. From the present perspective, it is possible to interpret these findings in terms of the following general derivations: Derivation 4(a) (from postulates 1 and 2): Given suYcient processing potential, the more relevant information will have a greater impact on judgments than the less relevant information. Derivation 4(b) (from postulates 1 and 2). In the absence of suYcient processing potential, the easier to process information (assuming an above threshold relevance) will have a greater judgmental impact than the more diYcult to process information.
Pierro et al. (2004) tested these notions in a series of three experimental studies. They all employed the same 2 2 2 2 factorial design with the variables: (1) processing motivation‐manipulated via accountability instructions (high versus low) (Tetlock, 1985), (2) valence of the early informational set about product features (positive or negative with respect to the attitude object), (3) valence of the later informational set about product features (positive or negative with respect to the attitude object), and (4) relative attitudinal relevance5 of the two informational sets (the early set less relevant than the later set or vice versa). Also common to all three studies was content of the target judgments, having to do with relative desirability of a given brand of a cellular phone compared to its competitors. The studies diVered, however, in contents of the information given 5
Ascertained via appropriate pilot studies.
ON THE PARAMETERS OF HUMAN JUDGMENT
291
on which basis participants were to reach their judgments. In the first study, both the early and the later information consisted of message arguments (about properties of the target phone), in the second study both consisted of heuristic information (namely, pertinent to the ‘‘consensus heuristic,’’ and containing information about opinion polls as to the target phone’s attributes), and in the third study, contrary to the typical sequence in persuasion studies‐ the early information consisted of message arguments and the later information of heuristic cues (again regarding consensus). All three experiments yielded the same general result: When the later and hence the more diYcult to process information was more subjectively relevant to the judgmental topic than the early information it exerted judgmental (persuasive) impact only under high motivation but not under low motivation. By contrast, the early, less relevant information exerted its eVect only under low motivation but not under high motivation. A very diVerent pattern obtained where the early information was more subjectively relevant than the latter information. Here, the impact of the early information invariably overrode that of the later information: Under low‐processing motivation, this may have been so because the earlier information was easier to process than the later information, and under high‐processing motivation because the early information was in fact more relevant than the later information (see Figs. 9–12 for illustrative results of our first study).
V. Recapitulation and Conclusions Judgmental activity insinuates itself into nearly all manner of people’s response to their social and physical environments, forging an indispensable launching pad for intelligent action. Indeed, psychological research on diverse levels of analysis, from psychophysics to the psychology of culture, has placed considerable emphasis on the study of judgments and their underlying processes. Over the last several decades, this work has resulted in an impressive yield of empirical findings, and a large number of conceptual models, typically adopting a ‘‘dual‐mode’’ perspective. Though some of these models incorporated various notions of continua (Fiske & Neuberg, 1990; Kahneman, 2003; Petty & Cacioppo, 1986),6 they were predominantly 6 Often these continua were anchored in qualitatively diVerent informational contents supporting the concept of a qualitative dichotomy. Thus, Petty and Cacioppo’s (1986) ‘‘elaboration likelihood’’ continuum extends from the brief processing of ‘‘peripheral’’ information to the thorough processing of ‘‘message and issue information.’’ Fiske and Neuberg’s (1990) continuum extends from the brief processing of ‘‘social category’’ information to the extensive (motivated and resource intensive) processing of ‘‘individuating’’ information, and so on.
292
ARIE W. KRUGLANSKI et al.
2.5
2.0
Attitudes
1.5
1.0
0.5
0.0 −0.5 −1.0 Low
High Accuracy motivation
Positive information
Negative information
Fig. 9. Later/more relevant information is more persuasive under high motivation than under low motivation (Pierro et al., 2004, Study 1).
committed to a qualitative dichotomy of judgmental process, rightfully earning their ‘‘dual‐mode’’ designation. Of interest, the binary notions of the various dual‐mode frameworks were quite disparate and did not readily map onto one another. For instance, the ‘‘peripheral’’ processing mode (Petty & Cacioppo, 1986) was depicted in diVerent terms than the ‘‘heuristic’’ mode (Chaiken et al., 1989) in turn characterized quite distinctly than the ‘‘associative’’ mode (Smith & DeCoster, 2000), the ‘‘impulsive’’ mode (Strack & Deutsch, 2004), or the ‘‘categorical’’ mode (Brewer, 1988; Fiske & Neuberg, 1990). Similarly, ‘‘central’’ processing (Petty & Cacioppo, 1986) was portrayed diVerently than ‘‘systematic’’ processing (Chaiken et al., 1989), ‘‘reflective’’ processing (Strack & Deutsch, 2004), or ‘‘rational’’ processing (Kahneman, 2003), and so on. In contrast to such conceptual diversity, the present analysis highlights the common threads shared by the various dual‐mode formulations. These were conceptualized as the psychological dimensions, or continuous parameters, on which judgmental situations may be ordered. The key proposed parameter
ON THE PARAMETERS OF HUMAN JUDGMENT
293
1.8 1.6 1.4
Attitudes
1.2 1.0 0.8 0.6 0.4 0.2 0.0 −0.2
Low
High Accuracy motivation
Positive information
Negative information
Fig. 10. Early/less relevant information is more persuasive under low motivation than under high motivation (Pierro et al., 2004, Study 1).
2.0 1.5
Attitudes
1.0 0.5 0.0 −0.5 −1.0
Low
High Accuracy motivation
Positive information
Negative information
Fig. 11. Early/more relevant information has greater persuasive impact than later/less relevant information under either low or high motivation (Pierro et al., 2004, Study 1).
294
ARIE W. KRUGLANSKI et al.
1.6 1.4
Attitudes
1.2 1.0 0.8 0.6 0.4 0.2 0.0 Low
High Accuracy motivation
Positive information
Negative information
Fig. 12. Later/less relevant information has less persuasive impact than early/more relevant information under either low or high motivation (Pierro et al., 2004, Study 1).
was that of information’s subjective relevance to a judgment. It rests on the assumption that judgments constitute inferences from evidential input, based on implicational rules conditionally linking the two in the knower’s mind. The remaining parameters, namely task demands, and individuals’ cognitive and motivational resources, represent auxiliary factors aVecting the knowers’ ability to appreciate the potential relevance of the information given to a requisite judgment. In brief, we have hypothesized that subjective relevance determines the information’s judgmental impact granting that individuals’ (cognitive and motivational) resources were suYcient for coping with the task demands, hence accurately gleaning the potential relevance of the information given. Evidence from heterogeneous judgmental domains (persuasion, attributions, judgments under uncertainty, or impression formation) yielded support for these assertions. Among others, it was found that where information extraneous to the message or the issue (e.g., about the communicator’s expertise) is presented lengthily and complexly, posing considerable processing demands, it exerts judgmental (persuasive) impact only under high degree of processing motivation and/or high degree of cognitive capacity. However, where the same information is presented briefly and simply it exerts impact under low motivation and/or capacity. Similarly, when message or issue information is presented lengthily and complexly, it exerts judgmental (persuasive) impact under high‐ processing motivation and cognitive capacity. However, where the message or
ON THE PARAMETERS OF HUMAN JUDGMENT
295
issue information is presented briefly and simply, it exerts judgmental (persuasive) impact under low motivation and/or capacity (Erb et al., 2003; Kruglanski & Thompson, 1999a; Pierro et al., 2005).7 We also found that the ability of traits to aVect impressions (the trait centrality eVect; Asch, 1946) depends on their subjective relevance to the specific impressions, in turn, contingent on individuals’ lay theories that tie the traits to the impressions (Fishbach et al., 2005). In a similar vein, prior research (Borgida & Brekke, 1980; Hilton, 1995) has demonstrated the role of subjective relevance in mediating the impact of statistical information (such as base rates) on likelihood estimates, and the possibility of teaching individuals various inferential rules, hence augmenting the subjective relevance of statistical information to specific estimates (Nisbett et al., 1987; Sedlmeier, 1999). Furthermore, Chun and Kruglanski (2006) have shown that the task demands posed by statistical or stereotypic information may vary, and that in both cases demands interact in an identical manner with individuals’ processing potential to determine the impact of (either type of ) information on specific judgments. Finally, research has shown that where the individuals’ processing potential suYces to cope with the task demands, the more relevant information overrides in its judgmental impact the less relevant information; by contrast, where the processing potential is insuYcient to cope with the task demands, the easier to process information (even if of lower relevance) overrides the more diYcult to process information (Pierro et al., 2004). The judgmental parameters identified in the present model aVord the integration of the various concepts and findings featured in prior judgmental models. Consider for example ‘‘intuitions’’ defined by Kahneman (2003) as highly accessible heuristics that come easily to mind. In present terms, such ‘‘intuitions’’ represent easy to process rules imposing low degree of cognitive demands on the individual. It is precisely for that reason that ‘‘intuitions’’ have been known to dominate judgments under conditions of limited cognitive resources (Kahneman, 2003). Similarly, in persuasion research ‘‘peripheral’’ or ‘‘heuristic’’ cues (Petty & Cacioppo, 1986) have been typically operationalized in a manner that may have rendered their processing appreciably less demanding than the processing of message or issue information; indeed, that could be the reason why the former type of information exerted greater impact under limited processing resources, and the latter type of information under 7
Note that our experimental studies pit predictions derived from the present theoretical formulation against those derived from the dual mode persuasion models, aVording them an equal opportunity to be validated. For instance, if message information turned out to be more impactful under high processing resources, and the peripheral information under low processing resources irrespective of their relative length or complexity, this would have validated the dual mode prediction regarding the importance of the type/content of information (i.e., claiming that message information is processed in a qualitatively diVerent manner than peripheral cue information).
296
ARIE W. KRUGLANSKI et al.
ample informational resources (Erb et al., 2003; Kruglanski & Thompson, 1999a,b; Kruglanski et al., 1999; Pierro et al., 2005). Finally, notions of ‘‘associative,’’ ‘‘impulsive,’’ or ‘‘automatic’’ processing are readily interpretable in terms of highly routinized rules (Anderson, 1983; Bargh, 1996; Schneider & ShiVrin, 1977) operating swiftly, eYciently, and often outside awareness, hence imposing low‐processing demands and capable of judgmental impact under conditions of low‐processing potential. The parametric approach of the unimodel aVords two types of integration as far as the dualistic frameworks are concerned: a within‐models integration and a between‐models integration. Rather than treating the two modes within each model as qualitatively distinct, the within‐models integration orders them on one or more dimensional continua. For instance, in much persuasion research ‘‘peripheral cues’’ and ‘‘message or issue arguments’’ have varied both on the parameter of task demands (the cues being easier to process than the arguments) and on the parameter of subjective relevance (the cues being typically perceived as less relevant than the arguments) (Kruglanski & Thompson, 1999a,b; Pierro et al., 2004, 2005). Similarly, ‘‘intuitive’’ heuristics have been defined as more accessible, hence lower on the parameter of task demands than ‘‘rational’’ thoughts (Kahneman, 2003), and the processing of ‘‘social categories’’ was assumed to require less motivation, hence also to be less demanding than the processing of ‘‘individuating information’’ (Brewer, 1988; Fiske & Neuberg, 1990). ‘‘Automatic’’ or ‘‘impulsive’’ processing was defined as more eYcient, hence posing lesser demands on the information processing system, than ‘‘reflective’’ processing (Strack & Deutsch, 2004), and so on. The present analysis suggests that it is such parametric diVerences between the modes (degree of task demands or of subjective relevance of information) rather than other possible distinctions (e.g., in the type or contents of the information processed, awareness, or swiftness of processing) that account for the empirical results on which numerous dual‐mode formulations were based. The evidence reviewed above (see also Chun & Kruglanski, 2006; Chun et al., 2002; Erb et al., 2003; Kruglanski & Thompson, 1999a,b; Pierro et al., 2004, 2005) is consistent with such an analysis. Because the proposed judgmental parameters constitute dimensions common to all judgmental contexts, the present model also aVords an integration between the various dual process models. In other words, the conceptual diversity of the dual‐mode formulations was based on qualitative model‐specific constructs (e.g., ‘‘peripheral cues,’’ ‘‘heuristics,’’ ‘‘automaticity,’’ ‘‘reflection’’). If, as presently suggested, such diversity is eVectively reducible to several continuous parameters, and if such parameters are common to judgmental domains explored by the dual‐mode models, the present formulation oVers a unified perspective on human judgment, aVording a synthesis of a rather fragmented field of psychological inquiry.
ON THE PARAMETERS OF HUMAN JUDGMENT
297
Perhaps more important than the integrative potential of our unified formulation is its focus on the several critical determinants of judgments and their continuous nature. These are conceptualized as orthogonal parameters defining a multidimensional space wherein the plethora of judgmental situations constitute separate points. Such a conception captures the considerable flexibility and malleability of the human judgmental capabilities. Thus, dynamic processes of rule‐learning determine the degree of relevance a given bit of information holds for a given individual. Vicissitudes of situational saliency and rule‐accessibility determine the diYculty of utilizing a given bit of information in a given context. Perceived degree of task importance, of cognitive busyness and the individual’s energy level, determine the degree to which she or he will tend to recruit the needed resources to address a given judgmental problem. Thus, the present framework views the judgmental process in terms of infinitely fine gradations on several intersecting dimensions, whose shifts smoothly transform into one another what initially may appear as qualitatively distinct phenomena. Finally but not of least importance, the central issue dealt with by the present model is of a considerable real‐world relevance. Essentially, it concerns the conditions under which the information given impacts individuals’ judgments, and those in which it does not. This question touches on a variety of intriguing phenomena such as recipients’ reluctance to be convinced by seemingly incontrovertible arguments, or, to be quickly persuaded by obviously specious ones, conflicted parties’ intransigence and inability to reach agreements despite the adversary’s generous concessions, the considerable challenges of intercultural communication, or the abysmal failures ‘‘to see it coming’’ despite the availability of ample, seemingly obvious, information, prompting costly debacles and tragedies of technical or military nature (see Bar‐Joseph & Kruglanski, 2003). A generalized understanding how judgments are formed may oVer insights into these and other topics related to the many complex issues facing individuals and groups in today’s world.
Acknowledgments We are indebted to Judson Mills and Wendy Wood for comments on a previous draft. This work was supported by NSF Grant SBR‐9417422.
References Albarracin, D., Johnson, B. T., & Zanna, M. P. (2005). The handbook of attitudes. Mahwah, NJ: Erlbaum. Anderson, J. R. (1983). The structure of cognition. Cambridge, MA: Harvard University Press.
298
ARIE W. KRUGLANSKI et al.
Anderson, J. R., Bothell, D., Byrne, M. D., Douglass, S., Lebiere, C., & Qin, Y. (2004). An integrated theory of mind. Psychological Review, 111, 1036–1060. Asch, S. E. (1946). Forming impressions of personality. Journal of Abnormal and Social Psychology, 41, 258–290. Baeyens, F., Crombez, G., Van den Bergh, O., & Eelen, P. (1988). Once in contact always in contact: Evaluative conditioning is resistant to extinction. Advances in Behavioural Research and Therapy, 10, 179–199. Baeyens, F., & De Houwer, J. (1995). Evaluative conditioning is a qualitatively distinct form of classical conditioning: A reply to Davey (1994). Behaviour Research and Therapy, 33, 825–831. Bar‐Hillel, M. (1990). Back to base‐rates. In R. M. Hogarth (Ed.), Insights in decision making: A tribute to Hillel J. Einhorn (pp. 200–216). Chicago, IL: University of Chicago Press. Bar‐Joseph, U., & Kruglanski, A. W. (2003). Intelligence failure and need for cognitive closure: On the psychology of the Yom Kippur surprise. Political Psychology, 24, 75–79. Bargh, J. A. (1996). Automaticity in social psychology. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 169–183). New York: Guilford Press. Bartlett, F. C. (1932). Remembering: A study in experimental and social psychology. Cambridge, UK: Cambridge University Press. Baumeister, R. F., Muraven, M., & Tice, D. M. (2000). Ego depletion: A resource model of volition, self‐regulation, and controlled processing. Social Cognition, 18, 130–150. Bhalla, M., & ProYtt, D. R. (1999). Visual‐motor recalibration in geographical slant perception. Journal of Experimental Psychology: Human Perception and Performance, 25, 1076–1096. Bodenhausen, G. V. (1990). Stereotypes as judgmental heuristics: Evidence of circadian variations in discrimination. Psychological Science, 1, 319–322. Borgida, E., & Brekke, N. (1981). The base rate fallacy in attribution and prediction. In J. H. Harvey, W. Ickes, & R. F. Kidd (Eds.), New directions in attribution research (Vol. 3, pp. 63–95). Hillsdale, NJ: Erlbaum. Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull & R. S. Wyer (Eds.), Advances in social cognition (Vol. 1, pp. 1–36). Hillsdale, NJ: Lawrence Erlbaum Associates. Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64, 123–152. Bruner, J. S. (1958). Social psychology and perception. In E. E. Maccoby, T. M. Newcomb, & E. L. Hartley (Eds.), Readings in social psychology (pp. 85–93). New York: Holt, Reinehart, and Winston, Inc. Bruner, J. S. (1973). Beyond the information given: Studies on the psychology of knowing. Oxford, England: W. W. Norton. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116–131. Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information processing within and beyond the persuasion context. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought. New York: The Guilford Press. Chaiken, S., Wood, W., & Eagly, A. H. (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. Chun, W. Y., & Kruglanski, A. W. (2006). The role of task demands and processing resources in the use of base rate and individuating information. Journal of Personality and Social Psychology, 91, 205–217. Chun, W. Y., Spiegel, S., & Kruglanski, A. W. (2002). Assimilative behavior identification can also be resource dependent: A unimodel perspective on personal‐attribution phases. Journal of Personality and Social Psychology, 83, 542–555.
ON THE PARAMETERS OF HUMAN JUDGMENT
299
De Houwer, J., Thomas, S., & Baeyens, F. (2001). Associative learning of likes and dislikes: A review of 25 years of research on human evaluative conditioning. Psychological Bulletin, 127, 853–869. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5–18. Ditto, P. H., & Lopez, D. F. (1992). Motivated skepticism: Use of diVerential decision criteria for preferred and nonpreferred conclusions. Journal of Personality and Social Psychology, 63, 568–584. Dosher, B. A., Liu, S. H., Blair, N., & Lu, Z. L. (2004). The spatial window of the perceptual template and endogenous attention. Vision Research, 44, 1257–1271. Dunning, D. (1999). A newer look: Motivated social cognition and the schematic representation of social concepts. Psychological Inquiry, 10, 1–11. Erb, H.‐P., Kruglanski, A. W., Chun, W. Y., Pierro, A., Mannetti, L., & Spiegel, S. (2003). Searching for commonalities in human judgment: The parametric unimodel and its dual mode alternatives. European Review of Social psychology, 14, 1–47. Fechner, G. T. (1860/1966). Elements of psychophysics. New York: Holt, Reinehart and Winston. Fishbach, A., Kruglanski, A. W., & Chun, W. Y. (2005). Trait centrality: A matter of subjective relevance. Unpublished manuscript. University of Maryland. Fiske, S. T., Lin, M., & Neuberg, S. L. (1999). The continuum model: 10 years later. In S. Chaiken & Y. Trope (Eds.), Dual process theories in social psychology (pp. 231–254). New York: Guilford Press. Fiske, S. T., & Neuberg, S. L. (1990). A continuum model of impression formation, from category‐based to individuating processes: Influences of information and motivation on attention and interpretation. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23, pp. 1–74). New York: Academic Press. Funder, D. C. (1987). Errors and mistakes: Evaluating the accuracy of social judgment. Psychological Bulletin, 101, 75–90. Garcia, J., McGowan, B. K., Ervin, F. R., & Koelling, R. A. (1968). Cues: Their relative eVectiveness as a function of the reinforcer. Science, 160, 794–795. Gawronski, B., & Bodenhausen, G. V. (2006). Associative and propositional processes in evaluation: An integrative review of implicit and explicit attitude change. Psychological Bulletin, 132, 692–731. Ginossar, Z., & Trope, Y. (1980). The eVects of base rates and individuating information on judgments about another person. Journal of Experimental Social Psychology, 16, 228–242. Gregory, R. L. (1997). Eye and brain: The psychology of seeing (5th ed.). Princeton, NJ: Princeton University Press. Helmholtz, J. A. (1910/2000). Treatise on physiological optics. New York: Dover. Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability and salience. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: handbook of basic principles (pp. 133–168). New York: Guilford Press. Hilton, D. J. (1995). Logic and language in causal explanation. In D. Sperber & D. Premack (Eds.), Causal cognition: A multidisciplinary debate. New York: Clarendon Press/Oxford University Press. Hilton, D. J., & Slugoski, B. R. (2001). Conversational processes in reasoning and explanation. In A. Tesser & N. Schwarz (Eds.), Blackwell handbook of social psychology: Intrapersonal processes (Vol. 1, pp. 181–206). Oxford, England: Blackwell. Holland, J. H., Holyoak, K. J., Nisbett, R. E., & Thaggard, P. R. (1986). Induction: processes of inference, learning and discovery. Cambridge, MA: Bradford Books/MIT Press.
300
ARIE W. KRUGLANSKI et al.
Holyoak, K. J., Koh, K., & Nisbett, R. E. (1989). A theory of conditioned reasoning: Inductive learning within rule‐based hierarchies. Psychological Review, 96, 315–340. Jacoby, L. L., Kelley, C., Brown, J., & Jasechko, J. (1989). Becoming famous overnight: Limits on the ability to avoid unconscious influences of the past. Journal of Personality and Social Psychology, 56, 326–338. Jarvis, W. B. G., & Petty, R. E. (1996). The need to evaluate. Journal of Personality and Social Psychology, 70, 172–194. James, W. (1890). The principles of psychology. New York: Henry Colt and Co. Kahneman, D. (2003). A perspective on judgment and choice: Mapping bounded rationality. American Psychologist, 58, 697–720. Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80, 237–251. Kelley, H. H. (1967). Attribution theory in social psychology. Nebraska Symposium on Motivation, 15, 192–238. Kelley, H. H. (1971). Moral evaluation. American Psychologist, 26, 293–300. Kersten, D., Mammassian, P., & Yuille, A. (2004). Object perception as Bayesian inference. Annual Review of Psychology, 55, 271–304. Kopetz, C., & Kruglanski, A. W. (2006). Piecemeal and category‐based evaluation: Playing a single theme in two diVerent keys. Unpublished manuscript. University of Maryland. Krechevsky, I. (1932). ‘‘Hypotheses’’ in rats. Psychological Review, 39, 516–532. Kruglanski, A. W. (1989). Lay epistemics and human knowledge: Cognitive and motivational bases. New York: Plenum Press. Kruglanski, A. W. (1999). Motivation, cognition, and reality: Three memos for the next generation of research. Psychological Inquiry, 10, 54–58. Kruglanski, A. W. (2004). The psychology of closed mindedness. New York: Psychology Press. Kruglanski, A. W., Chen, X., Pierro, A., Mannetti, L., Erb, H. P., & Spiegel, S. (2006). Persuasion according to the unimodel: Implication for cancer communication. Journal of Communication, 56, 105–122. Kruglanski, A. W., Pierro, A., Mannetti, L., & De Grada, E. (2006). Groups as epistemic providers: Need for closure and the unfolding of group centrism. Psychological Review, 113(1), 84–100. Kruglanski, A. W., Raviv, A., Bar‐Tal, D., Raviv, A., Sharvit, K., Ellis, S., et al. (2005). Says who? Epistemic authority eVects in social judgment. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 37, pp. 345–392). New York: Academic Press. Kruglanski, A. W., & Thompson, E. P. (1999a). Persuasion by a single route: A view from the unimodel. Psychological Inquiry, 10, 83–109. Kruglanski, A. W., & Thompson, E. P. (1999b). The illusory second mode or, the cue is the message. Psychological Inquiry, 10, 182–193. Kruglanski, A. W., Thompson, E. P., & Spiegel, S. (1999). Separate or equal? Bimodal notions of persuasion and a single process ‘‘unimodel’’. In S. Chaiken & Y. Trope (Eds.), Dual‐process theories in social psychology (pp. 293–313). New York: Guilford Press. Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: ‘‘Seizing’’ and ‘‘freezing’’. Psychological Review, 103, 263–283. Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480–498. Kunda, Z., & Sanitioso, R. (1989). Motivated changes in the self‐concept. Journal of Experimental Social Psychology, 25, 272–285. Kunda, Z., & Sinclair, L. (1999). Motivated reasoning with stereotypes: Activation, application, and inhibition. Psychological Inquiry, 10, 12–22. Lieberman, M. D., Gaunt, R., Gilbert, T. D., & Trope, Y. (2002). Reflection and reflexion: A social cognitive neuroscience approach to attributional inference. In M. P. Zanna (Ed.),
ON THE PARAMETERS OF HUMAN JUDGMENT
301
Advances in experimental social psychology (Vol. 34, pp. 199–249). New York: Academic Press. Logan, G. D. (1992). Attention and preattention in theories of automaticity. American Journal of Psychology, 105, 317–339. Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The eVects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098–2109. Moors, A., & De Houwer, J. (2006). Automaticity: A theoretical and conceptual analysis,. Psychological Bulletin, 132, 297–326. Murphy, G. L., & Medin, D. L. (1985). The role of theories in conceptual coherence. Psychological Review, 92, 289–316. Neal, D. T., Wood, W., & Quinn, J. M. (2006). Habits: A repeat performance. Current Directions in Psychological Science, 15(4), 198–202. Neuberg, S. L., & Fiske, S. T. (1987). Motivational influences on impression formation: Outcome dependency, accuracy‐driven attention, and individuating processes. Journal of Personality and Social Psychology, 53, 431–444. Newell, A. (1990). Unified theories of cognition. Cambridge, MA: Harvard University Press. Newman, L. S., & Uleman, J. S. (1989). Spontaneous trait inference. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 155–188). New York: Guilford Press. Nisbett, R. E., Fong, G. T., Lehman, D. R., & Cheng, P. W. (1987). Teaching reasoning. Science, 238, 625–631. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231–259. Pavelchak, M. A. (1989). Piecemeal and category‐based evaluation: An idiographic analysis. Journal of Personality and Social Psychology, 56, 354–363. Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 19, pp. 123–205). San Diego, CA: Academic Press. Petty, R. E., Wells, G. T., & Brock, T. C. (1976). Distraction can enhance or reduce yielding to propaganda: Thought disruption versus eVort justification. Journal of Personality and Social Psychology, 34, 874–884. Pierro, A., Mannetti, L., Erb, H. P., Spiegel, S., & Kruglanski, A. W. (2005). Informational length and order of presentation as determinants of persuasion. Journal of Experimental Social Psychology, 41, 458–469. Pierro, A., Mannetti, L., & Kruglanski, A. W. (2006). Length of source information as determinant of persuasion: Attitudes stability, behavioral intentions and behavior. Unpublished manuscript. University of Rome La Sapienza. Pierro, A., Mannetti, L., Kruglanski, A. W., & Sleeth‐Keppler, D. (2004). Relevance override: On the reduced impact of ‘‘cues’’ under high motivation conditions of persuasion studies. Journal of Personality and Social Psychology, 86, 251–264. Pizlo, Z. (2001). Perception viewed as an inverse problem. Vision Research, 41, 3145–3161. Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology, 32, 3–25. ProYtt, D. R., Bhalla, M., Gossweiler, R., & Midgett, J. (1995). Perceiving geographical slant. Psychonomic Bulletin and Review, 2, 409–428. ProYtt, D. R., Stefanucci, J., Banton, T., & Epstein, W. (2003). The role of eVort in perceived distance. Psychological Science, 14, 106–112. Rescorla, R. A. (1968). Probability of shock in the presence and absence of CS in fear conditioning. Journal of Comparative and Physiological Psychology, 66, 1–5.
302
ARIE W. KRUGLANSKI et al.
Rescorla, R. A. (1985). Conditioned inhibition and facilitation. In R. R. Miller & N. E. Spear (Eds.), Information processing as animals: Conditioned inhibition (pp. 299–326). Hillsdale, NJ: Erlbaum. Rescorla, R. A., & Holland, P. C. (1982). Behavioral studies of associative learning in animals. Annual Review of Psychology, 33, 265–308. Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the eVectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Procasy (Eds.), Classical Conditioning II: Current Theory and Research (pp. 64–99). New York: Appleton Century Crofts. Rock, I. (1983). The logic of perception. Cambridge, MA: The MIT Press. Santhi, N., & Reeves, A. (2004). The roles of distractor noise and target certainty in search: A signal detection model. Vision Research, 44, 1235–1256. Schneider, W., & ShiVrin, R. M. (1977). Controlled and automatic human information processing: I. Detection, search, and attention. Psychological Review, 84, 1–66. Schwarz, N. (2004). Metacognitive experiences in consumer judgment and decision making. Journal of Consumer Psychology, 14(4), 332–348. Schwarz, N., & Clore, G. L. (1983). Mood, misattribution and judgement of well‐being. Informative and directive functions of aVective states. Journal of Personality and Social Psychology, 45, 513–523. Schwarz, N., & Clore, G. L. (1996). Feelings and phenomenal experiences. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 433–465). New York: Guilford Press. Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer‐Schatka, H., & Simons, A. (1991). Ease of retrieval as information: Another look at the availability heuristic. Journal of Personality and Social Psychology, 61, 195–202. Schwarz, N., Servay, W., & Kumpf, M. (1985). Attribution of arousal as a mediator of the eVectiveness of fear‐arousing communications. Journal of Applied Social Psychology, 15, 178–188. Schwarz, N., Strack, F., Hilton, D. J., & Naderer, G. (1991). Base rates, representativeness, and the logic of conversation: The contextual relevance of ‘‘irrelevant’’ information. Social Cognition, 9, 67–84. Sedlmeier, P. (1999). Improving statistical reasoning: Theoretical models and practical implications. Mahwah, NJ: Lawrence Erlbaum Associates. Smith, E. R., & Branscombe, N. R. (1988). Category accessibility as implicit memory. Journal of Experimental Social Psychology, 24, 490–504. Smith, E. R., Branscombe, N. R., & Bormann, C. (1988). Generality of the eVects of practice on social judgment tasks. Journal of Personality and Social Psychology, 54, 385–395. Smith, E. R., & DeCoster, J. (2000). Dual process models in social and cognitive psychology: Conceptual integration and links to underlying memory systems. Personality and Social Psychology Review, 4, 108–131. Steele, C. M., & Josephs, R. A. (1990). Alcohol myopia: Its prized and dangerous eVects. American Psychologist, 45, 921–933. Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of social behavior. Personality and Social Psychology Review, 8, 220–247. Tetlock, P. E. (1985). Accountability: A social check on the fundamental attribution error. Social Psychology Quarterly, 48, 227–236. Tolman, E. C. (1932). Purposive behavior in animals and men. New York: Appleton Century‐ Crofts. Trope, Y. (1986). Identification and inferential processes in dispositional attribution. Psychological Review, 93, 239–257.
ON THE PARAMETERS OF HUMAN JUDGMENT
303
Trope, Y., & Alfieri, T. (1997). EVortfulness and flexibility of dispositional judgment processes. Journal of Personality and Social Psychology, 73, 662–674. Trope, Y., & Gaunt, R. (2000). Processing alternative explanations of behavior: Correction of integration? Journal of Personality and Social Psychology, 79, 344–354. Trope, Y., & Ginossar, Z. (1988). On the use of statistical and nonstatistical knowledge: A problem solving approach. In D. Bar‐Tel & A. W. Kruglanski (Eds.), The social psychology of knowledge (pp. 209–230). Cambridge, MA: Cambridge University Press. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124–1130. Uleman, J. S. (1987). Consciousness and control: The case of spontaneous trait inferences. Personality and Social Psychology Bulletin, 13, 337–354. Vallacher, R. R., & Wegner, D. M. (1985). A theory of action identification. Hillsdale, NJ: Erlbaum. Walther, E., Nagengast, B., & Trasselli, C. (2005). Evaluative conditioning in social psychology: Facts and speculations. Cognition and Emotion, 19, 175–196. Wason, P. C., & Johnson‐Laird, P. N. (1972). Psychology of Reasoning: Structure and Content, Oxford, England, Harvard University Press. Webster, D. M., & Kruglanski, A. W. (1998). Cognitive and social consequences of the need for cognitive closure. In W. Stroebe & M. Hewstone (Eds.), European review of social psychology (Vol. 8, pp. 133–141). Chichester, UK: Wiley. Webster, D. M., Richter, L., & Kruglanski, A. W. (1996). On leaping to conclusions when feeling tired: Mental fatigue eVects on impression formation. Journal of Experimental Social Psychology, 32, 181–195. Wilson, T. D., Centerbar, D. B., Kermer, D. A., & Gilbert, D. T. (2005). The pleasures of uncertainty: Prolonging positive moods in ways people do not anticipate. Journal of Personality and Social Psychology, 88, 5–21. Wilson, T. D., Dunn, D. S., Kraft, D., & Lisle, D. J. (1989). Introspection, attitude change, and attitude‐behavioral consistency: The disruptive eVects of explaining why we feel the way we do. Advances in Experimental Social Psychology, 22, 347–343. Wishner, J. (1960). Reanalysis of ‘‘impressions of personality’’. Psychological Review, 67, 96–112. Witt, J. K., ProYtt, D. R., & Epstein, W. (2004). Perceiving distance: A role of eVort and intent. Perception, 33, 577–590. Wood, W., Kallgren, C. A., & Preisler, R. M. (1985). Access to attitude‐relevant information in memory as a determinant of persuasion: The role of message attributes. Journal of Experimental Social Psychology, 21, 73–85. Wright, B. A., & Fitzgerald, M. B. (2003). Sound discrimination learning and auditory displays. In Proceedings of the International Conference on Auditory Display. ICAD‐2003, Boston, MA, pp. 228–232. Zanna, M. P., & Hamilton, D. L. (1972). Attribute dimensions and patterns of trait inferences. Psychonomic Science, 27, 353–354. Zukier, H., & Pepitone, A. (1984). Social roles and strategies in prediction: Some determinants of the use of base‐rate information. Journal of Personality and Social Psychology, 47, 349–360.
PANGLOSSIAN IDEOLOGY IN THE SERVICE OF SYSTEM JUSTIFICATION: HOW COMPLEMENTARY STEREOTYPES HELP US TO RATIONALIZE INEQUALITY Aaron C. Kay John T. Jost Anesu N. Mandisodza Steven J. Sherman John V. Petrocelli Amy L. Johnson
According to system justification theory, there is a general social psychological tendency to rationalize the status quo, that is, to see it as good, fair, legitimate, and desirable. This tendency is reminiscent of the dispositional outlook of Voltaire’s famous character, Dr. Pangloss, who believed that he was ‘‘living in the best of all possible worlds.’’ One of the means by which people idealize existing social arrangements is by relying on complementary (or compensatory) stereotypes, which ascribe compensating virtues to the disadvantaged and corresponding vices to the advantaged, thereby creating an ‘‘illusion of equality.’’ In this chapter, we summarize a program of research demonstrating that (1) incidental exposure to complementary gender and status stereotypes leads people to show enhanced ideological support for the status quo and (2) when the legitimacy or stability of the system is threatened, people often respond by using complementary stereotypes to bolster the system. We also show that (noncomplementary) victim‐blaming and (complementary) victim‐enhancement represent alternate routes to system justification. In addition, we consider a number of situational and dispositional moderating variables that aVect the use and eVectiveness of 305 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39006-5
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
306
AARON C. KAY et al.
complementary and noncomplementary representations, and we discuss the broader implications of stereotyping and other forms of rationalization that are adopted in the service of system justification. From time to time, Pangloss would say to Candide: There is a chain of events in this best of all possible worlds; for if you had not been turned out of a beautiful mansion at the point of a jackboot for love of Lady Cune´gonde, if you had not been clamped into the Inquisition, if you had not wandered about America on foot, and had not struck the Baron with your sword, and lost all those sheep you brought from Eldorado, you would not be here eating candied fruit and pistachio nuts. ‘‘That’s true enough,’’ said Candide; ‘‘but we must go and work in the garden.’’ —Voltaire, 1758/1947, Candide or Optimism, p. 144
I. Introduction Whether because of discrimination on the basis of race, ethnicity, religion, social class, gender, or sexual orientation, or because of policies and programs that privilege some at the expense of others, or even because of historical accidents, genetic disparities, or the fickleness of fate, certain social systems serve the interests of some stakeholders better than others. Yet historical and social scientific evidence shows that most of the time the majority of people—regardless of their own social class or position—accept and even defend the legitimacy of their social and economic systems and manage to maintain a ‘‘belief in a just world’’ (Jost & Banaji, 1994; Jost & Hunyady, 2002, 2005; Jost & Major, 2001; Lane, 1962; Lerner, 1980; Major, 1994; Moore, 1978; Sidanius & Pratto, 1999; Zinn, 1968). As Kinder and Sears (1985) put it, ‘‘the deepest puzzle here is not occasional protest but pervasive tranquility.’’ Knowing how easy it is for people to adapt to and rationalize the way things are makes it easier to understand why the apartheid system in South Africa lasted for 46 years, the institution of slavery survived for more than 400 years in Europe and the Americas, and the Indian Caste system has been maintained for 3000 years and counting. This point was made vividly by Elkins (1967), who drew parallels between the psychological situations faced by African‐American slaves and concentration camp survivors during the Nazi regime. He identified a number of bivalent stereotypes that seemed to help people to rationalize institutionalized abuse and exploitation under slavery and similar systems. For example, the stereotype of ‘‘‘Sambo’ in Southern lore was docile but irresponsible,
SYSTEM JUSTIFICATION THEORY
307
loyal but lazy, humble but addicted to lying and stealing . . .. His relationship to his master was one of utter dependence and childlike attachment; it was this childlike quality that was the very key to his being’’ (p. 395). Elkins also noted that for the 700,000 survivors of the Nazi holocaust, ‘‘the regime must be considered not as a system of death but as a way of life. These survivors did make an adjustment of some sort to the system; it is they themselves who report it.’’ He concluded that with respect to slaves and concentration camp survivors, ‘‘It is no wonder that their obedience became unquestioning, that they did not revolt, that they could not ‘hate’ their masters. Their masters’ attitudes had become internalized as a part of their very selves’’ (p. 410). Even in extraordinarily oppressive circumstances such as these, people find ways of adapting to circumstances that they cannot change, so that ‘‘the unwelcome force is idealized’’ (Dollard, 1937, p. 255; see also Jackman, 1994; Jost, 2001, pp. 91–92). These facts about the remarkable human capacity to rationalize existing social arrangements, however unjust those arrangements may be, were well‐ known to earlier generations of social psychologists such as Dollard (1937), Bettelheim (1943), Lewin (1941/1948), Allport (1954), and Elkins (1967), but they are frequently forgotten in the context of contemporary theories that stress self‐enhancement, self‐aYrmation, social identification, and in‐group favoritism as ubiquitous, almost ineradicable, motives. Until recently, there was very little research on the social and psychological processes by which people maintain relatively favorable views of the social systems that aVect them. How do people rationalize their own outcomes—whether good or bad—as well as the outcomes of others and, above all, the social systems that dictate those outcomes? What are the cognitive, motivational, and interpersonal mechanisms that enable people to cope with the intrapsychic conflict associated with participating in social systems that are, in many objective ways, arbitrary, capricious, and perhaps even unfair? System Justification Theory (SJT) tackles these and related questions by addressing the antecedents, contents, and consequences of thoughts, feelings, and behaviors that serve to maintain the societal status quo (Blasi & Jost, 2006; Jost & Banaji, 1994; Jost, Banaji, & Nosek, 2004; Jost & Hunyady, 2002, 2005; Jost & Kay, 2005; Kay & Jost, 2003; Kay, Jost, & Young, 2005; Mandisodza, Jost, & Unzueta, 2006). The theory was formulated more than a decade ago to explain a particularly vexing, but consistent, social psychological finding: the prevalence of out‐group favoritism among low‐status group members (Jost & Banaji, 1994). Since then, however, research on SJT has expanded greatly to include the empirical search for any and all social psychological processes that serve to maintain or bolster support for the social system (Jost et al., 2004). The research that we summarize in this chapter focuses on one psychological means of resolving the tension caused
308
AARON C. KAY et al.
by inequality, namely the formation and use of complementary stereotypes, that is, stereotypes that appear to compensate for intergroup disparities by assigning oVsetting advantages and disadvantages to low‐ and high‐status groups, respectively, thereby preserving a more just (i.e., equal) image of society than may be accurate. Drawing on both published and (to a lesser extent) unpublished research conducted over the past few years, our guiding thesis is that complementary stereotypes serve to rationalize inequality, allowing people to maintain their belief that the societal status quo is, generally speaking, fair, legitimate, and justified.
II. System Justification Theory A. MOTIVATION TO RATIONALIZE THE STATUS QUO At the most basic level, SJT posits that there is a general psychological tendency to justify and rationalize the status quo; in other words, there is a motivation to see the system as good, fair, legitimate, and desirable (Jost et al., 2004). System justification has negative consequences for some individuals. For example, members of disadvantaged groups exhibit lower personal and collective self‐esteem (Jost & Thompson, 2000; O’Brien & Major, 2005). Despite this, there are a number of psychological reasons why people would actively seek to justify the status quo. These reasons include (but are not limited to) cognitive‐motivational needs to believe in order, structure, closure, stability, predictability, consistency, and control (Allport, 1966; Crandall & Beasley, 2001; Festinger, 1957; JanoV‐Bulman & Yopyk, 2004; Kruglanski, 2004; Langer, 1975; Plaks, Grant, & Dweck, 2005) and to believe in a just world (Hafer & Begue, 2005; Lerner, 1980). Experimental studies in which people exhibit increased system justification following system threat, as we show in this chapter, provide further evidence that there is a general motive to defend and bolster the status quo (Jost & Hunyady, 2002; Jost, Kivetz, Rubini, Guermandi, & Mosso, 2005; Kay et al., 2005). There are social reasons as well not to ‘‘upset the apple cart’’: people, especially those who engage in system justification, derogate others who are perceived as complaining about discrimination and injustice (Kaiser, Dyrenforth, & Hagiwara, 2006). Thus, it appears that there are social norms that serve to uphold system‐justifying responses and punish system‐ challenging responses. Furthermore, women and others who defy stereotypes and otherwise threaten the status quo face the persistent prospect of backlash (Rudman & Fairchild, 2004; Rudman & Glick, 1999).
SYSTEM JUSTIFICATION THEORY
309
B. OUT‐GROUP FAVORITISM AND THE JUSTIFICATION OF ARBITRARY INEQUALITIES AMONG GROUPS One manifestation of the system justification motive is the pervasive tendency to use social judgments and stereotypes to justify arbitrary status and power diVerences between groups—that is, to portray both high‐ and low‐status groups as deserving of their position in the hierarchy (Jost, 2001; Jost & Burgess, 2000). For example, group members who are arbitrarily (and even illegitimately) ordained with high levels of relative power in an experimental setting tend to be perceived as more intelligent and responsible than group members who are arbitrarily placed into positions of low power; these perceptions are rendered by the powerful as well as the powerless (Haines & Jost, 2000). Furthermore, people tend to remember the bases of the power diVerences as more legitimate (and less arbitrary) than they actually were. There are also a great many studies showing that members of high‐status groups are both consciously and unconsciously preferred to members of low‐ status groups, even by members of low‐status groups (for reviews of this extensive literature, see Dasgupta, 2004; Jost et al., 2004). Although much of this work has employed the Implicit Association Test (IAT; Ashburn‐ Nardo, Knowles, & Monteith, 2003; Rudman, Feinberg, & Fairchild, 2004; Uhlmann, Dasgupta, Elgueta, Greenwald, & Swanson, 2002), out‐ group favoritism on the part of the disadvantaged has also been observed on other cognitive, aVective, and behavioral measures of implicit preference (Jost, Pelham, & Carvallo, 2002). From a system justification perspective, the internalization of inequality is an important instance of ‘‘buying into’’ the status quo (Jost et al., 2004).
C. ANTICIPATORY RATIONALIZATION OF LIKELY OUTCOMES Some further evidence for the existence of a motivation to justify the social system has emerged from research examining a diVerent form of system justification, namely the rationalization of the anticipated status quo (Kay, Jimenez, & Jost, 2002). Inspired in part by the writings of William J. McGuire (McGuire, 1960; McGuire & McGuire, 1991; see also Sherman, 1991), we hypothesized that to the extent that people are motivated to justify the status quo (whatever characteristics it might have), then they should begin to see highly probable events in increasingly favorable terms and highly improbable events in increasingly unfavorable terms (see also Elster, 1983; Jolls & Sunstein, 2006). Let us consider the example of the anticipated
310
AARON C. KAY et al.
outcome of a US presidential election, which, for many American residents, holds fairly obvious implications for the overarching social system. A system justification perspective suggests that as it becomes more (versus less) likely that a given candidate is going to win an election, engaged citizens should adjust their evaluations of that candidate accordingly. Furthermore, these desirability adjustments, according to Kay et al. (2002) and the general tenets of SJT, should take two forms: (1) a ‘‘sour grapes’’ rationalization, in which initially preferred outcomes that are deemed to be less and less likely should come to be rationalized as less desirable, and (2) a ‘‘sweet lemon’’ rationalization, in which initially nonpreferred outcomes that are deemed to be more and more likely should also come to be rationalized as more desirable. Two experimental studies provided evidence in support of both forms of anticipatory rationalization of the status quo (Kay et al., 2002; see also Gilbert, Pinel, Wilson, Blumberg, & Wheatley, 1998). In one study, which was conducted in the context of the 2000 presidential election between Al Gore and George W. Bush, participants were assigned to one of five conditions that manipulated likelihood beliefs concerning the chances of both Gore and Bush presidencies. Afterward, personal desirability ratings for both Gore and Bush presidencies were assessed. As predicted, desirability ratings of each president were influenced by the likelihood manipulations. When participants were led to believe that a Gore presidency was more (rather than less) likely, Democrats viewed him as even more desirable and Republicans viewed him as less undesirable (see Fig. 1A). Similarly, when participants were led to believe that a Bush presidency loomed, Republicans viewed him as more desirable and Democrats saw him as less undesirable (see Fig. 1B). Interestingly, nonpartisans, who were apparently less motivationally involved in the outcome of the election, showed no eVect of the experimental manipulations. These findings were conceptually replicated in a follow‐up experiment in which we manipulated the perceived likelihood of a large (i.e., high‐ motivational involvement) or small (i.e., low‐motivational involvement) tuition increase or decrease at Stanford University and then measured the desirability of these anticipated outcomes among Stanford students. Results from this study were very similar to those obtained in the election study. As large tuition increases were seen as more likely to occur, Stanford students adapted to the unwelcome news; they judged the tuition increase to be less undesirable. Conversely, as large tuition decreases were seen as less likely, Stanford students judged them to be less personally desirable. Participants assigned to the low‐motivational involvement conditions (i.e., those participants who were told that the tuition change would be small) were unaVected by the likelihood manipulations. These findings (and many others
311
SYSTEM JUSTIFICATION THEORY
A
Republicans
Democrats
Nonpartisans
Desirability of Gore presidency
9 8 7 6 5 4 3 2 1 43/51
B
45/49 47/47 49/45 Perceived likelihood of Gore victory
Republicans
53/41
Nonpartisans
Democrats
Desirability of Bush presidency
9 8 7 6 5 4 3 2 1 43/51
45/49
47/47
49/45
53/41
Perceived likelihood of Bush victory Fig. 1. (A) Desirability ratings of a Gore presidency as a function of manipulated likelihood. (Adapted from Kay et al., 2002.) (B) Desirability ratings of a Bush presidency as a function of manipulated likelihood. (Adapted from Kay et al., 2002.)
like them) are reminiscent of the mindset made famous by Voltaire’s (1758/ 1947) character, Dr. Pangloss, who repeatedly insisted that we are living ‘‘in the best of all possible worlds’’ (see also Blasi & Jost, 2006).
312
AARON C. KAY et al.
III. The System‐Justifying Function of Complementary Stereotypes SJT addresses a broad range of thoughts, feelings, and behaviors. A primary focus of research, however, has been on stereotyping and the guiding notion that stereotypes can serve to justify and rationalize inequalities within the social system (Jost, 2001; Jost & Hamilton, 2005; Jost et al., 2005). While much early work on stereotyping and prejudice stressed the aVective and motivational underpinnings of intergroup attitudes (Allport, 1954), contemporary approaches to the study of social stereotypes have tended to emphasize cognitive processes such as categorization (rather than, say, rationalization). Researchers in the 1980s and 1990s, for example, focused on stereotype accessibility, activation, and applicability, as well as issues of automaticity, control, and awareness (Devine, 1989; Fiske, 1998; Hamilton & Sherman, 1994. Macrae & Bodenhausen, 2000). During this period of heightened attention to relatively low‐level cognitive issues, issues of stereotype content, function, and origin were placed on the back burner. Despite this overall trend, a handful of researchers have addressed motivational processes underlying stereotyping and the social psychological functions that aVect the contents of stereotypes. For example, work on the stereotype content model (Fiske, Cuddy, Glick, & Xu, 2002) and on social dominance (Sidanius & Pratto, 1999), as well as models of ego justification and group justification (Fein & Spencer, 1997; Hogg & Abrams, 1988; Tajfel, 1981), contributed to an appreciation of the functional basis of stereotype content. From a system justification perspective, the content of stereotypes can serve to maintain ideological support for the prevailing social system by justifying and rationalizing inequality (see also Glick & Fiske, 2001; HoVman & Hurst, 1990; Jackman & Senter, 1983; Jackman, 1994; Jost, 2001; Jost & Banaji, 1994; Major, 1994; Sidanius & Pratto, 1999). The conflict between inequality, on one hand, and the desire to believe in a fair and just social world (Lerner, 1980), on the other hand, is likely to evoke a significant degree of ideological dissonance. Stereotypes, and in particular complementary stereotypes, we think, provide a common and eVective means of addressing and coping with this conflict. Although SJT emphasized the rationalization function of stereotyping from the start (Jost & Banaji, 1994), it has only very recently identified the role of complementary stereotyping in the rationalization process (Jost & Kay, 2005; Kay & Jost, 2003; Kay et al., 2005; Napier, Mandisodza, Andersen, & Jost, 2006). Status‐congruent stereotypes in which low‐status and high‐status groups are both seen as ‘‘deserving’’ of their relative positions received the bulk of the attention in the first wave of system
SYSTEM JUSTIFICATION THEORY
313
justification research (Jost, 2001; Jost, Burgess, & Mosso, 2001). Although derogating society’s ‘‘losers’’ and lionizing its ‘‘winners’’ is one powerful means of justifying the system, it is not the only means possible. Theorists have also noted the potentially system‐justifying consequences of complementary stereotypes, that is, stereotypes that serve to elevate low‐status groups and derogate high‐status groups (Jost et al., 2001; see also Glick & Fiske, 2001; Hunyady, 1998; Jackman, 1994; Lane, 1962). Such stereotypes often depict low‐ and high‐status groups as possessing their own unique strengths and weaknesses (or benefits and burdens). Indeed, several converging lines of research suggest that high‐ and low‐status individuals and groups are frequently assumed to have diVerent but complementary, or ‘‘balancing,’’ characteristics. We have shown that complementary stereotypical representations—those that depict relatively low‐status groups as having their own set of compensating rewards and high‐status groups as having certain drawbacks— contribute to the perceived legitimacy of the social system (Jost & Kay, 2005; Kay & Jost, 2003; Kay et al., 2005). These representations communicate that ‘‘no one group has it all’’ and thus encourage the feeling that things somehow balance out in a way that makes the system seem fair, or at least not unbearably unfair. That is, in lay thinking, a just social order is one in which no single group enjoys a monopoly over valued attributes and every group has something going for it. Thus, if equality cannot be achieved in actuality, complementary stereotypes, may help us to create a comforting illusion of equality.1 In this way, we are closer to living ‘‘in the best of all possible worlds.’’
A. THE CASE OF COMPLEMENTARY GENDER STEREOTYPES Perhaps the most familiar example of complementary stereotyping pertains to beliefs about gender diVerences. Researchers have often observed that women are stereotyped as communal but not agentic, whereas men are stereotyped as agentic but not communal. For example, researchers have noted that women in general are perceived as nicer, warmer, more supportive, and more
1 The hypothesis that stereotypes that reflect a more ‘‘balanced’’ social world will serve to justify the social system is consistent with interpersonal theories of justice, especially equity theory (Walster, Berscheid, & Walster, 1978). It also bears some relationship to Heider’s (1958; see also Crandall & Beasley, 2001) balance theory, although we are conceiving of balance in terms of a social equilibrium (in which benefits and burdens across groups ‘‘balance out’’), whereas Heider largely conceived of balance in terms of consistency or congruency.
314
AARON C. KAY et al.
interpersonally sensitive than men (Eagly & Mladinic, 1989). This has been characterized as the ‘‘women are wonderful’’ eVect (Eagly, Mladinic, & Otto, 1991). Additional studies carried out under the rubric of ‘‘benevolent sexism’’ have also found that women are often stereotyped as more moral, culturally refined, and deserving of protection than men (Glick & Fiske, 2001). It has been suggested that bivalent stereotypes of women as incompetent but also warm, friendly, caring, nurturing, honest, and morally superior to men serve to rationalize the patriarchal system (Jackman, 1994; Jost & Banaji, 1994). In support of this notion, Glick and Fiske (2001) demonstrated that such gender stereotypes are indeed widespread, that they are endorsed by women as well as men, and that they are especially prevalent in societies with extreme gender inequalities (as measured by objective indicators pertaining to the social and economic advancement of women). There are three main arguments that have been oVered for why complementary gender stereotypes might increase support for the system among women. First, stereotypic diVerentiation along communal/agentic lines can encourage people to treat each gender group as essentially well‐suited to occupy the positions and roles that are prescribed for them by society (see also Jost & Hamilton, 2005). Several studies have shown that people will spontaneously stereotype groups in ways that render them especially well‐ suited to fulfill their social roles. For example, people ascribe communal, nurturant characteristics to people (usually women) who are assumed to occupy stereotypically feminine roles such as the role of childcare‐giver (Eagly & SteVen, 1984; HoVman & Hurst, 1990). Second, according to Jackman (1994), ascribing communal characteristics to women prevents them from withdrawing completely from the system of gender relations by ‘‘sugarcoating’’ their low‐status position, leading them to more easily tolerate their status and position (see also Rudman, 2005). In other words, praising women for their communal qualities may flatter them into active cooperation with the patriarchal system. The third possibility, which is the one that we have focused on in our research, is that complementary gender stereotypes such as these maintain the sense that the system as a whole is fair, balanced, and legitimate (Jost & Kay, 2005). That is, complementary stereotypes of women may justify the current system of gender relations (including the division of labor in society and in the family), and the status quo in general, by reinforcing the notion that each gender group possesses its own set of strengths and weaknesses that supplement and balance out the strengths and weaknesses of the other group. We hasten to add that these explanations are not mutually exclusive. On the contrary, we think that these processes all work together as rationalizations for ongoing gender inequality in society.
SYSTEM JUSTIFICATION THEORY
315
1. EVects of Exposure to Complementary Stereotypes on Gender‐Specific System Justification Although the system‐justifying eVects of exposure to complementary stereotypes had never been tested experimentally, this general argument had been made persuasively by Bem and Bem (1970) in an influential chapter entitled ‘‘Case Study of a Nonconscious Ideology: Training the Woman to Know Her Place.’’ In this chapter, Bem and Bem observed that: In 1954 the United States Supreme Court declared that a fraud and hoax lay behind the slogan ‘‘separate but equal.’’ It is unlikely that any court will ever do the same for the more subtle motto that successfully keeps the woman in her place: ‘‘complementary but equal’’ . . .. The ideological rationalization that men and women hold complementary but equal positions in society appears to be a fairly recent invention. In earlier times—and in more conservative company today—it was not felt necessary to provide the ideology with an equalitarian veneer (p. 96).
Complementary stereotypes of women as more nurturing and caring than men, in other words, create an ‘‘illusion of equality’’ that helps men, and especially women, to justify ongoing gender disparities. This hypothesis—that common social and cultural representations of men and women lend ideological support to the status quo—may not at first appear to lend itself readily to psychological experimentation, but we concluded that an experimental test should in fact be possible. Social cognitive research on priming and implicit stereotyping suggests that subtle, unconscious reminders of stereotypical features (such as those communicated by character vignettes, certain visual images of members of a given social category, and even ‘‘ostensibly unrelated’’ descriptions of the stereotypical attributes of a given category exemplar) are adequate to increase the cognitive accessibility of stereotypes (Bargh, Chen, & Burrows, 1996; Higgins, 1996). In addition, research on stereotype activation suggests that once the cognitive accessibility of a social stereotype is increased, the perceptions and motivations that are generally associated with that stereotype tend to follow psychologically (Bargh, 1996; Dijksterhuis & Bargh, 2001). Thus, if certain stereotypes do serve to justify the social system, then exposing people to subtle reminders of those stereotypes should lead them to subsequently endorse the perceived legitimacy of the social system to a greater extent (see also Jost & Major, 2001). This prediction is also consistent with Bargh, Gollwitzer, Lee‐Chai, Barndollar, and Tro¨tschel (2001) automotive model of automatically triggered motivation, which suggests that when a repeated situation has led an individual to consciously adopt a specific goal often enough (so that the situation‐behavior link becomes adequately ‘‘tight’’), the situation itself will begin to automatically activate the goal. The idea is that a
316
AARON C. KAY et al.
consciously motivated process will eventually be replaced by an implicitly activated one. Thus, if repeated exposure to complementary stereotypes does serve to justify gender inequality in society, then simply increasing the cognitive accessibility of such stereotypes should automatically trigger an increase in the perceived fairness and legitimacy of the status quo. This prediction is also in line with other work demonstrating the eVects of temporarily activated ideologies on levels of prejudice and tolerance (Katz & Hass, 1988; Quinn & Crocker, 1999). Drawing on this line of reasoning, we conducted a series of experimental tests designed to examine the eVects of complementary stereotypes on beliefs concerning the fairness and legitimacy of the social system as a whole. This methodology was used in the specific context of examining the eVects of complementary gender stereotypes on subsequent system justification. In one study (Jost & Kay, 2005, Study 1), 100 participants were randomly assigned to 1 of 4 diVerent stereotype exposure conditions. In one condition, participants were asked to indicate whether five communal traits (considerate, honest, happy, warm, and moral) applied more to women or to men and to what degree. In another condition, participants were asked to indicate whether each of five agentic traits (assertive, competent, intelligent, ambitious, and responsible) applied more to women or to men. In a third condition, participants were asked to judge both communal and agentic traits. In all 3 experimental conditions, responses were given on a 10‐point scale ranging from 0 (e.g., ‘‘women are more considerate than men’’) to 9 (e.g., ‘‘men are more considerate than women’’).2 A control condition, in which participants were not exposed to gender stereotypes of any kind during the first part of the experiment, was also included in the design. Then, in an ostensibly unrelated second study, a scale measuring gender‐ specific system justification was administered. This questionnaire contained eight opinion statements regarding the current state of gender roles and sex role division. These items are listed in Table I. Responses were given on nine‐point scales, such that higher numbers indicated stronger agreement. An overall index of gender‐specific system justification was calculated by taking the mean of the eight items following reverse‐coding of two items. We predicted that participants who had been exposed to the communal stereotypes would evince higher levels of system justification than participants assigned to other conditions. 2
Although we used this methodology to activate the specific gender stereotypes rather than as a means of assessing the extent to which our participants themselves endorsed these stereotypes, responses to these questions were indeed recorded. The means did reflect the default stereotypes: Agentic traits were rated as significantly more characteristic of men, and communal traits were rated as significantly more characteristic of women.
SYSTEM JUSTIFICATION THEORY
ITEMS
FOR
317
TABLE I MEASURING GENDER‐SPECIFIC SYSTEM JUSTIFICATION
In general, relations between men and women are fair The division of labor in families generally operates as it should Gender roles need to be radically restructured (R) For women, the United States is the best country in the world to live in Most policies relating to gender and the sexual division of labor serve the greater good Everyone (male or female) has a fair shot at wealth and happiness Sexism in society is getting worse every year (R) Society is set up so that men and women usually get what they deserve Note: Participants indicated their agreement or disagreement on a scale ranging from one (strongly disagree) to nine (strongly agree). Items followed by ‘‘(R)’’ were reverse‐scored prior to data coding and analyses (Jost & Kay, 2005).
Gender-specific system justification
Women
Men
5.50 5.00 4.50 4.00 3.50 3.00 No
Yes
Exposure to communal stereotypes Fig. 2. Mean scores on gender‐specific system justification as a function of participant gender and exposure to communal stereotypes. (Adapted from Study 1 in Jost & Kay, 2005.)
The data were generally supportive of this hypothesis. As can be seen in Fig. 2, activating communal gender stereotypes served to increase women’s degree of support for the existing system of gender relations. For men, gender‐specific system justification was uniformly high, and it was unaVected by stereotype activation. This experiment, we believe, was the first to provide direct evidence of a causal connection between exposure to specific gender stereotype contents and ideological support for the system of gender relations. Interestingly, the extent to which participants personally endorsed
318
AARON C. KAY et al.
communal stereotypes did not significantly correlate with levels of system justification; nearly everyone subscribed to the same stereotypical beliefs. Thus, this study suggested that exposure to complementary gender stereotypes, regardless of the degree of conscious personal endorsement of those stereotypes, was suYcient to lead women to show increased support for the current system of gender relations in society. 2. EVects of Exposure to Complementary Stereotypes on DiVuse System Justification A follow‐up experiment sought to improve on and expand on these results in several important ways (Jost & Kay, 2005, Study 2). First, we examined the extent to which exposure to hostile and benevolent sexist ideals would mirror the eVects of exposure to agentic and communal stereotypes, respectively (see also Glick & Fiske, 2001). Second, we assessed the extent to which exposure to complementary gender stereotypes would aVect beliefs regarding the legitimacy of the overall social system, in general, and not just the specific system of gender relations. Third, we made use of an incidental exposure manipulation that did not require the participants to actually endorse or refute these stereotypes. Fourth and finally, we addressed the issue of whether any favorable statement regarding women would be suYcient to increase system justification levels or whether these eVects are dependent on the activation of preexisting, culturally available stereotypes associated with communal and agentic dimensions. To this end, participants in this study were randomly assigned to one of four diVerent conditions. Specifically, they were exposed to: (1) four benevolent sexism statements, (2) four hostile sexism statements, (3) two hostile and two benevolent sexism statements, or (4) four favorable, but nonstereotypical, statements about women (i.e., describing them as more creative, realistic, tactful, and resourceful than men). These four conditions were crossed with a manipulation of exposure versus endorsement. In the endorsement conditions, participants were again asked to indicate their degree of agreement or disagreement with each statement. In the exposure condition, participants were instead asked to proofread each item and to rate the ambiguity versus clarity of item wording. A control condition was also included in which participants were not exposed to any of these statements at all. Next, all participants completed a measure of general or diVuse system justification. The items are listed in Table II. An overall index was calculated by taking the mean of responses to all eight items following recoding. In terms of convergent validity, Kay and Jost (2003) found that these diVuse system justification scores correlated significantly with (1) scores on Lipkus’ (1991) Global Belief in a Just World scale, r(117) ¼ .67, p < .001; (2) Quinn
SYSTEM JUSTIFICATION THEORY
ITEMS
FOR
319
TABLE II MEASURING DIFFUSE (GENERAL) SYSTEM JUSTIFICATION
In general, you find society to be fair In general, the political system operates as it should American society needs to be radically restructured (R) The United States is the best country in the world to live in Most policies serve the greater good Everyone has a fair shot at wealth and happiness Our society is getting worse every year (R) Society is set up so that people usually get what they deserve Note: Participants indicated their agreement or disagreement on a scale ranging from one (strongly disagree) to nine (strongly agree). Items followed by ‘‘(R)’’ were reverse‐scored prior to data coding and analyses (Kay & Jost, 2003).
ITEMS
FOR
TABLE III MEASURING PERCEPTIONS OF ‘‘BALANCE’’ AND ‘‘COMPLEMENTARITY’’ IN THE SOCIAL WORLD
All in all, the world is a balanced place Some people have everything, while others have nothing (R) A person who has recently experienced a string of bad breaks probably has something good coming to him or her Masculine traits perfectly complement feminine traits (and vice versa) I agree with people who say that ‘‘everything comes out in the end’’ The dice are basically ‘‘loaded’’; positive outcomes are distributed disproportionately to the ‘‘winners’’ in society (R) Most people have both good and bad characteristics Everything has its advantages and disadvantages The social world is almost never in a state of ‘‘harmony’’ or ‘‘equilibrium’’ (R) Note: Participants indicated their agreement or disagreement on a scale ranging from one (strongly disagree) to seven (strongly agree). Items followed by ‘‘(R)’’ were reverse‐scored prior to data coding and analyses (Kay & Jost, 2003).
and Crocker’s (1999) Protestant Work Ethic (PWE) scale, r(49) ¼ .45, p < .001; and (3) a new measure of general beliefs concerning needs for ‘‘balance’’ and ‘‘complementarity’’ in the social world, r(117) ¼ .37, p < .001 (see items in Table III). The results from this study yielded several findings of note. First, exposure to the benevolent sexism items increased system justification among women, much as exposure to the communal items had in the previous study. That is,
320
AARON C. KAY et al.
for female participants, system justification was significantly higher following exposure to the benevolent sexism statements, compared to both the positively valenced, nonstereotypical control and ‘‘nothing’’ control conditions (see Fig. 3). Second, this experiment demonstrated that the system‐ justifying eVects of complementary gender stereotypes are not limited to beliefs regarding the legitimacy of gender relations, but they apply more generally to the societal status quo. Third, because the results for the incidental exposure (proofreading) conditions did not diVer substantially from those obtained in the endorsement conditions, we obtained further evidence that mere exposure to complementary stereotypes, even in the absence of opportunities for personal endorsement, aVects women’s degree of support for the societal status quo. Across these two studies, then, we found that the activation of communal and benevolent stereotypes was suYcient to increase system justification among women. Complementary gender stereotypes may be especially
Female participants
Male participants
Diffuse system justification
5.50
a
a
a
a
a
5.00
ab
a
ab
4.50 ab
4.00
b
3.50
N ot hi ng
co nt ro l
N on st er eo ty pi ca l H os til e se xi sm Be ne vo le C nt om se pl xi em sm en ta ry st er eo ty pe s
3.00
Fig. 3. Mean scores on diVuse system justification as a function of participant gender and content of stereotype exposure, collapsed across proofreading and endorsement instructions. (Adapted from Study 2 in Jost & Kay, 2005.) Note: DiVerent superscripts within each gender group diVer from one another according to Tukey tests of multiple comparison (p < .05).
SYSTEM JUSTIFICATION THEORY
321
eVective at rationalizing the status quo because they, almost by definition, appear to compensate for women’s deprivation in terms of status and power (Bem & Bem, 1970; Jackman, 1994). This interpretation is broadly consistent with the fact that only the communal and benevolent stereotypes about women (and not the agentic stereotypes about men) aVected women’s system justification scores in the first two studies. That is, the communal and benevolent stereotypes may have acted as a counterweight to the presumptive advantages enjoyed by men. 3. Altering the Presumptive Context of Gender Inequality If the above interpretation is correct, then a manipulation that temporarily reverses the status diVerences between men and women by portraying men as disadvantaged relative to women should reverse the eVects of complementary (or compensating) stereotypes on system justification. That is, when the presumptive social context is altered so that women are seen as having a status advantage over men, then agentic stereotypes about men (rather than communal or benevolent stereotypes of women) should lead to increased levels of system justification for men as well as women. In such a context, ascribing unique advantages to women should do little to ‘‘balance out’’ the system; only stereotypes that ascribe compensating advantages to men should be system justifying. The third study reported by Jost and Kay (2005) tested this reasoning. A group of Canadian and American participants was first exposed to a manipulation that either reinforced the presumed status advantage of men relative to women (i.e., the cultural default) or reversed that advantage. Afterward, participants were exposed to either communal stereotypes of women or agentic stereotypes of men using the same proofreading task that was used in the second study. Participants were asked to proofread four statements regarding either the communal nature of women or the agentic nature of men. To manipulate the relative advantage of men and women, participants read a description of alleged research findings that described either women or men as more naturally suited to being better managers; these materials were derived from actual scientific and journalistic accounts of female leadership qualities in business (Eagly & Johannesen‐Schmidt, 2001; Sharpe, 2000). In the ‘‘women are better managers’’ condition, participants read that: Research has demonstrated convincingly that the best managers in business settings tend to have excellent interpersonal skills and are able to communicate well and work closely with others. Consequently, the most eVective managers in recent years have tended to be women rather than men.
322
AARON C. KAY et al.
In the ‘‘men are better managers’’ condition, participants instead read that: Research has demonstrated convincingly that the best managers in business settings tend to have excellent individual leadership skills and are able to solve problems independently. Consequently, the most eVective managers in recent years have tended to be men rather than women.
After reading one of these two passages, participants were exposed to prescriptive stereotypical statements taken from research by Prentice and Carranza (2002). Finally, participants completed the diVuse system justification scale (see Table II). The experimental design for this study, therefore, was a 2 (presumptive context: women versus men are better managers) 2 (stereotype exposure: women are communal versus men are agentic) factorial, and the dependent variable was the perceived legitimacy of the societal status quo. As hypothesized, the interaction between the two independent variables was statistically significant. The pattern of means is illustrated in Fig. 4. When the presumptive context of male advantage was maintained, exposure to communal stereotypes about women was slightly more eVective at increasing system justification levels (although this pairwise comparison did not Women are better managers Men are better managers
Diffuse system justification
5.50
5.00
4.50
4.00
3.50
3.00 Men are agentic
Women are communal
Fig. 4. Mean scores on diVuse system justification as a function of managerial context and content of stereotype exposure. (Adapted from Study 3 in Jost & Kay, 2005.)
SYSTEM JUSTIFICATION THEORY
323
attain statistical significance). When the presumptive gender diVerence was reversed, however, so that women (because of their communal nature) were described as enjoying a superior managerial status, exposure to the male agentic stereotype led to significantly higher levels of system justification, as compared with exposure to the female communal stereotype. Thus, this experiment further strengthens our theoretical argument that complementary gender stereotypes can serve to maintain support for the status quo, and it oVers more precise evidence for the eVectiveness of the compensating (or balancing) mechanism in particular. 4. Summary: System‐Justifying Functions of Complementary Gender Stereotypes In three experiments first reported by Jost and Kay (2005), we have shown that exposure to communal and benevolent stereotypes is suYcient to bolster ideological support for the status quo (at least among women). Furthermore, exposure to complementary gender stereotypes increased both gender‐specific and diVuse forms of system justification. These results are quite consistent with several prominent accounts of the assumed connection between gender stereotyping and system maintenance (Bem & Bem, 1970; Eagly & SteVen, 1984; Glick & Fiske, 2001; HoVman & Hurst, 1990; Jackman, 1994; Jost & Banaji, 1994). In particular, we have highlighted the potential for complementary stereotypes to counteract the status disadvantage of women, thereby helping people to feel better about gender inequalities in society. Taken as a whole, this evidence provides strong support for novel hypotheses derived from SJT (see also Kay & Jost, 2003). One other aspect of these studies is worth mentioning. In earlier work, system justification was largely inferred from studies that documented either (1) the existence or ramifications of in‐group versus out‐group favoritism as a function of group status (Jost & Burgess, 2000; Jost et al., 2002; see also Jost & Hunyady, 2002 for a review) or (2) the degree to which (presumably) system‐justifying stereotypes were consensually endorsed by members of high‐ and low‐status groups (Jost et al., 2001, 2005). In these earlier studies, system justification had not yet been measured directly as an outcome variable. Nor had it been measured in previous studies of gender stereotyping as rationalization (Eagly & SteVen, 1984; Glick & Fiske, 2001; HoVman & Hurst, 1990). Thus, there was no direct evidence in the published literature that exposure to specific kinds of stereotypes serves to increase ideological support for the existing social system. The studies we have just described, in conjunction with a set of studies that we will describe next, represent the first experimental tests of the causal relationship between specific stereotypes and perceptions of system legitimacy.
324
AARON C. KAY et al.
B. BEYOND GENDER: COMPLEMENTARY STEREOTYPES RATIONALIZE STATUS DIFFERENCES IN GENERAL The fact that the system‐justifying potential of gender stereotypes can change depending on whether women are seen as occupying a high‐status or a low‐status position suggests that these eVects are more general than gender theorists may have assumed. That is, the system‐justifying consequences of complementary (or compensatory) stereotypes are probably not unique to gender per se, but may be part of a more general social psychological process of rationalizing inequality in general (see also Kay & Jost, 2003; Kay et al., 2005; Napier et al., 2006; Yzerbyt, Provost, & Corneille, 2005). In fact, if we divide the communal dimension into two subdimensions, one that captures warmth, friendliness, happiness, and likeability, and another that centers on honesty, morality, tradition, and virtue, the resulting subdimensions suggest important similarities to stereotypes of a number of groups other than gender groups. If ascribing these positive characteristics to low‐status groups is in fact system justifying, then we should observe a very general pattern of complementary stereotyping across a wide range of diVerent low‐status groups. 1. Regional and Ethnic Status Stereotypes We have addressed this issue in several studies with respect to stereotypes based on regional and ethnic status in the United States, Italy, England, and Israel (see Jost et al., 2001, 2005). In all of these very diVerent contexts, we have observed a strikingly similar pattern of results. In the United States, for instance, we find that both Northerners and Southerners tend to hold similar regional stereotypes that seem to be related to social and economic status diVerences (Jost et al., 2001). Specifically, a sample of students at the University of Cincinnati (which is located on the border between the North and the South) reported believing that Northerners in the United States are more intelligent and productive than Southerners but that Southerners are more happy and honest than Northerners (see Fig. 5). The fact that these complementary stereotypes are consensually shared by Northerners and Southerners is important because it suggests that these stereotypes serve system‐justifying (rather than ego‐ or group‐justifying) functions. Status diVerences between Northerners and Southerners in Italy are even weightier than in the United States (see Capozza, Bonaldo, & DiMaggio, 1982). They are summarized well by the racist slight, ‘‘Africa begins in Rome.’’ In a study conducted by Jost et al. (2005, Study 1), students from Northern, Central, and Southern universities in Italy (N ¼ 160) completed a series of stereotyping measures. As in the United States, Northerners were
325
SYSTEM JUSTIFICATION THEORY
Northerners
Southerners
7.00
Ratings of target group
6.00
5.00
4.00
3.00 Intelligent
Productive
Happy
Honest
Fig. 5. Trait ratings of Northerners and Southerners in the United States.
consensually perceived to be more intelligent and productive than Southerners. At the same time, Southerners were consensually perceived to be more happy (see Fig. 6), although they were not perceived as more honest, probably because of stereotypes linking Sicilians and other Southern Italians to the mafia. In England, where it is the Southerners rather than the Northerners who are seen as higher in social and economic status, the same general pattern of stereotypes emerges (see Jost et al., 2005, Study 2). Students from the North and the South (N ¼ 94) again stereotyped the higher status group (Southerners) as more intelligent and productive, but they stereotyped the lower status group (Northerners) as more happy and honest (see Fig. 7). Furthermore, we found that the perceived magnitude of status diVerences was associated with increased endorsement of complementary stereotypes, which, in turn, was associated with enhanced perceptions of the legitimacy and stability of the status hierarchy (Jost et al., 2005, pp. 320–321). In Israel, we replicated these results using ethnic rather than regional status diVerences (Jost et al., 2005, Study 3; N ¼ 135). In this context, the high‐status group of Ashkenazi Jews was stereotyped as more intelligent and responsible, whereas the low‐status group of Sephardic Jews was stereotyped
326
AARON C. KAY et al.
Northerners
Southerners
Trait ratings of target groups
7
6
5
4
3 Intelligent Fig. 6.
Productive
Happy
Honest
Trait ratings of Northerners and Southerners in England.
Northerners
Southerners
8.00 7.50
Ratings of target group
7.00 6.50 6.00 5.50 5.00 4.50 4.00 3.50 3.00 Intelligent
Productive
Happy
Fig. 7. Trait ratings of Northerners and Southerners in Italy.
Honest
327
SYSTEM JUSTIFICATION THEORY
Ashkenazim
Sephardim
8.00 7.50
Ratings of target group
7.00 6.50 6.00 5.50 5.00 4.50 4.00 Intelligent
Responsible
Happy
Family values
Fig. 8. Trait ratings of Ashkenazim and Sephardim in Israel.
as happier and as more supportive of family values (these respondents were not asked about their perceptions concerning honesty). As before, these complementary stereotypes were shared by members of both high‐ and low‐status groups. They were also endorsed more strongly following an experimental manipulation of system threat (see Fig. 8), providing further evidence that complementary stereotypes serve the function of bolstering the societal status quo (see also Kay et al., 2005; Napier et al., 2006). In any case, it appears that intergroup diVerentiation with respect to agentic and communal stereotypes is not at all unique to gender groups (see also Conway, Pizzamiglio, & Mount, 1996; Judd, James‐Hawkins, Yzerbyt, & Kashima, 2005; Ridgeway, 2001; Yzerbyt et al., 2005). On the contrary, the tendency to ascribe agentic characteristics to members of high‐status groups and communal characteristics to members of low‐status groups is pervasive indeed. 2. The Virtues of the Oppressed Kay and Jost (2003) observed that representations of the poor as happier and more honest than the rich abound in works of literature, film, and popular culture more generally. Characterizations found in such diverse
328
AARON C. KAY et al.
works as Dickens’ Great Expectations and A Christmas Carol, Moliere’s The Miser, Fitzgerald’s The Great Gatsby, Orson Welles’ Citizen Kane, Herman Hesse’s Siddhartha, and even Steve Martin’s The Jerk reflect a propensity to ‘‘balance out’’ accounts of diVerences in wealth with accounts of oVsetting or compensating traits and characteristics. This tendency to ascribe redeeming qualities to the poor and downtrodden was noted by Bertrand Russell (1950) in an essay entitled ‘‘The Superior Virtue of the Oppressed.’’ Russell wrote that: A rather curious form of this admiration for groups to which the admirer does not belong is the belief in the superior virtue of the oppressed: subject nations, the poor, women, and children. The eighteenth century, while conquering America from the Indians, reducing the peasantry to the condition of pauper laborers, and introducing the cruelties of early industrialism, loved to sentimentalize about the ‘‘noble savage’’ and the ‘‘simple annals of the poor’’ (p. 58).
Such representations are indeed quite common. Blasi and Jost (2006) noted that there are over 800 Google ‘‘hits’’ that contain English phrases such as the ‘‘happy vagrant’’ and the ‘‘carefree vagabond.’’ It seems likely that complementary stereotypes play a role in racial attitudes as well as in attitudes toward the rich and the poor. A short story by Langston Hughes (1933/1971) entitled ‘‘Slave on the Block’’ begins by excoriating a White couple for their romantic views about the virtues of Blacks: They were people who went in for Negroes—Michael and Anne—the Carraways. But not in the social‐service, philanthropic sort of way, no. They saw no use in helping a race that was already too lovely for words. Leave them unspoiled and just enjoy them, Michael and Anne felt. So they went in for the Art of Negroes—the dancing that had such jungle life about it, the songs that were so direct, so real (p. 19).
One is reminded also of the bigoted character in the (1957) film Twelve Angry Men, who insists that, ‘‘Oh, sure, there are some good things about them, too. Look, I’m the first to say that.’’ Kay and Jost (2003) proposed that by reinforcing the notion that the material advantages of the wealthy are accompanied by specific deficits in areas such as moral virtue and happiness and by stressing that the material deprivation of the poor is counterbalanced by other virtues, people are able to maintain an illusion of equality. Such representations, we hypothesized, help to preserve the legitimacy of the system, insofar as people are able to believe that there is, overall, an equal distribution of benefits and burdens throughout society. This system maintenance hypothesis had been suggested (but never directly tested) in previous work by, among others, Lane (1962) and Lerner (1980).
SYSTEM JUSTIFICATION THEORY
329
3. EVects of Exposure to ‘‘Poor but Happy’’ and ‘‘Poor but Honest’’ Stereotype Exemplars on DiVuse System Justification We designed four experiments to investigate the potential of complementary stereotypes concerning the rich and poor to serve as system‐justifying devices, that is, to aid individuals in shoring up support for the existing social system even in the face of visible inequality (see also Jost & Banaji, 1994; Napier et al., 2006). Our interest in these studies was not to gauge the prevalence of these representations per se—although they do seem to be quite common in popular culture, but rather to test for the eVects of exposure to ‘‘poor but happy,’’ ‘‘rich but miserable,’’ ‘‘poor but honest,’’ and ‘‘rich but dishonest’’ stereotype exemplars on the perceived legitimacy of the social system. In the first experiment (Kay & Jost, 2003, Study 1), under the guise of an impression formation task, we exposed participants to four character vignettes depicting a fictional character, Mark, as either rich and unhappy (complementary), rich and happy (noncomplementary), poor and happy (complementary), or poor and unhappy (noncomplementary). Thus, the design was a 2 2 factorial in which the protagonist was described as either rich or poor and as either happy or unhappy. Afterward, in what was purported to be a separate study, participants in all four conditions completed the diVuse system justification scale (see items in Table II). Our hypothesis was that participants would deem the system to be the most legitimate following exposure to the complementary (rich but unhappy, poor but happy) compared to the noncomplementary (rich and happy, poor and unhappy) stereotype exemplars. In other words, we predicted a two‐way interaction between level of wealth (rich versus poor) and level of happiness (happy versus unhappiness) of the protagonist, so that when Mark was described as poor, describing him also as happy (versus unhappy) would lead people to score higher on system justification, whereas when he was described as rich, describing him as unhappy (versus happy) would lead people to score higher on system justification. The vignettes were worded as follows (with alternate versions designated with the use of brackets and italics): Mark is from a large Northeast city. He is married and has two children, has brown hair and is 5 feet 11 inches. Mark was an athletic child and still closely follows all his local sports teams. Mark enjoys almost all aspects of his life [is not particularly happy with most aspects of his life], and [but] because of his high [low] salary, he has [has no] trouble getting the bills paid and keeping food on the table. In June, Mark will be turning 41.
Thus, the passages contained the relevant information pertaining to levels of wealth and happiness as well as some individuating information about the protagonist. The passages about Mark, it can be seen, had no clear or
330
AARON C. KAY et al.
explicit connection of any kind to the very general system‐related attitudes that are measured by the diVuse system justification scale. Nevertheless, and in line with predictions, a significant two‐way interaction between wealth and happiness levels of the protagonist was observed on participants’ system justification scores. As can be seen in Fig. 9, people who were exposed to the ‘‘poor but happy’’ protagonist judged the overarching social system to be more fair and legitimate, in comparison with people who were exposed to the ‘‘poor and unhappy’’ protagonist. Conversely, people who were exposed to the ‘‘rich and happy’’ protagonist judged the overarching social system to be less fair and legitimate, in comparison with people who were exposed to the ‘‘rich and unhappy’’ protagonist. Thus, the two complementary representations in which the poor were portrayed as happy and the rich as unhappy served to increase ideological support for the societal status quo. It is worth emphasizing a counterintuitive aspect of this pattern of results, namely, that when the level of wealth was held constant (as rich), increasing the happiness of the target actually decreased the perceiver’s satisfaction with the status quo. As Russell (1950), Lane (1962), Lerner (1980), and others have suggested, believing that the poor are more honest and virtuous than the rich (and perhaps also more likely to be rewarded in the afterlife) may be a powerful means of rationalizing status inequalities in society. Thus, we conducted a second experiment to investigate the system‐justifying potential of ‘‘poor but honest’’ and ‘‘rich but dishonest’’ complementary stereotypes (Kay & Jost,
Happy
Unhappy
Diffuse system justification
5.00 4.50 4.00 3.50 3.00 2.50 2.00 Rich protagonist
Poor protagonist
Fig. 9. EVects of exposure to rich versus poor protagonists who are described as happy versus unhappy on system justification. (Adapted from Study 1 in Kay & Jost, 2003.)
SYSTEM JUSTIFICATION THEORY
331
2003, Study 2). The vignettes for this study were as follows (with alternate versions italicized and set oV by brackets): George is from a large Northeast city. He is married and has two children, has brown hair and is 5 feet 11 inches. George was an athletic child and still closely follows all his local sports teams. George sometimes [never] cuts corners, and other people consider him to be somewhat dishonest [very honest]. Because of his high [low] salary, he has [has no] trouble getting the bills paid and keeping food on the table. In June, George will be turning 38.
Once again, we predicted that exposure to the complementary representations (i.e., ‘‘poor but honest,’’ ‘‘rich but dishonest’’) would produce an increase in participants’ levels of system justification, relative to their scores in the noncomplementary (i.e., ‘‘poor and dishonest,’’ ‘‘rich and honest’’) conditions. This is precisely what we observed. The protagonist’s wealth significantly interacted with his perceived level of honesty, so that when the protagonist was described as poor, the system was viewed more favorably when he was also described as honest rather than dishonest, but when he was described as rich, the system was viewed more favorably when he was also described as dishonest rather than honest (see Fig. 10). Thus, participants rated the system as more fair and legitimate following exposure to the complementary
Honest
Dishonest
General system justification
5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 Rich protagonist
Poor protagonist
Fig. 10. EVects of exposure to rich versus poor protagonists who are described as honest versus dishonest on system justification (Adapted from Study 2 in Kay & Jost, 2003.)
332
AARON C. KAY et al.
pairings of ‘‘poor but honest’’ and ‘‘rich but dishonest.’’ It is worth noting that these findings are quite diVerent from what one would expect on the basis of theoretical perspectives that stress balance, congruity, or ‘‘halo’’ eVects (see Kay & Jost, 2003). Nevertheless, the pattern of results obtained in these experiments nicely parallel the eVects on system justification that Jost and Kay (2005) observed after exposing participants to benevolent stereotypes of women (which portray them as, among other things, more virtuous than men). In both cases, stereotypes that depict the lower status group as more virtuous than the higher status group produced elevated perceptions of the legitimacy of the status quo (see also Mandisodza et al., 2006). This may help to explain why people are so eager to appreciate, as Russell (1950) put it, the ‘‘virtues of the oppressed.’’ 4. EVects of Exposure to ‘‘Poor but Happy’’ and ‘‘Poor but Honest’’ Stereotype Exemplars on Implicit Activation of the Justice Motive Inspired by the work of Carolyn Hafer (2000), who demonstrated that situations of injustice activate the justice motive at an implicit level of awareness, we sought to replicate our eVects using an implicit dependent measure. Hafer found that exposure to situations that are clearly unjust (such as a perpetrator of a crime not being brought to justice) led to increases in the activation of the justice motive, as measured by enhanced accessibility of justice‐related words (operationalized in terms of interference on a Stroop task). In other words, she demonstrated that unjust situations spontaneously activate justice concerns at a nonconscious level of awareness. Applying this logic, we endeavored to first expose participants to complementary (versus noncomplementary) stereotypical exemplars, and then to measure the implicit activation of the justice motive using a reaction time paradigm. To the extent that noncomplementary representations are perceived as less just than complementary representations, we expected to see increases in the accessibility of justice‐related words following exposure to them. We examined this hypothesis first in the context of ‘‘poor but happy’’/ ‘‘rich but miserable’’ complementary stereotypes (Kay & Jost, 2003, Study 3). Participants were exposed to one of two sets of stimuli that described two friends, one rich and one poor, in either complementary or noncomplementary terms. In the complementary condition, the rich friend was described as less happy than the poor friend. In the noncomplementary condition, the rich friend was described as happier than the poor friend. The first part of the vignette was identical in both conditions. It read: Joseph and Mitchell both grew up in the midwestern United States and now both live in Seattle. Joseph is 39 and Mitchell is 41. Joseph and Mitchell met in their twenties,
SYSTEM JUSTIFICATION THEORY
333
were good friends for almost ten years afterwards, but now because of their work schedules, they have lost touch over the last few years. Joseph has an excellent job now, lives in a beautiful, spacious house in a lavish neighborhood, and makes a very large salary. Mitchell spends a lot of his time watching and playing sports, but unlike Joseph, his job doesn’t pay him much, so his home, which is in a rather inexpensive part of town, is a bit cramped and not very nice‐looking.
In the complementary, ‘‘poor happier than rich’’ condition, the story concluded this way: Despite Mitchell’s smaller house and lower salary, he tends to be much happier with his life than Joseph is. Mitchell enjoys most aspects of his life and is known amongst his friends as that guy who’s always ‘‘broke but happy.’’ Joseph, on the other hand, lacks the feeling of general contentment that Mitchell has and is often thought of as that ‘‘rich but miserable guy.’’
In the noncomplementary, ‘‘rich happier than poor’’ condition, the story concluded as follows: Not only does Mitchell have a smaller house and lower salary than Joseph, he also tends to be much less happy with his life than Joseph is. Joseph enjoys most aspects of his life and is known amongst his friends as that guy who ‘‘has it all.’’ Mitchell, on the other hand, lacks the feeling of general contentment that Joseph has and is often thought of as that ‘‘broke, miserable guy.’’
Participants were instructed that this description would be used as part of a memory task, and they were asked to read the vignette as many times as necessary in order to be able to answer questions about it later. Before the memory task took place, however, participants were asked to complete a task pertaining to a separate study on the computer. This task, which was a lexical‐decision task, was used to gauge the relative accessibility of justice‐ related and non‐justice‐related words. Specifically, participants were required to judge, as quickly as possible, whether strings of letters presented to them were words or nonwords by pressing designated computer keys. Participants were exposed to justice‐related words (i.e., fair, legitimate, just, valid, justified), neutral words (e.g., volume, finger, calendar, candle), and nonwords; the justice‐related and neutral words were matched for familiarity. The extent to which participants were faster to categorize justice‐related (versus neutral) words as words was used as a measure of implicit activation of the justice motive. Given the results reported by Hafer (2000), we expected that the ‘‘rich happier than poor’’ vignette would lead people to exhibit increased justice concerns, relative to the ‘‘poor happier than rich’’ vignette. As illustrated in Fig. 11, we found that the accessibility of justice‐related words was indeed higher in the noncomplementary condition (as indicated
334
AARON C. KAY et al.
Noncomplementary
Complementary
Neutral words
Justice words
Mean reaction time (ms)
750
700
650
600
550
Fig. 11. EVects of exposure to complementary versus noncomplementary stereotype exemplars on accessibility of justice‐related and neutral words. (Adapted from Study 3 in Kay & Jost, 2003.)
by faster lexical decision reaction times), that is, when the rich friend was described as happier than the poor friend. The manipulation of perceived complementarity exerted no comparable eVects on response times to neutral control words. Thus, even when measured at an implicit level of awareness, it appears that the complementary ‘‘poor but happy’’ stereotype exemplar is more satisfying (in terms of system justification needs) than is its noncomplementary counterpart (rich and happy). In a follow‐up experiment, we substituted honesty and dishonesty for happiness and unhappiness, and obtained a nearly identical pattern of results (Kay & Jost, 2003, Study 4). The noncomplementary condition again appeared to threaten the justice (or justification) motive, as indicated by faster reaction times to justice‐related words. As before, the two conditions exerted no eVect on reaction times to the neutral words. Thus, using multiple independent and dependent measures, we have obtained consistent evidence in support of the notion that complementary stereotypes of the rich and poor can serve system‐justifying functions. As with complementary gender stereotypes of men as agentic but not communal and women as communal but not agentic, complementary representations of the rich and poor lead people to believe more strongly in the fairness, legitimacy, and overall desirability of the societal status quo. Maintaining
SYSTEM JUSTIFICATION THEORY
335
the notion that we live ‘‘in the best of all possible worlds’’ is no simple task in the context of social systems (such as patriarchy and capitalism) that are associated with dramatic and visible disparities between groups. Fortunately (or unfortunately), people show remarkable skill and flexibility in their capacity to invent or simply to latch onto ready‐made rationalizations for the status quo (see also Blasi & Jost, 2006; Jost & Hunyady, 2005). The studies first reported by Kay and Jost (2003) demonstrate the system‐ enhancing potential of several culturally familiar complementary (or compensatory) representations of the rich and poor. It appears that simply being reminded of one or more ‘‘virtues of the oppressed’’ leads people to profess greater satisfaction with the overarching social system at an explicit level of awareness and to exhibit decreased concerns for justice at an implicit level of awareness.
IV. Moderators of the EVect of Complementary Stereotypes on System Justification For several decades, much of the research in the social justice tradition has emphasized the ways in which people maintain their ‘‘belief in a just world’’ by engaging in processes of victim‐blaming or victim‐derogation. That is, we know that people frequently ascribe negative characteristics to members of disadvantaged groups, such as the poor and dispossessed, apparently in order to deflect blame away from the system itself and onto individual victims (Furnham & Gunter, 1984; Hafer & Begue, 2005; Lerner, 1980; Montada & Schneider, 1989). This approach stresses attributional consistency as an outcome, so that those who ‘‘succeed’’ are seen as good in various ways, and those who ‘‘fail’’ are seen as inherently flawed (see also Crandall & Beasley, 2001). Our research on the system‐justifying potential of complementary stereotypes, however, suggests that a process that is akin to victim‐enhancement— such that favorable characteristics are ascribed to members of disadvantaged groups—can also serve to increase allegiance to the overarching social system. Indeed, the seven experiments by Jost and Kay (2005; Kay & Jost, 2003, described above) show clearly that stereotypical representations that endow members of low‐status groups with compensating virtues and members of high‐status groups with leveling vices can serve to increase people’s faith in the fairness and legitimacy of the societal status quo. How do we reconcile our findings with the long tradition of research on victim‐blaming as a defensive response to injustice and inequality (Lerner, 1980; Ryan, 1976)? Our assumption is that in some contexts and for some
336
AARON C. KAY et al.
groups of people, victim‐derogating, noncomplementary justifications are especially eVective as system‐justifying devices, whereas in other contexts and for some people, victim‐enhancing, complementary stereotypes are more eVective. In other words, the system‐justifying potential of complementary stereotypes, such as ‘‘poor but happy’’ and ‘‘poor but honest’’ representations, may be moderated by both contextual constraints and individual diVerences. Thus, we have maintained an active interest in uncovering the moderating variables and boundary conditions that characterize our eVects (Kay & Jost, 2003; Kay et al., 2005; Kay, Czaplinski, & Jost, 2007a). To date, we have identified several situational and dispositional factors that aVect the means by which people attempt to maintain their belief in the legitimacy of the system as well as the eVectiveness of those means. Taken as a whole, this evidence suggests that there may be several alternative, possibly functionally equivalent routes to attaining system justification. In the language of goal systems theory, it appears that there are multiple, substitutable means of satisfying the goal of justifying the status quo (see Jost, Pietrzak, Liviatan, Mandisodza, & Napier, in press; Kruglanski, 1996; see also Tesser, Martin, & Cornell, 1996); some of these means presumably rely on direct, noncomplementary, victim‐derogating ‘‘strategies,’’ whereas others may be more subtle and indirect, like the complementary, victim‐enhancing approaches we have stressed in this chapter. Which route is ‘‘chosen’’ depends on a number of factors, including situational constraints, dispositional tendencies, and the contents of readily available (prepackaged) stereotypes. Our use of goal‐related terminology such as ‘‘strategy’’ and ‘‘choice’’ should not be taken to mean that these processes are necessarily consciously accessible to social actors and perceivers. Rather, we assume that system justification goals can operate implicitly or nonconsciously (see also Bargh et al., 2001).
A. PERCEPTIONS OF A CAUSAL LINK BETWEEN TRAIT AND OUTCOME Several social psychological theories assume that human beings are motivated to believe in a predictable and controllable social world (Allport, 1966; JanoV‐Bulman & Yopyk, 2004; Kluegel & Smith, 1986; Langer, 1975; Lerner, 1980; Major, 1994; Plaks et al., 2005). This motivation is thought to be so strong that when people encounter evidence that some events are uncontrollable, chaotic, or randomly determined, they generally respond by reconstruing things so as to minimize the threat to feelings of controllability (Hafer & Begue, 2005). One highly eVective means of coping with this situation is to explain seemingly arbitrary or random acts of injustice or misfortune as due to, or caused by, traits, dispositions, or other
SYSTEM JUSTIFICATION THEORY
337
characteristics that are internal to the victim in question. Accordingly, poor people are often assumed to be lazy and unintelligent (Jost & Banaji, 1994), obese people are assumed to lack self‐control (Crandall, 1994; Quinn & Crocker, 1999), and so on. This implies that the eVectiveness of victim‐ blaming attributions should depend on the perception of a specific causal link between the trait (e.g., intelligence) and the outcome (e.g., wealth). In contrast, when the trait is perceived as causally irrelevant to the outcome, victim‐derogation is unlikely to justify the social system. Victim‐enhancement, by contrast, is assumed to justify the system by implying a fair dispersion of benefits and burdens across social groups rather than a sense of prediction or control. Complementary stereotypes (which are victim‐enhancing) reinforce the desire to believe that no one ‘‘has it all’’ and that, as Dr. Pangloss had it, bad luck in one domain is oVset by good luck in others. It seems important that the compensating benefits commonly ascribed to members of disadvantaged groups (their ‘‘virtues’’) are causally unrelated to the dimension on which they are disadvantaged. In this manner, the system as a whole can be seen as fair, because in the long run everyone encounters both rewards and setbacks. There is, in other words, an illusion of equality. Thus, we propose that there are two distinct, alternative ways of justifying the status quo (and maintaining the belief in a just world) when confronted with unequal outcomes. First, one can blame victims for their own state of disadvantage, thereby reinforcing the extent to which events and outcomes are seen as predictable and controllable (Lerner, 1980). Second, one can conceive of ways of restoring the illusion of equality by stressing the virtues of the disadvantaged and the vices of the advantaged (Kay & Jost, 2003). If our theoretical logic is correct, then people will tend to pursue the first route to system justification (i.e., relying on victim‐derogating stereotypes and judgments) when a given trait is seen as causally related to the specific outcome in question, as when a person’s purported laziness is used to explain (or justify) his or her poverty. By contrast, people will tend to pursue the second route to system justification when there is no perceived causal link between the trait and the outcome, as when the poor are romanticized as happier or as more honest than the rest of us. We investigated these hypotheses in a series of experiments reported by Kay et al. (2005) on the moderating role of perceptions of trait–outcome causality. 1. EVects of Exposure to Victim‐Derogating Versus Victim‐Enhancing Representations on DiVuse System Justification In one experiment, we exposed people to victim‐derogating and victim‐ enhancing representations under diVerent circumstances and measured their subsequent degree of support for the societal status quo (Kay et al., 2005,
338
AARON C. KAY et al.
Study 2). Pretesting revealed that most people regard intelligence as causally related to economic outcomes such as wealth and poverty but as causally unrelated to physical attractiveness. Thus, study participants were exposed to one of four vignettes according to a 2 (victim‐enhancing, complementary versus victim‐derogating, noncomplementary representations) 2 (trait causally relevant versus irrelevant to outcome) factorial design. For conditions in which a perceived causal link existed between trait (intelligence) and outcome (wealth/poverty), the complementary (and, in italics, the noncomplementary) passage read: Mary and Sarah both grew up in the Midwestern United States and now both live in Seattle. Mary is 34 and Sarah is 33. Mary and Sarah met in their teens, were good friends for almost ten years afterwards, but now because of their work schedules they have lost touch over the last few years. Mary is very bright but not very wealthy [is both very bright and now very wealthy]. Sarah, on the other hand, is not very smart at all, but is now very wealthy [is not very wealthy at all, and is also much less intelligent than Mary]. Because of these diVerences, Mary is always thought of as that women ‘‘who had an easy time getting good grades, but now has a hard time paying the bills’’ [‘‘who is both smart and rich’’] and Sarah is always thought of as the girl ‘‘who is not very smart but is very rich’’ [‘‘a hard time getting decent grades and now has a hard time paying the bills’’].
For conditions in which no perceived causal connection existed between the trait (intelligence) and outcome (attractive/unattractive), the complementary (and, in italics, the noncomplementary) passage began the same way and continued as follows: Mary, although not particularly bright, is very pretty [is both very bright and very pretty]. Sarah, on the other hand, is not generally considered to be very good‐looking, but is without a question much more intelligent than Mary [is not generally considered to be very good looking, and is also much less intelligent than Mary]. Because of these diVerences, in college Mary was always thought of as that girl ‘‘who got lots of looks from the boys but no good grades from the professors’’ [‘‘who got lots of looks from the boys and lots of A’s from the teachers’’] and Sarah was always thought of as the girl who had ‘‘an easy time getting an A but a hard time getting a date’’ [‘‘a hard time getting decent grades and dates’’].
Afterward, participants completed the diVuse system justification scale (see Table II). We hypothesized that when a perceived causal connection between trait and outcome existed, exposure to noncomplementary representations (derogating the ‘‘loser,’’ lionizing the ‘‘winner’’) would lead to an increase in system justification. However, when no such causal connection was perceived, exposure to complementary, oVsetting representations (elevating the ‘‘loser,’’ downgrading the ‘‘winner’’) would lead to increased system justification.
SYSTEM JUSTIFICATION THEORY
339
These predictions were indeed supported by the results of our experiment (see Fig. 12). When intelligence was paired with economic outcomes, participants who were exposed to noncomplementary representations with victim‐ derogating consequences scored higher in terms of system justification than did participants who were exposed to complementary representations with victim‐enhancing consequences. When intelligence was paired with physical attractiveness (that is, when no trait–outcome connection was assumed), however, participants who were exposed to complementary representations with victim‐enhancing consequences scored higher in terms of system justification than did participants who were exposed to noncomplementary representations with victim‐derogating consequences. These data suggest that the system‐justifying potential of complementary and noncomplementary representations depends on whether people perceive a causal connection between trait and outcome. Victim‐blaming increases system justification only when there is an assumed link between trait and outcome (e.g., between a lack of intelligence and low income). Conversely, victim‐enhancement increases system justification only when there is no
Causal relevance (intelligence/wealth)
No causal relevance (intelligence/attractiveness)
6.50
Diffuse system justification
6.00 5.50 5.00 4.50 4.00 3.50 3.00 Noncomplementary (victim-blaming) condition
Complementary (victim-enhancing) condition
Fig. 12. EVects of exposure to complementary and noncomplementary representations on system justification as a function of causal relevance. (Adapted from Study 2 in Kay et al., 2005.)
340
AARON C. KAY et al.
perceived link between trait and outcome (e.g., between degrees of intelligence and physical attractiveness). The notion that victim‐blaming and victim‐enhancement are alternate routes to system justification was further explored by Kay et al. (2005) using a system threat experimental paradigm. 2. EVects of System Threat on Victim‐Derogation and Victim‐Enhancement We know from years of research that threatening the self‐concept stimulates the need to engage in ego‐defensive processing and aYrming the integrity of the self‐concept reduces ego‐defensiveness (Fein & Spencer, 1997; Sherman & Cohen, 2002; Steele, 1988). Similarly, it seems that threatening the social system stimulates the need to engage in system‐defensive processing and that aYrming the integrity of the system reduces such defensiveness (Jost & Hunyady, 2002; Jost et al., 2005). Both patterns are consistent with recent goal systems research demonstrating that attaining a desired end‐state (e.g., high self‐esteem, high system legitimacy) greatly diminishes the degree of motivated processing generally used to achieve that end‐state (Forster, Liberman, & Higgins, 2005; see also Jost, Pietrzak, et al., in press). Thus, to the extent that derogating victims on traits that are seen as causally related to status outcomes and enhancing victims on traits that are causally unrelated to outcomes both serve to justify the overarching social system, Kay et al. (2005) hypothesized that threatening the legitimacy of the system should increase both of these tendencies. In two experiments, we employed a system‐threat manipulation that was designed to increase the system justification motive (Kay et al., 2005, Studies 1a and b). Specifically, participants assigned to the high system threat condition read a newspaper article, attributed to a local journalist, which included the following passage: These days, many people in the United States feel disappointed with the nation’s condition. Many citizens feel that the country has reached a low point in terms of social, economic, and political factors. People do not feel as safe and secure as they used to, and there is a sense of uncertainty regarding the country’s future. It seems that many countries in the world are enjoying better social, economic, and political conditions than the U.S.
Participants assigned to the low system threat condition read a newspaper article that included the following passage instead: These days, despite the diYculties the nation is facing, many people in United States feel safer and more secure relative to the past. Many citizens feel that the country is relatively stable in terms of social, economic, and security factors. There is a sense of optimism regarding America’s future and an understanding that this is the only place
SYSTEM JUSTIFICATION THEORY
341
where American people can feel secure. It seems that compared with many countries in the world the social, economic and political conditions in the U.S. are relatively good.
Participants were asked to read the passage as many times as necessary to become familiar with it, and they expected to answer questions about it later in the session. Afterward, they completed ratings of the powerless and the obese, so that we could measure their tendencies to derogate versus enhance these groups on traits that had been pretested as either causally relevant or irrelevant to power and obesity. As hypothesized, system threat led people to exhibit increased: (1) victim‐ blaming on traits that were seen as causally related to status outcomes and (2) victim‐enhancement on traits that were seen as causally unrelated to these outcomes (see Fig. 13A and B). That is, people who had read the system‐ threatening (versus system‐aYrming) newspaper article were more likely to derogate ‘‘losers’’ on traits viewed as casually relevant to their plight, so that the powerless were judged to be less intelligent and the obese were judged to be lazier. They were also more likely to elevate these same ‘‘losers’’ on traits that were viewed as causally irrelevant to their plight, so that the powerless were judged as happier and the obese as more sociable. A noteworthy feature of these experiments is that all participants made ratings on causally relevant and causally irrelevant traits, which means that high (versus low) system threat led people to engage simultaneously in both complementary and noncomplementary means of restoring legitimacy to the system (see also Napier et al., 2006). B. POLITICAL ORIENTATION There is good reason to believe that the system‐justifying potential of complementary and noncomplementary representations will vary as a function of other variables in addition to perceptions of causality. A likely candidate is political orientation. Studies show that conservatives tend to explain social problems, in areas such as crime, poverty, and obesity, using models of personal responsibility and control (Carroll, Perkowitz, Lurigio, & Weaver, 1987; Crandall, 1994; Skitka, Mullen, GriYn, Hutchinson, & Chamberlin, 2002; Sniderman, Hagen, Tetlock, & Brady, 1986). With respect to poverty, for instance, conservatives are more likely than liberals to make internal attributions, seeing it as caused by a lack of intelligence or competence (Pellegrini, Queirolo, Monarrez, & Valenzuela, 1997; Skitka et al., 2002; Zucker & Weiner, 1993). Thus, it is reasonable to assume that they will rely more heavily on the noncomplementary, victim‐blaming route to system justification.
342
AARON C. KAY et al.
A
Causal relevance (independence/ intelligence)
No causal relevance (happiness)
8.00
Evaluation of the powerful
7.00 6.00 5.00 4.00 3.00 2.00 Low system threat
B
Causal relevance (laziness)
High system threat
No causal relevance (sociability)
8.00
Evaluation of the obese
7.00 6.00 5.00 4.00 3.00 2.00 Low system threat
High system threat
Fig. 13. (A) EVects of system threat manipulation on ratings of the powerful for causally relevant and irrelevant traits. (Adapted from Study 1a in Kay et al., 2005.) Note: Ratings were made on nine‐point scales, with higher numbers indicating that powerful people were judged to be more independent, intelligent, and happy (and powerless people judged to be less independent,
SYSTEM JUSTIFICATION THEORY
343
Liberals, on the other hand, may be more likely to pursue the complementary, victim‐enhancing route. For example, liberals are more likely than conservatives to acknowledge the role of contextual factors in creating and sustaining poverty (Bryan, Dweck, Ross, Kay, & Mislavksy, 2006; Skitka et al., 2002), they are less likely to believe that people have control over their circumstances (Zucker & Weiner, 1993), and they are more likely to endorse structural remedies (Cozzarelli, Wilkinson, & Tagler, 2001; Glaser, 2005). For all of these reasons, liberals may eschew victim‐blaming responses and instead rely on other ‘‘strategies’’ to maintain their faith in the legitimacy of the social system. To the extent that they are less likely to accept deservingness‐based justifications for economic inequality, they may be more accepting of complementary, victim‐enhancing justifications (see also Feather, 1999). We investigated the possibility that political orientation would moderate the eVects of exposure to complementary (i.e., victim‐enhancing) and noncomplementary (i.e., victim‐derogating) stereotypes on perceptions of system legitimacy in two studies conducted in Poland. In the first of these (Kay et al., 2007a, Study 1), participants were exposed to either a complementary representation in which a poor friend was described as happier than his rich friend or a noncomplementary representation in which a rich friend was described as happier than his poor friend. [These materials were modified slightly from those used in studies by Kay and Jost (2003); they were translated into Polish by Szymon Czaplinski.] This manipulation was crossed with self‐identified political orientation (leftists and centrists versus rightists), creating a 2 2 factorial design in which the two ideological groups were exposed to either complementary or noncomplementary representations of the rich and poor. Afterward, all participants completed a Polish version of the same diVuse system justification scale used in prior studies. As can be seen in Fig. 14, the complementary (victim‐enhancing) and noncomplementary (victim‐derogating) exemplars had opposite eVects on explicit system justification for leftists/centrists and rightists. For leftists and centrists, system justification was significantly higher following exposure to the complementary, ‘‘poor but happy/rich but miserable’’ vignette compared to the noncomplementary condition. For rightists, by contrast, system
intelligent, and happy). (B) EVects of system threat manipulation on ratings of the obese for causally relevant and irrelevant traits. (Adapted from Study 1b in Kay et al., 2005.) Note: Ratings were made on nine‐point scales, with higher numbers indicating that overweight people were judged to be more sociable and less lazy (and normal weight people judged to be less sociable and more lazy).
344
AARON C. KAY et al.
Complementary
Noncomplementary
Diffuse system justification
4.00
3.50
3.00
2.50
2.00 Not right-wing
Right-wing
Fig. 14. EVects of exposure to complementary and noncomplementary representations on system justification as a function of political ideology. (Adapted from Kay et al., 2007a.)
justification was higher following exposure to the noncomplementary, ‘‘poor and unhappy/rich and happy’’ vignettes. In a follow‐up experiment, we substituted honesty for happiness and examined the eVects of the ‘‘poor but honest/rich but dishonest’’ (complementary) and ‘‘poor and dishonest/rich and honest’’ (noncomplementary) stereotypes on system justification (Kay et al., 2007a, Study 2). As before, results generally supported our hypothesis. For right‐wingers, the noncomplementary (victim‐derogating) representation proved more eVective at justifying the system than did the complementary (victim‐enhancing) representation. For leftists/centrists, although the means were in the predicted direction (i.e., system justification was higher following exposure to the complementary representation), the diVerence did not reach statistical significance. Thus, self‐reported political orientation does appear to moderate the system‐justifying potential of complementary, victim‐enhancing and noncomplementary, victim‐derogating representations of the rich and poor. Whereas complementary, victim‐enhancing stereotypes proved most eVective at justifying the system for liberal and moderate participants, these same stereotypes proved least eVective for conservative participants. As hypothesized, conservatives scored consistently higher on system justification following exposure to noncomplementary, victim‐blaming stereotypical representations.
SYSTEM JUSTIFICATION THEORY
345
C. PROTESTANT WORK ETHIC The PWE is one of several system‐justifying belief systems, insofar as it fosters personal commitment to the capitalist system (see also Jost & Hunyady, 2005). It refers to the assumption that hard work is a moral prerequisite and that it will be rewarded in the long run, in heaven if not on earth (Jones, 1997; Mirels & Garrett, 1971). Studies show that people who score high on individual diVerence measures of the PWE are more likely to exhibit victim‐blaming tendencies and to believe that the lack of success is due to laziness and poor self‐control (Biernat, Vescio, & Theno, 1996; Crandall, 1994; Katz & Hass, 1988; Quinn & Crocker, 1999). Thus, people who embrace the PWE, like political conservatives, may be more prone to rationalizing economic inequality through the noncomplementary, victim‐derogating route than the complementary, victim‐enhancing route. In the context of ‘‘poor but happy’’ complementary stereotypes, this is indeed what we observed (Kay & Jost, 2003, Study 3). Whereas participants who scored low on Quinn and Crocker’s (1999) version of the PWE scale viewed the system as more legitimate following exposure to complementary (versus noncomplementary) representations, participants who scored high on the PWE scale were unmoved by the complementary stereotypes (see Fig. 15). (There was also a tendency for people who scored high on the PWE scale to endorse the system justification items more strongly.) Thus, at least
Complementary
Noncomplementary
Diffuse system justification
6.00 5.50 5.00 4.50 4.00 3.50 3.00 Low PWE
High PWE
Fig. 15. EVects of PWE of exposure to complementary versus noncomplementary representations on system justification as a function of PWE. (Adapted from Study 3 in Kay & Jost, 2003.)
346
AARON C. KAY et al.
in the context of the poor but happy complementary stereotype, the eVects of PWE are similar to those of political conservatism. We also examined the moderating role of PWE in the context of the ‘‘poor but honest’’ complementary stereotype, but in this case the pattern did not replicate (Kay & Jost, 2003, Study 4). Rather, low PWE scorers showed no appreciable diVerence between complementary and noncomplementary conditions, whereas high PWE scorers were more supportive of the societal status quo following exposure to ‘‘poor but honest’’ and ‘‘rich but dishonest’’ complementary representations. Thus, it appears that low PWE scorers are more aVected by ‘‘poor but happy’’ stereotype exemplars, whereas high PWE scorers are more aVected by ‘‘poor but honest’’ stereotype exemplars. This is broadly consistent with some accounts of the ideological function of the PWE, which is said to provide ‘‘a moral justification for the accumulation of wealth’’ but also to encourage people to ‘‘eschew immoderate consumption and participation in worldly pleasures’’ (Mirels & Garrett, 1971, p. 40).
V. Implicit Complementary Versus Noncomplementary Stereotypical Associations There are questions concerning the overall prevalence of complementary versus noncomplementary stereotypical representations that we have not yet addressed. In the case of gender stereotypes, it certainly does seem that complementary stereotypes of women as communal but not agentic and men as agentic but not communal reflect cultural ‘‘default’’ stereotypes (Eagly et al., 1991; Glick & Fiske, 2001; Jost & Kay, 2005). However, the most prevalent economic and racial stereotypes may well be noncomplementary (and victim‐derogating) rather than complementary (and victim‐enhancing). A pilot study by Kay and Jost (2003) found, for instance, that at an explicit, conscious level participants believed that poor people are more honest than rich people (a complementary stereotype) but also less happy than rich people (a noncomplementary stereotype). Stereotypes of African‐Americans and other racial minority groups do seem to possess some favorable characteristics (e.g., athletic, religious, musical), but also a preponderance of unfavorable characteristics (e.g., aggressive, uneducated, loud, hostile, and so on; see Devine, 1989; Judd, Park, Ryan, Braver, & Kraus, 1995; Katz & Hass, 1988; Stangor, Sechrist, & Jost, 2001). The fact that people possess implicit, automatic stereotypes raises another interesting question concerning the extent to which complementary stereotypes have permeated our unconscious (as well as conscious) minds. Do implicit associations primarily reflect noncomplementary (victim‐derogating) or complementary (victim‐enhancing) stereotypes?
SYSTEM JUSTIFICATION THEORY
347
A. IMPLICIT ASSOCIATIONS CONCERNING RICH/POOR AND HAPPINESS/HONESTY Issues of implicit complementary versus noncomplementary stereotyping were addressed by Sherman, Petrocelli, Johnson, and Jost (2005). Specifically, they conducted IAT experiments in order to gauge the extent to which ‘‘poor but happy’’ and ‘‘poor but honest’’ stereotypical associations are held unconsciously. They also examined ideological and other correlates of complementary versus noncomplementary associations, which allows us to better understand explicit moderators of the associations held at an implicit level of awareness (see also Jost et al., 2004). In one set of experiments, Sherman et al. (2005) constructed IATs to measure both rich/poor þ happy/unhappy and rich/poor þ honest/dishonest associations (N ¼ 56 Indiana University students). There were eight stimuli terms related to the concept ‘‘rich’’ (rich, wealthy, aZuent, well‐oV, prosperous, fortune, millionaire, and well‐to‐do) and eight terms related to ‘‘poor’’ (poor, underprivileged, deprived, needy, broke, poverty, bankrupt, and penniless). There were also eight stimuli terms related to the concept ‘‘happy’’ (happy, content, cheerful, pleased, joyful, delighted, elated, and blissful) and eight terms related to ‘‘unhappy’’ (unhappy, discontented, gloomy, sad, depressed, miserable, despondent, and cheerless). Finally, there were eight stimuli terms related to the concept ‘‘honest’’ (honest, truthful, sincere, candid, genuine, trustworthy, frank, and honorable) and eight terms related to ‘‘dishonest’’ (dishonest, deceitful, liar, untruthful, fraudulent, ingenuine, cheater, and corrupt). Results indicated that participants did indeed respond more quickly to complementary pairings of rich/dishonest and poor/honest (M ¼ 951.78 ms) than to their noncomplementary counterparts (M ¼ 1542.07 ms; t[25] ¼ 9.34, p < .001). With regard to happiness, however, the opposite pattern was the dominant one: participants were significantly faster at pairing the noncomplementary rich/happy and poor/unhappy concepts together (M ¼ 865.36 ms) than the complementary rich/unhappy and poor/happy concepts (M ¼ 1478.66 ms; t[29] ¼ 11.16, p < .001). These findings mirror the results of Kay and Jost’s (2003) pilot test concerning explicit attitudes. At both implicit and explicit levels of awareness, people appear to endorse the ‘‘poor are more honest than rich’’ complementary stereotype, but they also endorse the (noncomplementary) notion that the ‘‘rich are happier than the poor.’’ It is worth pointing out that the latter association, at least, is consistent with the available evidence concerning the actual relationship between income and subjective well‐being (Diener & Diener, 1995; Ryan & Deci, 2000). In a follow‐up set of experiments (N ¼ 251 Indiana University students), Sherman et al. (2005) employed IATs in which the first names of individuals
348
AARON C. KAY et al.
were used as stimuli for the categories of rich and poor (e.g., Winthrop, Carlton, Trevor, Gavin, and Parker versus Billy Bob, Spike, Freddy, Bubba, and Darryl). Under these conditions, the tendency for people to see the poor as more honest disappeared. In fact, participants were faster to associate ‘‘poor’’ names with dishonesty and ‘‘rich’’ names with honesty (M ¼ 698.55 ms) than the reverse (M ¼ 831.67 ms; t[125] ¼ 9.21, p < .001). They were also faster to associate these same ‘‘poor’’ names with unhappiness and ‘‘rich’’ names with happiness (M ¼ 680.13 ms) than vice versa (M ¼ 844.79 ms; t[124] ¼ 10.11, p < .001). These results, taken as a whole, are consistent with the notion that people have diVerent reactions toward general (group) and specific (individual) cases. That is, diVerent cognitive processes seem to operate when people encounter and store general versus specific case information (Sherman, Beike, & Ryalls, 1999). This suggests that, although the poor in general might be stereotyped at both implicit and explicit levels of awareness as more honest than the rich, any individual poor person may not be seen as more honest than others. Once personalization enters, it appears that individual targets may be evaluated quite diVerently than the group as a whole (Jenni & Loewenstein, 1997; Small & Loewenstein, 2003).
B. IMPLICIT ASSOCIATIONS CONCERNING WHITE/BLACK AND HAPPINESS/HONESTY Sherman et al. (2005) also conducted a series of race‐based IATs (N ¼ 94 Indiana University students) in which first names that had been pretested as either stereotypical of Black males (Jamal, Theo, Darnell, Leroy, Tyrone, Lavon, Marcellus, and Wardell) or White males (Adam, Chip, Josh, Matthew, Brad, Greg, Paul, and Todd) were paired with the same honest/ dishonest and happy/unhappy stimuli used in the earlier study. Under these circumstances, it appears that the noncomplementary, victim‐derogating associations dominated any trace of complementary, victim‐enhancing associations. That is, participants were faster at pairing Black names with dishonesty and White names with honesty (M ¼ 822 ms) than the reverse pairing (M¼ 1231 ms; t[30] ¼ 7.31, p < .001), and they were faster at pairing Black names with unhappiness and White names with happiness (M ¼ 920.64 ms) than the reverse (M ¼ 1122.42 ms; t[30] ¼ 6.57, p < .001). An additional IAT revealed that participants were also more likely to associate Black names with ‘‘poor’’ words and White names with ‘‘rich’’ words (M ¼ 852.89 ms) than vice versa (M ¼ 988.67 ms, t[31] ¼ 5.04, p < .001). Two more race‐based IAT experiments conducted by Mandisodza, Jost, and Kenyon (2006)—in which photographs of Blacks and Whites were
SYSTEM JUSTIFICATION THEORY
349
paired with words related to honesty/dishonesty and happiness/unhappiness—revealed the same overall pattern of results as in the Sherman et al. (2005) study using Black and White names. In one study, participants (N ¼ 64 NYU undergraduates) were faster at pairing photographs of Black individuals with negative characteristics (i.e., dishonesty and unhappiness) and photographs of White individuals with positive characteristics (M ¼ 844.87 ms) than the reverse (M ¼ 986.92 ms; F [1, 63] ¼ 54.62, p < .001). Furthermore, participants who scored higher on the diVuse system justification scale (see items in Table II) were more likely to implicitly associate Blacks with dishonesty and Whites with honesty (r ¼ .34, p < .001) and to associate Blacks with unhappiness and Whites with happiness (r ¼ .38, p < .001), suggesting that the dominant (noncomplementary) associations were themselves system‐justifying in the racial context. In a second study involving 88 NYU undergraduates, we manipulated whether the photographs of Black and White targets were of individuals or groups. Consistent with the notion that people treat general (group) and specific (individual) cases diVerently (Sherman et al., 1999), reaction times to the two types of stimuli diVered. Specifically, participants were faster to ascribe stereotypical associations (whether complementary or noncomplementary) to individuals than groups, F(1, 87) ¼ 6.29, p < .05. Nevertheless, collapsing across individual and group targets, participants were again faster at pairing Black targets with negative traits (dishonesty and unhappiness) and White targets with positive traits (M ¼ 828.71 ms) than the reverse (M ¼ 990.07 ms; F [1, 87] ¼ 113.02, p < .001). The tendency to associate Black targets with dishonesty and White targets with honesty was again correlated with participants’ scores on the diVuse system justification scale (r[63] ¼ .35, p < .01) and also with their scores on Jost and Thompson’s (2000) economic system justification scale (r[63] ¼ .33, p < .01). However, implicit racial associations pertaining to happiness were not consistently correlated with system justification scores in this study. Thus, in the case of implicit racial stereotypes, noncomplementary (or victim‐derogating) associations clearly trump complementary (or victim‐ enhancing) associations, regardless of whether target race is conveyed through the use of names or photographs and whether members are depicted as individuals or as a group. The preponderance of correlational evidence also suggests that these more common victim‐derogating associations in the racial context are positively (rather than negatively) related to system‐ justifying tendencies. It may be that in the case of stereotypes about Blacks, characteristics other than honesty and happiness, such as athleticism, are more likely to be used in a victim‐enhancing manner to create the ‘‘illusion of equality,’’ thereby contributing to the perceived legitimacy of the societal status quo.
350
AARON C. KAY et al.
VI. Concluding Remarks: SJT and Stereotyping as Rationalization The program of research that we have summarized in this chapter leads to the conclusion that, as previous theorists have suggested (Bem & Bem, 1970; Glick & Fiske, 2001; Jackman, 1994; Jost & Banaji, 1994; Lane, 1962; Lerner, 1980), complementary stereotypes do indeed help people to rationalize substantial social and economic inequalities in society. Presumably, this is because such stereotypes serve to create an ‘‘egalitarian veneer,’’ allowing people to tolerate disparities that might otherwise provide strong grounds for complaint or dissatisfaction with the status quo. In several studies, we have provided strong and consistent experimental evidence that exposing people to complementary (or compensatory) gender and status stereotypes increases their faith in the fairness and legitimacy of the overarching social system (Jost & Kay, 2005; Kay & Jost, 2003; Kay et al., 2005). Furthermore, we have shown that when the legitimacy or stability of the system is threatened, people often respond by using complementary stereotypes to bolster the system (Jost et al., 2005; Kay et al., 2005). Although we have focused on stereotyping in this chapter, we know that people engage in rationalization and ‘‘cognitive restructuring’’ much more generally (Festinger, 1957; Walster, Berscheid, & Walster, 1978). The kinds of system‐justifying biases we have identified with respect to complementary stereotypes are probably also at work in many other instances of cognition, perception, and memory (see also Haines & Jost, 2000). For instance, we have begun to find that people seem to rationalize everyday, self‐relevant threats to the justice motive, such as the lucky and unlucky breaks people experience in their everyday lives (e.g., winning or losing a coin flip). In one such study, participants who experienced a good break were subsequently more likely to call to mind previous bad breaks, and participants who were faced with a bad break were more likely to recall good breaks in the past, suggesting that they were motivated to compensate for the fickleness of fate by ‘‘balancing out’’ good and bad breaks over time (Kay et al., 2007a). The system‐justifying potential of complementary (or compensatory) beliefs about the social world may also aVect the dynamics of interpersonal interaction and attraction. For example, Lau, Kay, and Spencer (2007) found that an experimental manipulation of high system threat, which presumably increases the strength of system justification motives, led men to express increased attraction to women who embraced romantic ideals associated with benevolent sexism (e.g., women who described themselves as delicate, fragile, old‐fashioned, and in need of male protection). This suggests that psychological needs to satisfy system justification goals could influence a much wider range of social outcomes, including interpersonal
SYSTEM JUSTIFICATION THEORY
351
attraction, impression management strategies, and the choice of specific types of interaction partners. It is tempting to conclude that Panglossian forms of rationalization, which help us to maintain a favorable, even idealized image of the status quo, are innocuous and even propitious. Perhaps we should be thankful for the gift of system justification. As Gilbert (2005) wrote in an Op‐Ed piece that appeared in the New York Times: Research suggests that human beings have a remarkable ability to manufacture happiness . . .. Our ability to spin gold from the dross of our experience means that we often find ourselves flourishing in circumstances we once dreaded. We fear divorces, natural disasters and financial hardships until they happen, at which point we recognize them as opportunities to reinvent ourselves, to bond with our neighbors and to transcend the spiritual poverty of material excess. When the going gets tough, the mind gets going on a hunt for silver linings, and most linings are suYciently variegated to reward the mind’s quest.
Viewed from this perspective, the human capacity to rationalize suVering, exploitation, injustice, and inequality seems like a godsend. But, as Gilbert goes on to note, there is a price, namely complacency about change: Many of the heroes and redeemers we most admire were unhappy people who found it impossible to change how they felt about the world—which left them no choice but to change the world itself. Outrage, anger, fear and frustration are unpleasant emotions that most of us vanquish through artful reasoning; but unpleasant emotions can also be spurs to action—clamorous urges that we may silence at our peril.
In our research, we have found that many system‐justifying beliefs and ideologies do indeed serve the palliative function of reducing distress and fostering positive aVect and general satisfaction (Jost & Hunyady, 2002; Jost, Pelham, Sheldon, & Sullivan, 2003). As Gilbert (2005) suggests, however, satisfaction with the status quo breeds inaction. Kay, Gaucher, Peach, Spencer, and Zanna (2007b) have found that the system justification motive leads people to reinterpret what ‘‘is’’ (i.e., descriptive norms) as what ‘‘should be’’ (i.e., injunctive norms). For example, learning that the Canadian House of Commons is composed of people in the top 90th percentile of income, or that university professors are typically male, causes participants in experimental settings to actively argue that politicians should be wealthier than the rest, and university professors should be male (an eVect that becomes more pronounced under conditions of high system threat). This system‐justifying bias has obvious consequences for motivations to redress social inequality. Furthermore, we find that system justification undermines moral outrage, which is an important motivator of egalitarian reform and eVorts to help the disadvantaged (see Wakslak, Jost, Tyler, & Chen, 2007).
352
AARON C. KAY et al.
Thus, to the extent that people embrace complementary stereotypes and other appealing notions that allow them to continue believing that they live ‘‘in the best of all possible worlds,’’ they will undoubtedly feel better about the inequalities and even injustices all around them, but they will also be much less likely to do anything to change them. Acknowledgments Some of the research summarized in this chapter was presented at Cornell University, Syracuse University, New York University, the University of Wisconsin, and the University of Waterloo as well as the 2005 Society of Experimental Social Psychology (SESP) conference in San Diego. We are grateful for the many useful comments we received on those occasions. This work was supported in part by a Social Science and Humanities Research Council of Canada Grant to A.C.K., National Science Foundation (NSF) Award # BCS‐0617558 to J.T.J., and a Ford Foundation Fellowship administered through the National Research Council of the National Academies to A.M.
References Allport, G. W. (1954). The nature of prejudice. Cambridge, MA: Addison Wesley. Allport, G. W. (1966). Traits revisited. American Psychologist, 21, 1–10. Ashburn‐Nardo, L., Knowles, M. L., & Monteith, M. J. (2003). Black Americans’ implicit racial associations and their implications for intergroup judgment. Social Cognition, 21, 61–87. Bargh, J. A. (1996). Automaticity in social psychology. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 169–183). New York: Guilford. Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct eVects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71, 230–244. Bargh, J. A., Gollwitzer, P. M., Lee‐Chai, A., Barndollar, K., & Tro¨tschel, R. (2001). The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology, 81, 1014–1027. Bem, S. L., & Bem, D. J. (1970). Case study of a nonconscious ideology: Training the woman to know her place. In D. J. Bem (Ed.), Beliefs, attitudes, and human aVairs (pp. 89–99). Belmont, CA: Brooks/Cole Publishing Co. Bettelheim, B. (1943). Individual and mass behavior in extreme situations. Journal of Abnormal and Social Psychology, 38, 417–452. Biernat, M., Vescio, T. K., & Theno, S. A. (1996). Violating American values: A ‘‘value congruence’’ approach to understanding outgroup attitudes. Journal of Experimental Social Psychology, 32, 387–410. Blasi, G., & Jost, J. T. (2006). System justification theory and research: Implications for law, legal advocacy, and social justice. California Law Review, 94, 1119–1167. Bryan, C., Dweck, C. S., Ross, L., Kay, A. C., & Mislavsky, N. (2007). Ideology as mindset: EVects of personality‐relevant priming on political assessments. Manuscript in preparation. Capozza, D., Bonaldo, E., & DiMaggio, A. (1982). Problems of identity and social conflict: Research on ethnic groups in Italy. In H. Tajfel (Ed.), Social identity and intergroup relations (pp. 299–334). Cambridge, UK: Cambridge University Press.
SYSTEM JUSTIFICATION THEORY
353
Carroll, J., Perkowitz, W., Lurigio, A., & Weaver, F. (1987). Sentencing goals, causal attributions, ideology, and personality. Journal of Personality and Social Psychology, 52, 107–118. Conway, M., Pizzamiglio, M. T., & Mount, L. (1996). Status, communality, and agency: Implications for stereotypes of gender and other groups. Journal of Personality and Social Psychology, 71, 25–38. Cozzarelli, C., Wilkinson, A. V., & Tagler, M. J. (2001). Attitudes toward the poor and attributions for poverty. Journal of Social Issues, 57, 207–228. Crandall, C. S. (1994). Prejudice against fat people: Ideology and self‐interest. Journal of Personality and Social Psychology, 66, 882–894. Crandall, C. S., & Beasley, R. K. (2001). A perceptual theory of legitimacy: Politics, prejudice, social institutions, and moral value. In J. T. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations (pp. 77–102). New York: Cambridge University Press. Dasgupta, N. (2004). Implicit ingroup favoritism, outgroup favoritism, and their bahavioral manifestations. Social Justice Research, 17, 143–169. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5–18. Diener, E., & Diener, M. (1995). Cross‐cultural evidence correlates of life satisfaction and self‐ esteem. Journal of Personality and Social Psychology, 68, 653–663. Dijksterhuis, A., & Bargh, J. A. (2001). The perception‐behavior expressway: Automatic eVects of social perception on social behavior. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 33, pp. 1–40). San Diego, CA: Academic Press. Dollard, J. (1937). Caste and class in a southern town. New Haven, CT: Yale University Press. Eagly, A., & SteVen, V. (1984). Gender stereotypes stem from the distribution of women and men into social roles. Journal of Personality and Social Psychology, 46, 735–754. Eagly, A. H., & Johannesen‐Schmidt, M. L. (2001). Characteristics of women’s leadership: The leadership styles of men and women. Journal of Social Issues, 57, 781–797. Eagly, A. H., & Mladinic, A. (1989). Gender stereotypes and attitudes toward women and men. Personality and Social Psychology Bulletin, 15, 543–558. Eagly, A. H., Mladinic, A., & Otto, S. (1991). Are women evaluated more favorably than men? An analysis of attitudes, beliefs, and emotions. Psychology of Women Quarterly, 15, 203–216. Elkins, S. (1967). Slavery and personality. In R. S. Lazarus & E. M. Opton, Jr. (Eds.), Personality (pp. 395–420). Middlesex, UK: Penguin. Elster, J. (1983). Sour grapes: Studies in the subservience of rationality. Cambridge, UK: Cambridge University Press. Feather, N. T. (1999). Judgments of deservingness: Studies in the psychology of justice and achievement. Personality and Social Psychology Review, 3, 86–107. Fein, S., & Spencer, S. J. (1997). Prejudice as self‐image maintenance: AYrming the self through negative evaluations of others. Journal of Personality and Social Psychology, 73, 31–44. Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 2, pp. 357–411). New York: McGraw‐Hill. Fiske, S. T., Cuddy, A., Glick, P., & Xu, J. (2002). A model of (often mixed) stereotype content: Competence and warmth respectively follow from perceived status and competition. Journal of Personality and Social Psychology, 82, 878–902. Forster, J., Liberman, N., & Higgins, E. T. (2005). Accessibility from active and fulfilled goals. Journal of Experimental Social Psychology, 41, 220–239.
354
AARON C. KAY et al.
Furnham, A., & Gunter, B. (1984). Just world beliefs and attitudes towards the poor. British Journal of Social Psychology, 23, 265–269. Gilbert, D. T. (2005, January 20). Four more years of happiness. New York Times, p. A23. Retrieved October 29, 2006 from http://select.nytimes.com/gst/abstract.html?res¼ F50E10FF3B5C0C738EDDA80894DD404482 Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. (1998). Immune neglect: A source of durability bias in aVective forecasting. Journal of Personality and Social Psychology, 75, 617–638. Glaser, J. (2005). Intergroup bias and inequity: Legitimizing beliefs and policy attitudes. Social Justice Research, 18, 257–282. Glick, P., & Fiske, S. T. (2001). An ambivalent alliance: Hostile and benevolent sexism as complementary justifications for gender inequality. American Psychologist, 56, 109–118. Hafer, C. L. (2000). Do innocent victims threaten the belief in a just world? Evidence from a modified Stroop task. Journal of Personality and Social Psychology, 79, 165–173. Hafer, C. L., & Begue, L. (2005). Experimental research on just‐world theory: Problems, developments, and future challenges. Psychological Bulletin, 131, 128–167. Haines, E. L., & Jost, J. T. (2000). Placating the powerless: EVects of legitimate and illegitimate explanation on aVect, memory, stereotyping. Social Justice Research, 13, 219–236. Hamilton, D. L., & Sherman, J. W. (1994). Stereotypes. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Handbook of social cognition, Vol. 1: Basic processes (2nd ed., pp. 1–68). Hillsdale, NJ: Erlbaum. Heider, F. (1958). The psychology of interpersonal relations. Oxford, England: Wiley. Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability, and salience. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 133–168). New York: Guilford. HoVman, C., & Hurst, N. (1990). Gender stereotypes: Perception or rationalization? Journal of Personality and Social Psychology, 58, 197–208. Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of intergroup relations and group processes. London: Routledge. Hughes, L. (1933/1971). Slave on the block. In L. Hughes (Ed.), The ways of white folks: Stories by Langston Hughes. New York: Vintage. Hunyady, Gy. (1998). Stereotypes during the decline and fall of communism. New York: Routledge. Jackman, M. R. (1994). The velvet glove: Paternalism and conflict in gender, class, and race relations. Berkeley, CA: University of California Press. Jackman, M. R., & Senter, M. S. (1983). DiVerent, therefore unequal: Beliefs about trait diVerences between groups of unequal status. Research in Social Stratification and Mobility, 2, 309–335. JanoV‐Bulman, R., & Yopyk, D. (2004). Random outcomes and valued commitments: Existential dilemmas and the paradox of meaning. In J. Greenberg, S. L. Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology. New York: Guilford. Jenni, K., & Loewenstein, G. (1997). Explaining the identifiable victim eVect. Journal of Risk and Uncertainty, 14, 235–257. Jolls, C., & Sunstein, C. R. (2006). Debiasing through law. The Journal of Legal Studies, 35, 199–242. Jones, H. B. (1997). The Protestant ethic: Weber’s model and the empirical literature. Human Relations, 50, 757–778. Jost, J. T. (2001). Outgroup favoritism and the theory of system justification: An experimental paradigm for investigating the eVects of socio‐economic success on stereotype content.
SYSTEM JUSTIFICATION THEORY
355
In G. Moskowitz (Ed.), Cognitive social psychology: The Princeton symposium on the legacy and future of social cognition (pp. 89–102). Mahwah, NJ: Erlbaum. Jost, J. T., & Banaji, M. (1994). The role of stereotyping in system justification and the production of false consciousness. British Journal of Social Psychology, 22, 1–27. Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system justification theory: Accumulated evidence of conscious and unconscious bolstering of the status quo. Political Psychology, 25, 881–919. Jost, J. T., & Burgess, D. (2000). Attitudinal ambivalence and the conflict between group and system justification motives in low status groups. Personality and Social Psychology Bulletin, 26, 293–305. Jost, J. T., Burgess, D., & Mosso, C. (2001). Conflicts of legitimation among self, group, and system: The integrative potential of system justification theory. In J. T. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations (pp. 363–390). New York: Cambridge University Press. Jost, J. T., & Hamilton, D. L. (2005). Stereotypes in our culture. In J. Dovidio, P. Glick, & L. Rudman (Eds.), On the nature of prejudice: Fifty years after Allport (pp. 208–224). Oxford: Blackwell. Jost, J. T., & Hunyady, O. (2002). The psychology of system justification and the palliative function of ideology. European Review of Social Psychology, 13, 111–153. Jost, J. T., & Hunyady, O. (2005). Antecedents and consequences of system‐justifying ideologies. Current Directions in Psychological Science, 14, 260–265. Jost, J. T., & Kay, A. C. (2005). Exposure to benevolent sexism and complementary gender stereotypes: Consequences for specific and diVuse forms of system justification. Journal of Personality and Social Psychology, 88, 498–509. Jost, J. T., Kivetz, Y., Rubini, M., Guermandi, G., & Mosso, C. (2005). System‐justifying functions of complementary regional and ethnic stereotypes: Cross‐national evidence. Social Justice Research, 18, 305–333. Jost, J. T., & Major, B. (Eds.) (2001). The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations. New York: Cambridge University Press. Jost, J. T., Pelham, B. W., & Carvallo, M. R. (2002). Non‐conscious forms of system justification: Cognitive, aVective, and behavioral preferences for higher status groups. Journal of Experimental Social Psychology, 38, 586–602. Jost, J. T., Pelham, B. W., Sheldon, O., & Sullivan, B. N. (2003). Social inequality and the reduction of ideological dissonance on behalf of the system: Evidence of enhanced system justification among the disadvantaged. European Journal of Social Psychology, 33, 13–36. Jost, J. T., Pietrzak, J., Liviatan, I., Mandisodza, A., & Napier, J. (in press). System justification as conscious and nonconscious goal pursuit. In J. Shah & W. Gardner (Eds.), Handbook of motivation science. New York: Guilford. Jost, J. T., & Thompson, E. P. (2000). Group‐based dominance and opposition to equality as independent predictors of self‐esteem, ethnocentrism, and social policy attitudes among African Americans and European Americans. Journal of Experimental Social Psychology, 36, 209–232. Judd, C. M., James‐Hawkins, L., Yzerbyt, V. Y., & Kashima, Y. (2005). Fundamental dimensions of social judgment: Understanding the relations between competence and warmth. Journal of Personality and Social Psychology, 89, 899–913. Judd, C. M., Park, B., Ryan, C. S., Braver, M., & Kraus, S. (1995). Stereotypes and ethnocentrism: Diverging interethnic perceptions of African Americans and White American youth. Journal of Personality and Social Psychology, 69, 460–481.
356
AARON C. KAY et al.
Kaiser, C. R., Dyrenforth, P. S., & Hagiwara, N. (2006). Why are attributions to discrimination interpersonally costly? A test of system‐ and group‐justifying motivations. Personality and Social Psychology Bulletin, 32, 1523–1536. Katz, I., & Hass, R. G. (1988). Racial ambivalence and American value conflict: Correlational and priming studies of dual cognitive structures. Journal of Personality and Social Psychology, 55, 893–905. Kay, A. C., Czaplinski, S., & Jost, J. T. (2007a). System‐justifying eVects of victim‐derogation and victim‐enhancement: The moderating role of political ideology. Manuscript submitted for publication. Kay, A. C., Gaucher, D. M., Peach, J. M., Spencer, S. J., & Zanna, M. P. (2007b). System justification and the shift from descriptive to injunctive norms. Unpublished data. University of Waterloo. Kay, A. C., Jimenez, M. C., & Jost, J. T. (2002). Sour grapes, sweet lemons and the anticipatory rationalization of the status quo. Personality and Social Psychology Bulletin, 28, 1300–1312. Kay, A. C., & Jost, J. T. (2003). Complementary justice: EVects of ‘‘poor but happy’’ and ‘‘poor but honest’’ stereotype exemplars on system justification and implicit activation of the justice motive. Journal of Personality and Social Psychology, 85, 823–837. Kay, A. C., Jost, J. T., & Young, S. (2005). Victim‐derogation and victim‐enhancement as alternate routes to system‐justification. Psychological Science, 16, 240–246. Kinder, D. R., & Sears, D. O. (1985). Public opinion and political action. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology (3rd ed., Vol. 2, pp. 659– 741). New York: Random House. Kluegel, J. R., & Smith, E. R. (1986). Beliefs about inequality: Americans’ views of what is and what ought to be. New York: Aldine de Gruyter. Kruglanski, A. W. (1996). Motivated social cognition: Principles of the interface. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp. 493–520). New York: Guilford Press. Kruglanski, A. W. (2004). The psychology of closed mindedness. New York: Psychology Press. Lane, R. E. (1962). Political ideology: Why the American common man believes what he does. New York: Free Press. Langer, E. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311–328. Lau, G. P., Kay, A. C., & Spencer, S. J. (2007). EVects of system justification and benevolent sexism on interpersonal attraction. Manuscript in preparation. Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York: Plenum. Lewin, K. (1941/1948). Self‐hatred among Jews. In K. Lewin (Ed.), Resolving social conflicts: Selected papers on group dynamics. New York: Harper & Brothers. Lipkus, I. M. (1991). The construction and preliminary validation of a global belief in a just world scale and the exploratory analysis of the multidimensional belief in a just world scale. Personality and Individual DiVerences, 12, 1171–1178. Macrae, C. N., & Bodenhausen, G. V. (2000). Social cognition: Thinking categorically about others. Annual Review of Psychology, 51, 93–120. Major, B. (1994). From social inequality to personal entitlement: The role of social comparisons, legitimacy appraisals, and group memberships. In M. Zanna (Ed.), Advances in experimental social psychology (Vol. 26, pp. 293–355). New York: Academic Press. Mandisodza, A. N., Jost, J. T., & Kenyon, G. K. (2006). Complementary stereotypes: An implicit approach. Unpublished data. New York University. Mandisodza, A. N., Jost, J. T., & Unzueta, M. M. (2006). ‘‘Tall Poppies’’ and ‘‘American Dreams’’: Reactions to rich and poor in Australia and the United States. Journal of Cross‐ Cultural Psychology, 37, 659–668.
SYSTEM JUSTIFICATION THEORY
357
McGuire, W. J. (1960). A syllogistic analysis of cognitive relationships. In M. J Rosenberg, C. I. Hovland, W. J. McGuire, R. P. Abelson, & J. W. Brehm (Eds.), Attitude organization and change (pp. 65–111). New Haven: Yale University Press. McGuire, W. J., & McGuire, C. V. (1991). The content, structure, and operation of thought systems. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Advances in social cognition (Vol. IV, pp. 1–78). Hillsdale, NJ: Erlbaum. Mirels, H. L., & Garrett, J. B. (1971). The Protestant ethic as a personality variable. Journal of Consulting and Clinical Psychology, 36, 40–44. Montada, L., & Schneider, A. (1989). Justice and emotional reactions to the disadvantaged. Social Justice Research, 3, 313–344. Moore, B., Jr. (1978). Injustice: The social bases of obedience and revolt. White Plains, NY: Sharpe. Napier, J. L., Mandisodza A. N., Andersen, S. M., Jost J. T. (2006). System justification in responding to the poor and displaced in the aftermath of Hurricane Katrina. Analyses of Social Issues and Public Policy, 6, 57–73. O’Brien, L. T., & Major, B. (2005). System‐justifying beliefs and psychological well‐being: The roles of group status and identity. Personality and Social Psychology Bulletin, 31, 1718–1729. Pellegrini, R. J., Queirolo, S. S., Monarrez, V. E., & Valenzuela, D. M. (1997). Political identification and perceptions of homelessness: Attributed causality and attitudes on public policy. Psychological Reports, 80, 1139–1148. Plaks, J. E., Grant, H., & Dweck, C. S. (2005). Violations of implicit theories and the sense of prediction and control: Implications for motivated person perception. Journal of Personality and Social Psychology, 88, 245–262. Prentice, D. A., & Carranza, E. (2002). What women should be, shouldn’t be, are allowed to be, and don’t have to be: The contents of prescriptive gender stereotypes. Psychology of Women Quarterly, 26, 269–281. Quinn, D. M., & Crocker, J. (1999). When ideology hurts: EVects of belief in the Protestant ethic and feeling overweight on the psychological well‐being of women. Journal of Personality and Social Psychology, 77, 402–414. Ridgeway, C. L. (2001). The emergence of status beliefs: From structural inequality to legitimizing ideology. In J. T. Jost & B. Major (Eds.), The psychology of legitimacy: Emerging perspectives on ideology, justice, and intergroup relations (pp. 257–277). New York: Cambridge University Press. Rudman, L. A. (2005). Rejection of women? Beyond prejudice and antipathy. In J. F. Dovidio, P. Glick, & L. A. Rudman (Eds.), On the nature of prejudice: Fifty years after Allport (pp. 106–120). Malden, MA: Blackwell Publishing. Rudman, L. A., & Fairchild, K. (2004). Reactions of counterstereotypic behavior: The role of backlash in cultural stereotype maintenance. Journal of Personality and Social Psychology, 87, 157–176. Rudman, L. A., Feinberg, J., & Fairchild, K. (2004). Minority members’ implicit attitudes: Automatic ingroup bias as a function of group status. Social Cognition, 20, 294–320. Rudman, L. A., & Glick, P. (1999). Feminized management and backlash toward agentic women: The hidden costs to women of a kinder, gentler image of middle managers. Journal of Personality and Social Psychology, 77, 1004–1010. Russell, B. (1950). The superior virtue of the oppressed. In B. Russell (Ed.), Unpopular essays (pp. 58–64). New York: Simon & Schuster. Ryan, R. M., & Deci, E. L. (2000). On happiness and human potentials: A review of research on hedonic and eudaimonic well‐being. Annual Review of Psychology, 52, 141–166. Ryan, W. (1976). Blaming the victim. New York, NY: Vintage Books. Sharpe, R. (2000, November 20). As leaders, women rule. Business Week, p. 74.
358
AARON C. KAY et al.
Sherman, D. K., & Cohen, G. L. (2002). Accepting threatening information: Self‐aYrmation and the reduction of defensive biases. Current Directions in Psychological Science, 11, 119–123. Sherman, S. J. (1991). Thought systems for the past as well as for the future. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Advances in social cognition (Vol. 4, pp. 173–195). Hillsdale, NJ: Lawrence Erlbaum. Sherman, S. J., Beike, D. R., & Ryalls, K. R. (1999). Dual‐processing accounts of inconsistencies in responses to general versus specific cases. In S. Chaiken & Y. Trope (Eds.), Dual‐ process theories in social psychology (pp. 203–227). New York: Guilford Press. Sherman, S. J., Petrocelli, J. V., Johnson, A. L., & Jost, J. T. (2005, September). The poor are honest, but people who are poor are not: System justification and implicit associations of general and specific cases of rich and poor. Talk presented at the Annual Conference of the Society of Experimental Social Psychology (SESP). San Diego, CA. Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social hierarchy and oppression. New York: Cambridge University Press. Skitka, L. J., Mullen, E., GriYn, T., Hutchinson, S., & Chamberlin, B. (2002). Dispositions, scripts, or motivated correction? Understanding ideological diVerences in explanations for social problems. Journal of Personality and Social Psychology, 83, 470–487. Small, D. A., & Loewenstein, G. (2003). Helping a victim or helping the victim: Altruism and identifiabilty. Journal of Risk and Uncertainty, 26, 5–16. Sniderman, P. M., Hagen, M. G., Tetlock, P. E., & Brady, H. E. (1986). Reasoning chains: Causal models of policy reasoning in mass publics. British Journal of Political Science, 16, 405–430. Stangor, C., Sechrist, G. B., & Jost, J. T. (2001). Changing racial beliefs by providing consensus information. Personality and Social Psychology Bulletin, 27, 486–496. Steele, C. M. (1988). The psychology of self‐aYrmation: Sustaining the integrity of self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261–302). New York: Academic Press. Tajfel, H. (1981). Human groups and social categories. Cambridge: Cambridge University Press. Tesser, A., Martin, L. L., & Cornell, D. P. (1996). On the substitutability of self‐protective mechanisms. In P. M. Gollwitzer & J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 48–68). New York: Guilford Press. Uhlmann, E., Dasgupta, N., Elgueta, A., Greenwald, A. G., & Swanson, J. (2002). Subgroup prejudice based on skin color among Hispanics in the United States and Latin America. Social Cognition, 20, 198–225. Voltaire (1947). Candide or optimism (J. Butt, Trans.). New York: Penguin Books. Wakslak, C., Jost, J. T., Tyler, T. R., & Chen, E. (2007). Moral outrage mediates the dampening eVect of system justification on support for redistributive social policies. Psychological Science. (In press). Walster, E., Berscheid, E., & Walster, G. W. (1978). Equity: Theory and research. Boston: Allyn & Bacon. Yzerbyt, V., Provost, V., & Corneille, O. (2005). Not competent but warm . . .. really? Compensatory stereotypes in the French‐speaking world. Group Processes & Intergroup Relations, 8, 291–308. Zinn, H. (1968/2002). Disobedience and democracy: Nine fallacies on law and order. Cambridge, MA: South End Press. Zucker, G. S., & Weiner, B. (1993). Conservatism and perceptions of poverty: An attributional analysis. Journal of Applied Social Psychology, 23, 925–943.
FEELING THE ANGUISH OF OTHERS: A THEORY OF VICARIOUS DISSONANCE
Joel Cooper Michael A. Hogg
Fifty years ago, Leon Festinger introduced a new concept into the language of social psychology. In his A Theory of Cognitive Dissonance, Festinger (1957) explained that two elements of knowledge that are discrepant with each other could cause a state of arousal that he called cognitive dissonance. Dissonance, experienced as an unpleasant and uncomfortable tension state akin to a drive, needs to be reduced. An individual who is in such a state is motivated to add, subtract, or otherwise alter his or her cognitions in order to be rid of the unpleasant feeling of dissonance. From these basic propositions, the theory has spawned controversy, support, and an ample share of criticism. Over the decades, the culmination of thousands of empirical studies and theoretical refinements is testimony to the impact and significance of cognitive dissonance on individual agents and actors. In this chapter, we suggest that Festinger’s focus on the experience of the individual whose cognitions are discrepant was important, but narrow. We suggest that the reach of cognitive dissonance can be far more sweeping than Festinger had imagined. Under the appropriate circumstances, dissonance can be experienced not only by an actor but also by an audience that observes the actor’s attitude‐discrepant behavior. Moreover, the eVect of dissonance on the audience can be substantially similar to the eVect on an individual, leading to changes of attitudes, thoughts, and behaviors. We call this phenomenon vicarious cognitive dissonance. In this chapter, we argue for the plausibility of our theoretical notion that dissonance can be experienced vicariously. We present data to support the presence of vicarious dissonance and examine the circumstances in which it arises. We then argue that the state of vicarious dissonance provides a new 359 ADVANCES IN EXPERIMENTAL SOCIAL PSYCHOLOGY, VOL. 39 DOI: 10.1016/S0065-2601(06)39007-7
Copyright 2007, Elsevier Inc. All rights reserved. 0065-2601/07 $35.00
360
JOEL COOPER AND MICHAEL A. HOGG
way to translate research in basic social processes to areas of physical and mental health as well helping to sharpen an analysis of the interplay between the experience of dissonance and culture.
I. Introduction Imagine the following scenario: You are a university professor and have been at your university for several years. So, too, are the people gathered with you at the lunch table. All of you like your university and think it is a very good place to work. Sometimes, there are issues that need attending to such as a recent administration proposal to increase student enrollment at the university. You have been discussing this controversial question with your colleagues and you are gratified to know that, like you, all of the people at your table disagree with the plan, fearing it will increase class sizes and teaching loads. But on this particular afternoon, an unusual event occurs. Another colleague walks to your table and asks you all for a favor. Would one of you help to convince a visiting reporter that there is some merit to the proposed new enrollment plan? Your colleague tells you that he is trying to get the reporter to keep an open mind and he would surely appreciate it if one of you would help convince her that a larger enrollment would be good for the college. Inwardly, you are amused for you believe no one will take him up on his request. You know that most of the faculty at your school are generally opposed to the plan and you know for certain that your colleagues at this table are opposed to the plan. Surely, no one is going to volunteer to express contrary points of views, especially to a reporter. That is when the surprise occurs: One of your tablemates, Professor Adams, agrees to do it. Despite what you know to be his opinion on the matter, he acquiesces to the request and agrees to forcefully convince the reporter that the enrollment plan has merit. We know from the many studies conducted in cognitive dissonance theory that Adams will likely experience cognitive dissonance. We know this because he has freely agreed to deliver a statement about the enrollment issue that is discrepant from his true opinion. We know this because the conditions for dissonance seem to be in place. The statement is attitude‐discrepant, he freely chose to make the statement, and the possibility that he might convince the reporter would be the kind of consequence that makes the dissonance experience likely to occur. As social psychologists, if we had administered our famous Likert‐type attitude scales to Prof. Adams prior to this incident, and
A THEORY OF VICARIOUS DISSONANCE
361
administered it again after he agreed, we would expect to find that Adams had changed his attitude in the direction of his attitude‐discrepant behavior. The new question that we pose to you is: How do you feel after witnessing Adams’s decision to convince the reporter? Surely, you were surprised, maybe even a little disappointed. We believe, however, that if the social psychologists had given you an attitude scale, you would have changed your attitude as well. So would the other people sitting at the table. Why? Because you would have experienced Adams’s dissonance vicariously. Like Adams, you would have experienced an unpleasant and uncomfortable tension state and, like Adams, you will have changed your attitude as well.
II. Personal Cognitive Dissonance Almost all of the research in cognitive dissonance has focused on the individual actor who finds him or herself with cognitive elements that are inconsistent. Festinger (1957) explained that the perceived psychological inconsistency between elements gives rise to ‘‘pressures to reduce or eliminate the dissonance’’ (p. 18) and that the pressure was proportional to the magnitude and importance of the inconsistency (relative to the magnitude and importance of supporting or consistent cognitive elements). This relatively straightforward notion inspired and precipitated literally thousands of research studies that supported, criticized, and altered the concept of dissonance. The extraordinary history of dissonance research has been reviewed elsewhere (see Cooper, 2007; Harmon‐Jones & Mills, 1999). SuYce it to say that the major activity in the first‐quarter century of dissonance research was directed toward deriving and confirming some of the nonobvious predictions and illustrations that flowed from the theory. One of those, known as induced compliance, was the basis for our illustration highlighting the dilemma of Professor Adams in the preceding section. Festinger and Carlsmith (1959) demonstrated that making a statement contrary to one’s belief creates attitude change. The basis for their prediction was that one cognitive element— that is, knowledge about one’s attitude—is discrepant from another cognitive element—that is, knowledge about one’s behavior. This pair of cognitions is dissonant and establishes the pressure to reduce it. Changing one’s memory for what one did is diYcult; changing one’s attitude to restore consistency is far more psychologically defensible. Festinger (1957) taught us that people will reduce dissonance by changing the cognition least resistant to change. Thus, induced compliance typically leads to attitude change in the direction implied by the behavioral act.
362
JOEL COOPER AND MICHAEL A. HOGG
What made cognitive dissonance so controversial was the predicted eVect of receiving a monetary reward on the magnitude of cognitive dissonance. Festinger and Carlsmith (1959), in their landmark demonstration of induced compliance, predicted that people who received a large financial inducement to make a statement contrary to their attitude about how exciting a tedious laboratory task had been, would change their attitude as an inverse function of the magnitude of incentive. Their well‐known procedure had participants try to convince a person they believed was another student waiting to participate in the experiment that the experimental task was interesting and exciting, despite the fact that the participants knew the task was dull and boring. Participants who had been promised $20 for their statement changed their attitude toward the experimental task significantly less than participants who had only been promised $1. They argued that the larger and more important the cognition supporting their behavior (in this case, taking home a financial prize), the less the magnitude of dissonance and the less need for participants to change their attitudes. The small inducement was a trivial cognitive element supporting the behavior; the large inducement was an important one. That is why the relationship of attitude change to incentive magnitude should be inverse. There was a variety of situations that produced cognitive dissonance and several research traditions were spawned by the studies of the 1950s and 1960s. Jack Brehm’s (1956) free choice research showed that people distort the attractiveness of choice alternatives following a decision such that the chosen alternative becomes more attractive and the rejected alternative becomes less attractive. Aronson and Mills (1959) began yet another research tradition by showing that people change their attitudes to justify eVort they expend in order to reach a goal. Erlich, Guttman, Schonbach, and Mills (1957) showed that people actively seek information that supports a chosen alternative and avoid information that supports a rejected alternative. And, in a research paradigm that seems like the negative image of induced compliance, Aronson and Carlsmith (1962) showed that children who were asked not to play with an attractive toy devalued that toy as a way of reducing their dissonance, but devalued it less as the magnitude of the threat for playing with it increased. Although many of the research methods or paradigms played significant roles in the maturation and development of cognitive dissonance theory, the research using the method of induced compliance has arguably been the most productive. Partly as an outgrowth of interesting controversies along the way, research in the intervening decades has: supported the prediction that people are motivated to change their attitudes following counter‐attitudinal behavior;
A THEORY OF VICARIOUS DISSONANCE
363
supported Festinger and Carlsmith’s original finding that attitude change is an inverse function of the magnitude of incentive for performing the behavior; concluded that the behavior must be engaged in freely; the appearance of coercion eliminates cognitive dissonance; shown that dissonance seems to depend on potentially producing an unwanted event as a consequence of the behavior; and supported the prediction that cognitive dissonance is a motivational state that can be measured physiologically (Croyle & Cooper, 1983; Elkin & Leippe, 1986; Losch & Cacioppo, 1990) and psychologically (Elliot & Devine, 1994), and can be reduced by misattribution of the arousal (Zanna & Cooper, 1974, 1976). Moreover, attitude change following induced compliance can be increased if the physiological arousal is increased (e.g., by the ingestion of an amphetamine) or decreased by the ingestion of a sedative (Cooper, Zanna, & Taves, 1978).
A. DISSONANCE THEORY EVOLVES Cognitive dissonance theory has not remained static. Although Festinger’s theoretical ideas were elegantly straightforward, they could not quite predict some of the limiting conditions of the phenomenon. Why did a counter‐ attitudinal statement produce dissonance if it was engaged in freely but not if it was in response to an order or command? Why did dissonance apparently depend on a counter‐attitudinal statement bringing about an unwanted event? In Cooper and Worchel’s (1970) follow‐up to Festinger and Carlsmith’s study, for example, we found that the inverse relationship between incentive magnitude and attitude change only occurred when the waiting subject was convinced that the experiment was going to be dull, but not if he or she was unconvinced. Why does the occurrence of dissonance occur only when the aversive event was foreseeable and not when it was unforeseeable (Goethals, Cooper, & Naficy, 1979)? To deal with these anomalies, Cooper and Fazio (1984) proposed the New Look model of cognitive dissonance in which dissonance is aroused when a person takes personal responsibility for his or her behavior and that behavior has the potential to lead to foreseeable, unwanted consequences. Stone and Cooper (2001) expanded the theory further in the self‐standards model of dissonance as a way to accommodate the role of the self in the dissonance process. Other models have also been proposed to try to shed further light on the motivational basis of the theory (Harmon‐Jones, 1999). The distinctions among the models are not trivial for purposes of theoretical clarity and prediction. However, a discussion of those distinctions takes us beyond the
364
JOEL COOPER AND MICHAEL A. HOGG
current focus. The important conclusion from the 50 years of research since Festinger’s original statement is that cognitive dissonance occurs following freely chosen attitude‐discrepant behavior; it is experienced as unpleasant discomfort and it is dependent on, or at least strengthened by, the occurrence of a foreseeable aversive consequence emanating from the behavior.
B. DISSONANCE PROCESSES WITHIN GROUPS Until recently, cognitive dissonance has been studied from the perspective of an individual actor. Dissonance has been conceived of as a function of the discrepancy among cognitive elements within the individual. Very little attention has been paid to the social group in which the individual is embedded (see Cooper & Stone, 2000, for a review). Ironically, one of the very earliest studies supporting Festinger’s theory of cognitive dissonance was conducted in the context of an in vivo social group. In the mid‐1950s Ms. Marion Keech’s ‘‘Seekers’’ prophesized a cataclysm that would terminate all life on Earth. Festinger, Riecken, and Schachter (1956) reported that the members of the Seekers sought social support as a way of reducing dissonance, once their belief was shown to be false. However, with just a few exceptions (Cooper & Mackie, 1983; Sakai, 1999; Zanna & Sande, 1987) the subsequent focus of dissonance research almost immediately shifted to the study of the individual in isolation from the group. The emergence of the empirical study of the self (see Baumeister, 1998, 1999; Leary & Tangney, 2003; Sedikides & Gregg, 2003) helped to open the door to a more focused study of cognitive dissonance in a group context. Several investigators criticized Festinger’s original statement of cognitive dissonance theory because it did little to emphasize the importance of the self. Aronson (1968) was probably the first theorist to suggest that cognitive dissonance was not created by just any discrepant cognitions, but rather by cognitions that involved the self. Several years later, Steele and his colleagues echoed this emphasis in the formulation of self‐aYrmation theory (Steele, 1988; see Sherman & Cohen, 2006). For both views, preserving the integrity of the self‐concept lies at the heart of cognitive dissonance. Similarly, in Stone and Cooper’s (2001) self‐standards model, the self plays a critical role in determining when dissonance will be aroused and how dissonance will be reduced. The self, however, is multidetermined and multifaceted and it is probably more accurate to talk about selves rather than the self. For example, Brewer and Gardner (1996) distinguish between three forms of self: (1) the individual self, defined by personal traits that diVerentiate the self from all others; (2) the relational self, defined by dyadic relationships that assimilate the self
A THEORY OF VICARIOUS DISSONANCE
365
to significant other persons; and (3) the collective self, defined by group membership that diVerentiates ‘‘us’’ from ‘‘them.’’ Markus and Kitayama (1991) distinguished between the independent self, grounded in a view of the self as autonomous, separated from other people, and revealed through one’s inner thoughts and feelings, and the interdependent self, grounded in one’s connection to and relationships with other people and expressed through one’s roles and relationships. Social identity theorists (Tajfel & Turner, 1979; see Hogg & Abrams, 1988) distinguish between social identity—a self defined in terms of one’s group memberships, and personal identity—a self defined in terms of one’s idiosyncratic attributes and personal relationships. What is significant for the present chapter is the fact that a meaningful portion of our identity and sense of self is shared with other people—it is configured and defined by interdependence or common category membership with others. If self can be shared, or grounded in shared experience in this way, then the possibility arises that dissonance can be shared. More specifically, we may be able to experience dissonance vicariously by witnessing another member of a group with which we identify experiencing dissonance or behaving in a counter‐attitudinal or hypocritical manner. This simple idea may have far reaching consequences for ways in which large numbers of people can be persuaded to change their attitudes and perhaps behaviors. In this chapter, we ground our theory of vicarious dissonance specifically in social identity theory. Our analysis will show that when Professor Adams, in our prior vignette, agreed to speak against his privately held attitudes, every other professor at the table experienced cognitive dissonance vicariously. Every other professor was motivated to change his or her attitudes to become more favorable to the formerly disliked position.
III. Social Identity and Vicarious Dissonance Intersubjectivity, the ability to experience another person as oneself, is at the root of the experience of vicarious dissonance. The general idea of intersubjectivity is addressed in diVerent ways by a number of literatures in social psychology. For example, research on empathy and altruism discusses how the ability to empathize with another person, which can be enhanced by perception of similarity, increases the probability of altruistic and helpful acts (Batson & Oleson, 1991; Batson, Duncan, Ackerman, Buckley, & Birch, 1981). Closely related is research on perspective taking (Batson, Early, & Salvarani, 1997; Oswald, 1996) that shows that the ability to take the perspective of another person, which can be enhanced by perceived similarity or
366
JOEL COOPER AND MICHAEL A. HOGG
common group membership, may protect against prejudice and build more cooperative behavior (Galinsky & Ku, 2004; Galinsky & Moskowitz, 2000; Galinsky, Ku, & Wang, 2005; Vescio, Sechrist, & Paolucci, 2003; but see Maner et al., 2002). Research on intergroup emotions has shown that people who share a group membership or identify with the same group can experience shared feelings about an out‐group, and that this sense of shared identity is critical (Gordijn, Yzerbyt, Wigboldus, & Dumont, 2006; Mackie & Smith, 2002; Mackie, Devos, & Smith, 1999; Mackie, Silver, & Smith, 2004). A development of this research has also shown how shared identity can sponsor feelings of collective guilt (Lickel, Schmader, Curtis, Scarnier, & Ames, 2005; Maitner, Mackie, & Smith, in press). Together this research furnishes a precedent for the general idea that people may be able to experience others’ feelings, and that this may be enhanced under circumstances in which people perceive a bond between self and other. It is a real possibility that feelings of dissonance also may be shared or transmitted in this way—producing feelings of vicarious dissonance that may have exactly the same eVects on attitudes and behavior as nonvicarious, personal dissonance. The key question revolves around what social‐cognitive process instantiates the bond of intersubjectivity that kicks the vicarious process into play. Social identity theory provides a nice answer to this question. Social identity theory was originally conceived by Tajfel in the early 1970s (Tajfel, 1972) and has now developed into an influential theory of group processes, intergroup relations, and the collective self (for recent statements and overviews see Hogg, 2003, 2005a, 2006). Social identity theory (Tajfel & Turner, 1979) and self‐categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) argue that people represent social groups in terms of prototypes that maximize metacontrast—prototypes simultaneously embody in‐group defining attributes (values, beliefs, attitudes, behaviors, feelings) and accentuate diVerences between one’s in‐group and relevant out‐groups. The process of categorizing someone depersonalizes your perception of them—you view them through the lens of the category prototype, assigning them the prototypical attributes of their group. Depersonalization transforms social perception and perceptually renders members of a group more similar to one another, via their perceived assimilation to the category prototype. A key insight of self‐categorization theory is that the process of social categorization applies to self in precisely the same way—self‐ perception is depersonalized with the added eVect that one’s attitudes, feelings, and behaviors actually conform to the group prototype. Social categorization‐based depersonalization is a mechanism that binds people together through a shared social identity and thus produces the sort
A THEORY OF VICARIOUS DISSONANCE
367
of intersubjectivity discussed above. Simply recognizing that someone is a member of a particular group is not enough to engage this process—the social categorization must be psychologically salient as the basis of self‐ conception and social perception in a particular context (Oakes, Haslam, & Turner, 1994; Turner, Oakes, Haslam, & McGarty, 1994). People draw on accessible social categorizations—ones that are valued, important, and frequently employed aspects of self‐conception and social perception (they are chronically accessible in memory), and/or because they are self‐evident and perceptually salient in the immediate situation (they are situationally accessible). People are very ready to use accessible categories to make sense of their social context, investigating how well the categorization accounts for similarities and diVerences among people (structural or comparative fit) and how well the stereotypical properties of the categorization account for why people behave as they do (normative fit). People cycle through categorizations until an optimal level of fit is obtained—a process that is quick and automatic as people strive to reduce feelings of uncertainty about self‐conception, social interaction, and people’s behavior (Hogg, 2000, 2007). The categorization that has optimal fit becomes psychologically salient in that context as the basis of self‐categorization, group identification, and prototype‐based depersonalization. It triggers social identity‐related perceptions, cognitions, aVect, and behavior, and brings into play the intersubjectivity that allows in‐group members to share their attitudes, feelings, and perceptions. The social identity notion of depersonalization has been approached in a slightly diVerent way by Wright and his colleagues who discuss the self being extended to incorporate the ‘‘other’’ as part of the self (Aron & McLaughlin‐ Volpe, 2001; Wright, Aron, & Tropp, 2002). This idea comes from research showing that people in close interpersonal relationships tend to confuse the boundary between self and partner (Aron & Aron, 1996) and research on false consensus (Marks & Miller, 1987), self‐anchoring (Cadinu & Rothbart, 1996) and social projection (Robbins & Krueger, 2005) showing that people can project their own attributes onto other people and thus make them appear more similar to, or interchangeable with, self than they really are (cf. Otten, 2002). From our perspective, this supports the depersonalization thesis that under some circumstances related to common group identification, conditions of intersubjectivity may arise. Social identity theory identifies a process that might produce vicarious dissonance based on self‐categorization and group identification. The key prediction is that witnessing a fellow group member, or person with the same social identity as yourself, experiencing dissonance or behaving in ways that produce dissonance, will engender dissonance in you, the observer. This will happen to the extent that a common category membership/social identity is
368
JOEL COOPER AND MICHAEL A. HOGG
psychologically salient. Furthermore, the vicarious dissonance eVect will be accentuated to the extent that you identify strongly with the shared group membership, the identity furnished by the group is subjectively a more central and important aspect of who you are, and you and the other person are both highly prototypical members of the group. Another factor that may play a key moderator role is uncertainty reduction. According to uncertainty‐identity theory (Hogg, 2000, 2007), a key motivation for social identification is the need to reduce uncertainty about oneself. Identification achieves this because it circumscribes one’s self definition and prescribes what one should think, how one should behave, and how one’s interactions with others might play out. This process may be particularly relevant in the context of dissonance because the counter‐attitudinal behavior of fellow in‐group members intrinsically threatens the normative integrity of one’s group, which will tend to raise uncertainty and perhaps strengthen subsequent identification. For a discussion of social identity and attitudinal processes see Terry and Hogg (1996) and Hogg and Smith (in press). Let us now examine the way in which social identity aVects the observation of a group member’s inconsistent behavior to create the tensions state of vicarious cognitive dissonance. We will look at the empirical evidence and then explore the wider implications of vicarious dissonance for the study of culture and also as an intervention strategy for creating positive changes in health behaviors and attitudes (also see Cooper & Hogg, 2002; Monin, Norton, Cooper, & Hogg, 2004; Norton, Monin, Cooper, & Hogg, 2003).
IV. Evidence for Vicarious Dissonance A. A FIRST STUDY Our initial foray to find empirical confirmation of vicarious dissonance was a study conducted at Princeton University (Norton et al., 2003, Study 1). The general idea was to have participants observe another student agreeing to write an essay that took a position that they, the observer, thought was counter‐attitudinal. The question we asked was whether the participants would experience cognitive dissonance and change their own attitudes, solely because of the behavior of the fellow student. From our theoretical analysis of social identity and dissonance processes, we expected that the mere observation of counter‐attitudinal behavior would lead to attitude change in our participants, provided that the participants and the essay writer were
A THEORY OF VICARIOUS DISSONANCE
369
members of the same social group and provided that the participants identified with and felt positive about being a member of their group. The method we used for the study serves as a paradigm for much of the research we consider in this chapter, so we will present it in some detail. We created a fictitious cover story that participants found intriguing and involving. For our research purposes, the cover story created a rationale for a student to witness a fellow student agree to write a counter‐attitudinal message and to learn whether the student was a member of the participant’s social group. At Princeton University, all entering undergraduate students are assigned at random to one of five residential colleges. Each student lives and eats in one of the colleges and each college has its own social and academic activities. The student’s residential college served as the crux of the in‐group versus out‐group manipulation because each participant believed that he or she was witnessing an interaction with a student who happened to be a fellow resident of his or her residential college (in‐group member) or happened to be a resident of a diVerent residential college (out‐group member). The students arrived for a study of ‘‘linguistic subcultures’’ in groups of two, although each reported to a separate room, separated by two‐way mirrors. We told the students that we were interested in how people in diVerent residential colleges come to speak in slightly diVerent ways, learning to use slightly diVerent inflections or terms in their spoken behavior. For example, we know that people who live in the Midwest develop a slightly diVerent pattern of speech than people who live in South Carolina or Massachusetts. The experimenter explained that the purpose of the students’ participation in the current study was to see if these speech patterns occur in microcosms, that is, small groups within a larger context. We told the students that, in this study, we wanted to see if the speech patterns of students in the diVerent residential colleges at Princeton University were diVerent from one another and whether we could measure them. We explained that one of the two students, selected at random, was going to deliver a speech on a given topic and the other student was going to listen carefully, and then respond to several questions about the speaker’s speech patterns. Each participant was told that he or she was the one who had been randomly picked to rate the speech, while the student in the other room was assigned to give the speech. The procedure allowed us to make the student’s residential college group salient and manipulate systematically whether the speaker’s residential college group was the same (in‐group) or diVerent (out‐ group) from the participant’s. The experimenter found a pretext to turn the lights on briefly, which allowed the participants to see that there truly was another student in the other room. The illumination was kept low so that the students’ identities could not be accurately discerned. What students did
370
JOEL COOPER AND MICHAEL A. HOGG
not realize was that each of them had been assigned the role of listener. All information about what the alleged other student said or did was manipulated by instruction or audio tape. The experimenter left the room, ostensibly to instruct the other participant about the speech he or she was to make. During the intervening period, participants filled out various measures, including measures of how much they liked and felt identified with their residential college on a scale developed by Hogg and his associates (Hogg & Hains, 1996; Hogg, Cooper‐Shaw, & Holzworth, 1993; Hogg, Hains, & Mason, 1998). In a few minutes, the experimenter returned with a tape recording that included the completed speech and the experimenter’s alleged conversation with the other student. On the tape, the experimenter explained that he was fortunate to be able to combine two studies into one. The Dean’s oYce had asked for a study trying to assess student opinion about the possibility of raising tuition fees by a more than typical amount. Using a cover story developed by Linder, Cooper, and Jones (1967), the experimenter asked the student to write a strong and forceful speech advocating a spike in tuition fees. He explained that this would be the speech that the other subject (i.e., the real participant) would rate for its linguistic features and that it would then be sent on to the Dean’s oYce. The experimenter also asked the alleged other student how he or she felt about raising tuition and the student responded, ‘‘Well . . . I’d be against it.’’ The participants thus had a credible, albeit fabricated, story that allowed them to overhear an in‐group or an out‐group member make a counter‐ attitudinal speech on a controversial topic. The tape recorder was stopped while the writer supposedly organized his or her thoughts and then restarted for the participant to hear the alleged speech. The speech was a relatively brief and bland exposition on how higher tuition rates could allow the university to hire more faculty, purchase more books for the library, and so forth. Before rating the speech for its linguistic properties, participants were asked about their own attitudes toward tuition increases at the university. This served as the dependent measure of our study. Would students change their attitude after observing another person write a counter‐attitudinal speech? The results of the study by Norton et al. (2003, Study 1) are shown in Fig. 1. Our prediction was that overhearing a fellow student makes a counter‐ attitudinal speech that would lead to attitude change provided that (1) the other student was an in‐group, rather than an out‐group member and (2) participants felt highly identified with their group. We did not expect attitude change if the other student was an out‐group member or if participants did not like or identify with the residential college of which they were a member. The results of the regression analysis show that the prediction was supported. Attitude change occurred after hearing an in‐group member make
A THEORY OF VICARIOUS DISSONANCE
371
Group of speaker Out-group In-group 2
−1
0
1
2
Identification (standardized) Fig. 1. Regression slopes with attitude change predicted by the speaker’s group and participants’ level of in‐group identification.
the counter‐attitudinal speech and, as predicted, did so as a direct function of the participants’ in‐group identification.
B. VICARIOUS DISSONANCE OR PERSUASION? The results of the first study were consistent with the notion that people can experience dissonance vicariously. However, there were some anomalies in the data of the study that make its interpretation ambiguous. It may be that participants were actually persuaded by the content of the pro‐tuition speech. Although we intended the speech to be relatively bland and straightforward, it did contain arguments in support of a tuition increase. It has been shown elsewhere (Mackie, Worth, & Asuncion, 1990) that group membership can aVect the depth of information processing that people engage in. Messages from in‐group members are more deeply processed and this may have contributed to the persuasiveness of the essay. It would also follow that people who are more highly identified with their group, relative to those who do not care for the group they are in, would be more likely to engage in deeper and more elaborated processing of an in‐group message.
372
JOEL COOPER AND MICHAEL A. HOGG
In our second study (Norton et al., 2003, Study 2), we sought to eliminate diVerential processing or any other form of persuasion as an explanation for the results. We did this by eliminating the actual speech. We know from research on personal cognitive dissonance that it is not necessary for a person actually to deliver a speech or write an argument for dissonance to arise. The necessary condition for dissonance for a person is to agree to engage in counter‐attitudinal advocacy (Linder, Cooper & Wicklund, 1968; Waterman & Katkin, 1967). The cognition that one has agreed to make a statement to the Dean in favor of a tuition increase is discrepant from the cognition that one’s attitude is against the increase and is suYcient to produce cognitive dissonance. We transported this logic from personal dissonance into our research and designed a study in which participants overheard an in‐group member agree to make a speech that was discrepant with his or her attitudes. We also wished to be certain that attitude change occurred only when the fellow in‐group member made a speech that was specifically counter‐ attitudinal. Dissonance should only occur in that circumstance. If the speaker were actually in favor of the increase, it might raise a number of issues, but would not arouse dissonance. Our underlying theoretical analysis requires that the observer (i.e., the participant) understands that an in‐group member has agreed to act in a counter‐attitudinal manner. If the persuasiveness of the message is the key variable, then the speaker’s agreeing with the position or disagreeing with it should not have much impact. But cognitive dissonance requires that the behavior be discrepant with the speaker’s attitude. Therefore, we varied whether the speaker was in favor, or opposed, to a tuition increase. In response to the experimenter’s inquiry, the in‐group member was heard to say, ‘‘I think that (a tuition increase) is a pretty good idea,’’ or ‘‘I’d be against (a tuition increase).’’ Figure 2 supports our analysis. When the agreement to make the speech was in line with the group member’s attitude, then it had no impact on the participant, regardless of the participant’s level of identification with the in‐group. However, when the in‐group member’s agreement was counter‐attitudinal, then it led to attitude change in the observer as a direct, monotonic function of in‐group identification.
C. IS VICARIOUS ATTITUDE CHANGE VICARIOUS DISSONANCE? How certain can we be that the attitude change that occurs when observing a group member act counter‐attitudinally is really dissonance? Our approach in the next study (Norton et al., 2003, Study 3) was to manipulate in a single
373
A THEORY OF VICARIOUS DISSONANCE
6
4
2
0 Opinion of speaker Agree 2
Disagree −2
−1
0 1 Identification (standardized)
2
Fig. 2. Regression showing increasing attitude change as a function of the speech writer’s disagreement with the position and the participant’s level of identification with the in‐group.
study the variables that are known to aVect personal dissonance. From the decades of research on personal dissonance, we know that dissonance only occurs when a person is responsible (i.e., makes an act of free choice) to act in a counter‐attitudinal fashion (Cooper, 1971; Davis & Jones, 1960; Linder et al., 1967) and when the act has potential unwanted consequences (Cooper & Fazio, 1984; Cooper & Worchel, 1970; Scher & Cooper, 1989; Stone & Cooper, 2001). There are some models of cognitive dissonance theory that take issue with whether an unwanted consequence is a necessary condition for dissonance to occur (Beauvois & Joule, 1999; Harmon‐Jones, Brehm, Greenberg, Simon, & Nelson, 1996), but all agree that producing an unwanted event is, at the very least, an important cognition for the arousal of dissonance. Our third study (Norton et al., 2003, Study 3) used our familiar procedure, although this time at the University of Queensland in Australia. The attitude issue was the imposition of up‐front fees (payment of tuition up‐front rather than through a loan to be repaid later on), which was a very unpopular
374
JOEL COOPER AND MICHAEL A. HOGG
position at the time for students at universities throughout Australia. Not surprisingly, we requested the speech writer in this experiment to make a strong and forceful statement supporting the collection of up‐front fees. We manipulated the two crucial independent variables: choice of the speaker to make the counter‐attitudinal speech and whether the speech had the potential to lead to an aversive consequence. The participants’ level of in‐group identification was measured with the scale used in Norton et al. (2003) Study 1 described above. The results are depicted in Fig. 3. For ease of presentation, we analyzed the data separately for low identifiers and high identifiers, although regression analysis confirms the results of the factorial analysis of variance. Participants changed their attitudes in the direction of the position advocated in the essay. In addition to main eVects for choice, consequences, and identification, there was the predicted second order interaction: Attitudes changed in the direction of the discrepant position when the speaker had high choice, when the speech had a potential negative consequence, and when the participants felt strongly identified with their fellow students at the University of Queensland. The first two of those variables are the ones that typically enable personal dissonance to occur (high choice and high consequences). Thus, dissonance in the observer occurred in conditions in which dissonance would have been experienced by the speaker—but only when the observers were highly identified with and attracted to their in‐group.
4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Low choice Av. Cons. - Hi ID Av. Cons. -Low Id
High choice No Cons.-Hi Id No Cons. Low ID
Fig. 3. Vicarious attitude change as a function of choice, consequence, and level of identification with the in‐group.
A THEORY OF VICARIOUS DISSONANCE
375
D. THE AFFECTIVE EXPERIENCE OF VICARIOUS DISSONANCE What does vicarious dissonance feel like? From the data presented thus far, we know that people change their attitudes when they observe an in‐ group member act in a way that should cause that member to experience personal dissonance. This raises the important question of what propels or activates vicarious dissonance. In the case of personal dissonance, evidence has supported Festinger’s (1957) notion that dissonance is uncomfortable (Elliot & Devine, 1994; Losch & Cacioppo, 1990) and arousing (Croyle & Cooper, 1983; Losch & Cacioppo, 1990; Zanna & Cooper, 1974). Attitude change is at the service of reducing that uncomfortable state of arousal. But what motivates vicarious dissonance? Is it also experienced as unpleasant aVect? In the first two studies reported above, we had no success finding evidence for the experience of unpleasant aVect in the participants. We created an index of personal discomfort by asking participants to rate the degree to which they felt uncomfortable, bothered, and uneasy. These items were the ones used by Elliot and Devine (1994) and several other researchers to corroborate the occurrence of personal discomfort. The items were highly correlated in our studies ( ¼ .94) but produced no reliable eVects. At least on these items, vicarious dissonance did not produce discomfort.
1. In Search of Vicarious Discomfort Our current conception of vicarious discomfort is indebted to research traditions dating back to the foundations of social psychology. Some of the very early theories of social behavior emphasized vicarious arousal. McDougall (1908) emphasized ‘‘primitive passive sympathy’’ in which the emotion of an actor is directly and automatically transferred to an observer. Similarly, Lipps (1926) discussed ‘‘inner imitation’’ in which an observer automatically emulates the behaviors of an actor and thus ‘‘catches’’ the emotion of others (Hatfield, Cacioppo, & Rapson, 1994). A useful distinction often made by researchers who study empathy is between the automatic experience of another’s emotion and experiences in which the observers more actively work to understand the situation faced by the actor. La Pierre and Farnsworth (1949), for example, discuss empathic emotion as: Mentally putting oneself in the place of another and reacting more or less intensely to the stimuli that actually impinge upon that other person. Thus should a person with whom we have closely identified ourselves cut his finger in our presence, we would vicariously ‘‘feel the pain of that hurt’’ (p. 225).
376
JOEL COOPER AND MICHAEL A. HOGG
In other words, feeling the pain requires taking the other’s perspective and feeling what the closely identified other feels. A similar distinction has been oVered by Asch (1952) and Heider (1958) discussing the distinction between emotional contagion and sympathy. In emotional contagion, an actor’s emotion simply spills over to an observer’s aVective state, with a need to understand the actor’s situation. True sympathy, on the other hand, refers to the emotions one experiences as a result of understanding the psychological condition of the other person. It does not imply that the emotion be identical, but only congruent, with the emotion experienced by the actor. A person chatting with a friend who has discovered true love does not catch the same experience as his friend but, with his ability to take his friends perspective, understands his friend’s emotion and experience a congruent emotion of joy. Thus, it is important to distinguish between two diVerent types of discomfort that observers might experience when witnessing an actor’s dilemma in the vicarious dissonance studies: ‘‘vicarious discomfort’’ and ‘‘personal discomfort’’ (Norton et al., 2003). The former comes from understanding the plight of the person who engaged in the counter‐attitudinal behavior. It comes from an understanding, perhaps at a level below awareness, of the discomfort experienced by a person who was responsible for bringing about an aversive event by agreeing to write an essay in favor of higher fees. But it is not the same experience. Vicarious discomfort is not an automatic, empathic ‘‘catching’’ of emotional discomfort but is more akin to the concept of sympathy when imagining how one would feel in the actor’s shoes (Norton et al., 2003). Accordingly, in the study described above conducted at the University of Queensland whose attitude data are depicted in Fig. 3 above (Norton et al., 2003, Study 3), we added questions that had not been included in the first two studies. We continued to ask participants to rate themselves on negative and positive personal aVect dimensions including how bothered, uneasy, uncomfortable, good, optimistic, and happy they felt. In addition, we also asked participants to rate themselves on vicarious aVect dimensions. The question asked, ‘‘How do you think you would feel if you were in the speech writer’s position?’’ Participants responded on the same dimensions as for personal aVect (bothered, uneasy, comfortable, good optimistic, and happy). Although personal aVect scores continued to show no diVerences as a function of condition, the scores for vicarious aVect showed the same pattern as the attitude data shown in Fig. 3. Vicarious discomfort was high in the conditions in which students who were highly identified with their in‐group observed a fellow group member write a speech favoring higher fees, but only when that speech writer voluntarily agreed to make the speech and the speech resulted in a potentially aversive consequence (Fig. 4). That is, the same second order interaction found for attitudes was also found for the measure of vicarious aVect.
377
A THEORY OF VICARIOUS DISSONANCE
4.0 3.5 3.0 2.5
Av.Con.-Hi Id Av.Con.-Lo Id No Con.-Hi Id No Con.-Lo Id
2.0 1.5 1.0 0.5 0.0 Low choice
High choice
Fig. 4. Vicarious discomfort as a function of choice, consequence, and level of identification with the in‐group.
2. Vicarious Attitude Change at the Service of Vicarious Discomfort In the study of personal dissonance, there is empirical support for the idea that attitude change reduces personal aVect. Elliot and Devine (1994) varied the order in which they asked participants in an induced compliance study about their attitudes and their personal aVect. They found that when they asked about aVect prior to attitudes, there was evidence for personal discomfort and there was attitude change. However, when they reversed the order, they found that participants in the high‐dissonance condition did not show heightened personal discomfort. In other words, when participants had a chance to change their attitudes in the direction of their counter‐attitudinal behavior, they had eVectively dealt with their dissonance and no longer experienced discomfort. Elliot and Devine (1994) interpreted this finding as consistent with the idea that attitude change is at the service of reducing psychological discomfort (see also Galinsky, Stone & Cooper, 2000). In the experiment by Norton et al. (2003, Study 3), we also varied the order of the aVect and attitude measure. There was no indication of personal discomfort regardless of the order of the questions. However, the results of the vicarious aVect measure tell a diVerent story. Recall that when vicarious discomfort was measured first, it produced the results shown in Fig. 4. However, when participants were asked about their vicarious discomfort following the attitude measure, vicarious discomfort was significantly reduced. This suggests that attitude change is an eVective way to reduce the discomfort that a person can imagine experiencing if he or she were in the speechwriter’s position. Overall, the research shows that observers who witness an in‐group member act in a counter‐attitudinal fashion show vicarious attitude change under
378
JOEL COOPER AND MICHAEL A. HOGG
the same set of conditions that would cause the in‐group member to experience personal dissonance: under conditions of free choice and high aversive consequences. Moreover, this does not occur if the speech writer is from a diVerent group and it only occurs if the observer is highly identified with his group. In addition, vicarious dissonance is characterized by the experience of vicarious discomfort, which in turn can be reduced by attitude change. This is the most direct evidence that vicarious attitude change is indeed based on cognitive dissonance processes.
E. GROUP PROTOTYPICALITY AND VICARIOUS DISSONANCE The theoretical basis for our investigation of vicarious cognitive dissonance centers on the notion that when people identify strongly with a group, they experience intersubjectivity in which fellow members’ internal states become their own internal states. One possible moderator of the strength of this eVect is group prototypicality. In salient groups, members are attentive to information reflecting on what is and what is not group prototypical, and who is more prototypical than whom—prototypicality is the yardstick of group life as it defines one’s identity and one’s place within the group (Haslam, Oakes, McGarty, Turner, & Onorato, 1995; Hogg, 2005b). Social identity research has, for example, shown how members of a crowd are guided by prototype‐relevant behavior of others (Reicher, 1984; Reicher & Hopkins, 2003) and how being prototypical endows leaders of salient groups with a marked advantage in leadership endorsement and eVectiveness (Hogg & van Knippenberg, 2003). Overall, more prototypical members are also more likely to identify with a given group, and members are more likely to believe that highly prototypical others also identify strongly with the group. The implication for vicarious dissonance is quite clear: Vicarious dissonance will be stronger to the extent that self and/or other prototypicality is greater. One prediction, then, is that vicarious dissonance is strongest in conditions in which an in‐group member who agrees to act in a counter‐ attitudinal fashion is prototypical of the group. We found some tentative support for this in a study by Monin et al. (2004) that will be explained in more detail in a subsequent section. Monin et al. (2004) found a significant correlation between the rated typicality of the group member who made the counter‐attitudinal speech and the degree of attitude change manifested by the participants. Building on this finding, we designed a study to investigate this prediction directly (Hogg & Cooper, 2006). Our study used the linguistic subculture paradigm described above (Norton et al., 2003, Study 1). University of
A THEORY OF VICARIOUS DISSONANCE
379
Queensland students heard a fellow student agree to make a statement supporting the collection of up‐front fees under conditions of high or low choice. The level of unwanted consequences was always set high in that the Dean would see the speeches supporting the up‐front fees. The major independent variable in the study was the prototypicality of the speaker as well as the prototypicality of the participant. Prior to the explanation of linguistic subcultures, the participants were asked to complete a battery of questionnaires. As in our previous research, one of those questionnaires was a measure of the students’ identification with his or her group: University of Queensland undergraduates. Other measures asked a series of questions about values, attitudes, and behaviors whose sole purpose was to make credible the information we were about to deliver. When the alleged purpose of the experiment was explained and participants understood that we were studying linguistic subcultures among students at various universities in Australia, the experimenter informed the participant and his or her partner that the questionnaires they had filled out had been analyzed and that the data had been entered into the larger data file that was composed of all responses from University of Queensland student who had taken the test. The experimenter continued that he knew that students usually like to see how their data profile looks in relation to other students at the university. Accordingly, the participant was shown a scatterplot drawing on a computer screen. He explained that each dot represented a composite score of students’ overall values profile. Students whose scores were near the center of the scatterplot were typical UQ students. Scores at the fringe of the array were nontypical of UQ students. The computer screen showed the array of dots. In order to emphasize the point about typicality, a red circle was drawn near the center of the scatterplot. The partner’s score and the participant’s score were shown in vivid color, either inside or outside the circle. Prototypicality was manipulated for each score independently so that the participant was or was not prototypical and the partner was or was not prototypical. The experiment then continued using the basic vicarious dissonance paradigm established in our prior work. The participant was selected as the observer and the partner was chosen as the speech giver. The partner’s freedom to decline the request to make the speech was varied. This resulted in a 2 (decision freedom of the speaker) 2 (speaker prototypicality) 2 (participant prototypicality) design with the strength of identification with the in‐group (UQ students) measured as a continuous variable. The results of the study revealed that in the high‐choice conditions, participants changed their attitudes toward up‐front fees, but especially when they believed that both they and the fellow group member were prototypical members of the group (see Fig. 5). In that circumstance, the participant—having
380
JOEL COOPER AND MICHAEL A. HOGG
5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 High selfprototypical
Low selfprototypical
Other high prototypical
Other low prototypical
Fig. 5. EVect of self and other prototypicality on attitude change.
been shown that he is a prototypical group member—seemed to identify more closely with his prototypical partner and experienced vicarious dissonance. It is important, too, that the eVect was moderated by the degree of identification the participants felt toward the group. With one fascinating exception that we will describe below, attitude change via vicarious dissonance occurred for group members as a function of the identification with their in‐ group. The higher the identification, the higher the attitude change. These results are interesting for two reasons. First, they confirm the notion that intersubjectivity occurs most easily with a group member who is perceived to be prototypical of a group. Second, it shows that the participants themselves are more likely to experience intersubjectivity if they perceive themselves to be prototypical. The perception that both self and other are highly prototypical group members seems to be the most eVective determinant of strong intersubjectivity. The one intriguing exception to this finding was among students who found that they and the speaker were at the periphery of the student group and who did not feel particularly attached to their group on the individual diVerence measure of group attachment. For these students, greater dissonance was experienced when the speaker agreed to write the counter‐ attitudinal essay. This group of participants, paradoxically, makes a strong case for a social identity explanation of vicarious dissonance. In eVect, the student is a marginal group member or outsider who discovers a fellow traveler—another outsider—to bond with and perhaps form an embryonic minority group with ‘‘those people who do not like the university.’’
A THEORY OF VICARIOUS DISSONANCE
381
Given that shared marginal status may be a particularly potent source of social identity (cf. Abrams, Hogg, & Marques, 2005; Mugny & Pe´rez, 1991) from a vicarious dissonance perspective, this situation might be particularly conducive to strong identification‐based intersubjectivity.
V. The Process of Vicarious Dissonance The results of the studies presented so far tell a story about the conditions that give rise to vicarious dissonance. As depicted in Fig. 6, we view the process of vicarious dissonance to begin as a group‐based phenomenon that is made possible when people share membership in a group that is subjectively important and attractive, and with which they identify. When membership in the social group is particularly accessible, intersubjectivity occurs such that there is a ‘‘fusing’’ of identities between prototypical group members. Witnessing a group member (O) act in an attitude‐discrepant fashion causes other group members (P) to experience the uncomfortable aVective state of vicarious cognitive dissonance. The necessary conditions for O’s counter‐attitudinal behavior to result in vicarious dissonance for P seem to be the same as those that facilitate personal dissonance. P must perceive that the group member, O, freely chose the inconsistent behavior, and based on accessible standards of judgment, the behavior potentially can result in an unwanted consequence. This set of perceptions produces the uncomfortable vicarious aVect, which is ultimately reduced when group members (P) change their attitudes. A deeper level question is still open for speculation. What is actually accomplished by attitude change following the group member’s dissonance‐ producing behavior? One possibility is that people are shifting attitudes strategically in order to accomplish interpersonal or intergroup goals. For example, an observer may change his or her attitude in order to support the fellow in‐group member. The intersubjectivity that occurs in highly attractive groups may cause people to intuit what one’s partner is going to experience following his or her counter‐attitudinal speech writing and what new attitude the partner is likely to form. Because all of us have had experience with dissonance‐ arousing situations in ordinary life, we all understand what the group member is going through. Through an empathic process, we feel the pain and discomfort of the group member. We then realize that the group member is going to change his or her opinion about tuition or up‐front fees or whatever the issue happens to be. It is possible that the participants change their attitudes in order to more closely approximate the group member’s new
Act is less aversive
P
P feels vicarious discomfort P
P engages in vicarious attitude change
O agrees to discrep. act
O O
Increase in group cohesiveness Intersubjectivity based on common group membership
P observes O’s behavior
P’s experiences as a group member
Reduce uncertainty
Consequence of vicarious attitude change Fig. 6. A model of vicarious cognitive dissonance.
A THEORY OF VICARIOUS DISSONANCE
383
attitude, either as a sign of support for the group member or as an attempt to conform more closely to the new group norm. A robust finding from research on social identity and conformity is for group members to express opinions in line with the group norm (Abrams & Hogg, 1990; Turner & Oakes, 1989). We looked for evidence for this approach in a study by Monin et al. (2004). Princeton University students experienced the vicarious aVect after watching a fellow student make an attitude‐discrepant speech advocating the unpopular position that parents of university students should have unlimited access to students’ health records. As in the previous research, we found evidence to support the arousal of vicarious aVect and we found attitude change in the predicted conditions. When the speaker was provided with high choice to make the speech in favor of giving parents access to student health records (Monin et al., 2004, Study 1), when there was a foreseeable negative consequence (Study 2), and when the speaker made it explicit that he or she was personally against that position, the participants’ vicarious negative aVect was high and there was a significant shift in attitudes (Monin et al., 2004). However, in both studies, we asked the participants what they thought the speaker’s attitude was both before and after he wrote his essay supporting unlimited access. The results showed that the participants remembered correctly that the speaker was against the position prior to writing his speech. In fact, the more the subject believed the speaker was hostile to the position in the essay, the more vicarious attitude change the participants showed. However, when asked whether the speaker changed his or her attitude after having written the speech, the participants said ‘‘no.’’ There was no systematic change reported by our participants in their belief about the speech maker’s new attitude. Apparently, conformity to the speaker’s new position, at least in these studies, did not underlie the vicarious dissonance eVect. Monin et al. (2004) also examined the relationship of vicarious dissonance with a measure of empathy. Using Davis’ (1983) multidimensional Interpersonal Reactivity Index, we found a relationship between participants’ score on the empathic concerns subscale and our measure of vicarious negative aVect. There was also a correlation between the measure of personal distress and vicarious attitude change, but there was no correlation between the empathic concerns subscale and vicarious attitude change. The only measure that directly links aVect and attitude change was our measure of vicarious aVect. Moreover, as in the Norton et al. (2003) experiments, asking about vicarious aVect following the measure of vicarious attitude change showed that the negative vicarious aVect was eliminated. This oVers continued support for the notion that vicarious attitude change is at the service of reducing the negative feeling of vicarious discomfort.
384
JOEL COOPER AND MICHAEL A. HOGG
If vicarious dissonance is not a way of conforming to a group norm and if it is not an entirely empathic process of sharing a fellow group member’s pain, what is it? We feel that vicarious dissonance reduction has the potential to serve a particularly potent collective function for the group as a whole. At the individual level, vicarious dissonance reduction through attitude change is satisfying—it reduces the discomfort of (vicariously experienced) dissonance. Now picture this process playing out in a face‐to‐face setting of several group members witnessing one of their own behaving counter‐ attitudinally. Vicarious dissonance reduction through attitude change is a collective event that, we might speculate, would have a particularly powerful eVect on group life. The fact that all members shift their attitude will create a general atmosphere of enhanced harmony and cohesion within the group. It might also provide consensual validation of members’ social identity in the group and thus reduce feelings of self‐conceptual uncertainty that might have been raised by vicarious dissonance—a process that will enhance identification and solidarity (Hogg, 2000, in press). These ideas concerning the collective dimension of vicarious dissonance are as yet speculative, but are the focus of subsequent research.
VI. Vicarious Dissonance and the Collective A. PERSONAL COGNITIVE DISSONANCE IN EAST AND WEST There has been considerable speculation and research concerning the possible cultural specificity of cognitive dissonance. Miller (1984) was one of the first to argue that many of the processes studied by social psychologists in North America and Europe are not necessarily generalizable to other cultures. Echoing Hofstede’s (1980) classic distinction between individualistic and collectivistic cultures, Miller drew a useful distinction between cultures that are basically agentic in orientation and those that are holistic. In the former category are cultures socialized with a western European influence while cultures in Asia may be characterized as having a more holistic orientation. In the latter, interconnectedness within social and societal groups are more prominent than in agentic societies. Duties, roles, and obligations play a larger role in determining appropriate social behavior than do individual responsibility and freedom. In her empirical work, Miller (1984) asked Hindus living in Mysore, India, and North Americans living in Chicago to make attributions about the causes of people’s behaviors in a variety of social settings. She found that her Indian sample, as representative of people socialized in a holistic culture, was more likely to see the cause of people’s actions as
A THEORY OF VICARIOUS DISSONANCE
385
residing in their roles and social obligations, whereas Americans attributed the same behavior to the free will and individual characteristics of the actors in the scenarios. Markus and Kitayama (1991) made a similar distinction among cultures by focusing on the degree to which societies configure the self as independent or interdependent (cf. Oyserman, Kemmelmeier, & Coon, 2002). The self in independent culture is focused inward and is focused on the meaning of their own behavior for the way in which they express their characteristics, attitudes, and traits. The self in interdependent cultures is intrinsically connected to others. Individual traits are not more important than the quality of relationships. The ability to create harmony in relationships and to act for the good of the people with whom you are connected is equally important to the interdependent self as any personal belief, attitude, or trait. The self‐ concept of people in interdependent cultures is based on a joint function of the worthiness of their individual dispositions and their ability to maintain pleasant and harmonious relationships with others with whom they are connected. In the Western cultures that feature independent selves, it is appropriate to say what we believe, believe what we say, and take responsibility for our actions. Consistency and responsibility are admired traits. In interdependent countries, according to Markus and Kitayama’s (1991) analysis, saying what one believes and believing what one says are more complicated. There is as much self‐defining value in controlling what one says if it is directed to the self‐defining aspect of creating harmony in relationships. Markus and Kitayama raised the question of whether the motive for cognitive consistency is as important for people in interdependent cultures as it is for people in independent cultures. They suggested that dissonance reduction may not be a universal motivation after all, but rather may be a phenomenon restricted to the agentic, individualistic cultures of North America, Western Europe, and Australasia. Heine and Lehman (1997) collected data relevant to the distinction between independent and interdependent cultures in the arousal of dissonance. The study was conducted in Vancouver, Canada on culturally distinct samples of participants. One group of participants was composed of Canadian citizens of European descent. The other group was recruited from visitors to Canada whose country of origin was Japan and who were visiting Canada on short‐term visas. Using the ‘‘free choice’’ research method introduced by Brehm (1956), participants were asked to choose between two musical CDs. Decades of research in independent Western cultures has shown that, following a choice between two attractive decision alternatives, the chosen alternative is seen as more attractive and the rejected alternative is seen as less attractive after the decision than it was prior to the decision. This basic
386
JOEL COOPER AND MICHAEL A. HOGG
dissonance prediction was replicated for the Canadian citizens. It was not, however, replicated for the Japanese sample. As Heine and Lehman (1997) had predicted, participants representing the interdependent culture saw no need to change their personal attitudes toward the music as a consequence of their having made a decision. The Heine and Lehman (1997) study raised the question of whether cognitive dissonance, which relies on an individual’s responsibility for the consequences of his or her decisions, is a uniquely Western rather than a general phenomenon. Recent research has shown that the more interpersonal consequences that accrue to a decision, the more likely it is that dissonance will be experienced and resolved, even in interdependent cultures. For example, Hoshino‐Browne et al. (2005) found evidence that Asian participants changed their attitudes about chosen and rejected choice alternatives if they were making decisions for their friends rather than themselves. Kitayama, Snibbe, Markus, and Suzuki (2004) also found that Asian participants spread the choice alternatives consistent with dissonance theory, but only when they were instructed or primed to think about their social relationships. Therefore, the trend of current research is to suggest that cognitive dissonance is not aroused similarly across cultures, although evidence for dissonance can still be found if participants experiencing dissonance are intertwined with others in an interpersonal context.
B. VICARIOUS COGNITIVE DISSONANCE IN INTERDEPENDENT CULTURES In contrast to personal dissonance, vicarious cognitive dissonance is a uniquely social event. It occurs in groups, precisely because of the relationship among group members. It occurs because people care about their group membership and the interpersonal relationships among group members. Relative to personal dissonance, vicarious cognitive dissonance should be a characteristic feature of interdependent cultures. We collected new data in South Korea to support this conclusion. Chong and Cooper (2007) asked students at Korea University in Seoul to participate in a study on the eVect of increased tuition rates at their university. The study was run by a Korean speaking researcher in the Korean language but was a faithful replication of the study reported by Norton et al. (2003). Participants were recruited as part of a study on linguistic subcultures. They believed that they were overhearing another student from their university, or from Yonsei University, agree to make a speech in favor of increasing tuition for university students in South Korea. Yonsei University was selected as the alleged home university of the other student because the two universities are
A THEORY OF VICARIOUS DISSONANCE
387
perceived as rival universities in South Korea. Korea University participants who believed the other student was from Yonsei believed that they were witnessing an out‐group member deliver the speech; students who believed the speaker was from Korea University believed they were overhearing an in‐ group member. As in the Norton et al. (2003) study, the speech maker either was asked or told to make a pro‐tuition speech, which, in turn, was to be used by the Dean of the university in forming his own policy about tuition increases. The participants then indicated their own agreement with the proposal to increase tuition. Recall that Norton et al. (2003) found that the stronger the participants’ identification with the in‐group, the stronger was the eVect of vicarious dissonance. Therefore, Korea University students’ strength of identification with Korea University was also measured with a Korean language translation of the identification scale used in the Norton et al. (2003) studies. The results depicted in Fig. 7 strongly support the existence of vicarious dissonance in the interdependent society of South Korea. Korea University students changed their attitudes to become more favorable to a tuition increase when a pro‐tuition statement was made by an in‐group member who freely decided to make the speech. Moreover, as predicted, the stronger the students’ identification with their in‐group, the more they changed their attitudes. An induced compliance also study run at Korea University using a personal dissonance paradigm showed no attitude change at all (Chong & Cooper, 2007). Students from Korea University were asked or told to make a speech favoring tuition increases at all Korean University. In the high‐choice
16 14 12 10 8 6 4 2 0 Speaker-in-group High identify
Speaker-out-group Low identify
Fig. 7. Vicarious attitude change toward tuition increase at South Korean universities.
388
JOEL COOPER AND MICHAEL A. HOGG
conditions, all of the students agreed to make the speech. Yet, attitudes measured following the pro‐tuition statement showed no change on the tuition question. Consistent with the findings of Hoshino‐Browne et al. (2005), who found dissonance among East Asians only in a social context, choosing to make a speech in favor of an unwanted policy did not produce dissonance in this study. However, as shown in Fig. 7, when the highly interpersonal vicarious dissonance procedure was invoked, then students in Korea showed the eVects of cognitive dissonance.
C. PRIMING CULTURAL IDENTITIES People belong to many diVerent social groups and categories that can vary in subjective importance and strength of identification, and in contextual and cognitive accessibility. The social identity process of salience described earlier determines which social identity is psychologically salient in which context—and thus whether the intersubjectivity necessary for vicarious dissonance exists. A university student who is a member of a university basketball team and also the university’s debate team will think of his or her basketball identity when in basketball attire and more about debate team membership when dressed in a suit. Many people also identify with several cultural groups. An Irish‐Canadian living in Montreal may sometimes identify with his Irish heritage and feel intersubjectivity with other prototypical Irish citizens. At other times, this person may identify with Canada and still other times with French‐speaking Quebec. Similarly, an Asian‐American student living in New Jersey may sometimes identify more closely with her Asian heritage and sometimes identify more closely as a US citizen. In one study, Asian American students at Princeton University were influenced to think more about their Asian or university identities by use of a straightforward priming procedure (Cooper, 2003). All of the participants were women who were either freshman or sophomores at Princeton and all identified themselves as Asian Americans. When they came to the laboratory, the experimenter (also an Asian American female) explained the linguistic subculture story (Norton et al., 2003). However, she also explained that she was doing an unrelated study as part of her thesis and would first like the participants to write a brief paragraph about some of their experiences at the university. Half of the students were asked to write about what it was like to be an Asian in the context of their first years at the university. The other half were asked to write about their most and least favorite courses during their time at Princeton. In this way, we hoped to make Asian identity more salient for the first group and the more local Princeton identity
389
A THEORY OF VICARIOUS DISSONANCE
salient for the latter group. We expected that the diVerential accessibility of their cultural identities would make one group use standards of judgment more typical of interdependent Asian culture and the other group use standards of judgment more typical of Western, independent cultures. Vicarious dissonance was aroused in a way similar to the previous investigations. Linguistic subcultures served as the overall cover story and Princeton University served as the in‐group. However, some of the students entered this phase of the experiment with their Asian identities primed while others entered with their university identities primed. The speech writer was asked, or told, to compose a speech arguing that tuition rates should be increased. Following the other student’s decision to make the speech, the participants were asked about their own attitudes toward tuition increase. The vicarious dissonance prediction is that participants who observed another university student agree to make the counter‐attitudinal speech would change their attitudes to become more favorable to tuition increases. We expected that this eVect would be moderated by cultural accessibility. We expected students with their interdependent Asian culture primed would experience more vicarious dissonance and change their attitudes to a greater extent than students whose Princeton identities had been primed. The results supported this prediction. Table I shows that there was an overall vicarious dissonance eVect. The main eVect for choice was highly significant. Participants showed greater support for tuition increases following a group member’s free choice to make the counter‐attitudinal speech than when the speech was commanded by the experimenter. However, the degree of attitude change was greatest in the choice condition when the students’ Asian identity had been made accessible. The significant interaction between choice and identity showed that the vicarious dissonance eVect for people primed with their Asian identity was greater than for participants primed with their Princeton identity. In conclusion, evidence now shows that dissonance is aVected by cultural norms. Some of the activities that arouse cognitive dissonance in one culture produce less consternation in others. In terms of the self‐standards model of
TABLE I ATTITUDES TOWARD TUITION INCREASE BY ASIAN‐AMERICAN STUDENTS Prime Asian
Princeton
Choice
6.1
3.2
No choice
2.2
1.9
390
JOEL COOPER AND MICHAEL A. HOGG
dissonance (Stone & Cooper, 2001), what is judged to be an aversive behavioral consequence in one culture may be considered less aversive in other cultures, as diVerent standards of judgment are brought to bear on the consequences of behavior (Kitayama, Ishii, Imada, Takemura, & Ramaswamy, 2006). However, it seems to be the case that cultures in which the self is interdependent provide an especially eVective context for the arousal of vicarious cognitive dissonance.
VII. Vicarious Hypocrisy: Translating Dissonance into Actions and Attitudes to Improve Health That cognitive dissonance can be experienced vicariously suggests that it occurs far more frequently than we had previously realized. Dissonance not only motivates people to change attitudes and behaviors as a function of their own behaviors but also motivates change as a consequence of group members’ actions. The fact that mere observation of others’ behavior can cause dissonance that may promote attitude and behavior change has implications for attitude and behavior change on a large scale. Of particular interest is the possibility that vicarious dissonance could be used as a mechanism for large‐scale improvement in people’s health‐related attitudes and behaviors. The personal and societal costs of dysfunctional health‐ related attitudes and behavior are enormous and are a major priority for social intervention. In many cases, people are well aware of the ingredients that comprise a healthy lifestyle—that they should exercise, eat healthily, not smoke, minimize sun exposure, and so forth. And yet they smoke, take no exercise, and gorge on obscene portions of fatty fast food and syrupy sodas. Most smokers know the harmful eVects of nicotine, yet persist in smoking. Consumers know they should avoid artery‐blocking trans fat, yet continue to order French fries at their nearby fast food outlet. Sexually active people understand the importance of safe sex practices but often fail to use condoms. Dissonance theory provides a mechanism for bringing behaviors into line with attitudes. Vicarious dissonance amplifies the opportunity for such change.
A. HYPOCRISY: SAYING WHAT I BELIEVE, BUT NOT WHAT I DID In this section, we begin with a brief description of a research paradigm that arouses personal dissonance and has been successful at bringing behaviors into line with attitudes. At the heart of the hypocrisy procedure is having
A THEORY OF VICARIOUS DISSONANCE
391
people make a statement that is consistent with their attitudes and then making them mindful that their past behaviors have been less than consistent— that is, that they have not always practiced what they just preached. The realization of hypocrisy creates dissonance and pressure to resolve the inconsistency. Research has shown that hypocrisy is typically resolved by people acting more strongly and forcefully in line with their attitudes. For example, Stone, Aronson, Crain, Winslow, and Fried (1994) showed that the hypocrisy paradigm could be used to have sexually active college students purchase condoms. The participants in the hypocrisy condition were asked to agree to make a videotape that would be shown to high school students advocating safe sex practices. The participants crafted their own speech and then delivered it in front of a video camera. After completing the speech (or thinking about such a speech), some of the participants were made mindful of times that they did not engage in safe sex practices. The experimenter explained that it was important to know more about the circumstances that made condom use diYcult. Participants read a list of common reasons that people typically give for failing to use condoms and were asked to identify circumstances that surrounded their own past failure to use condoms. In the nonhypocrisy conditions, the participants were asked merely to think about what they might say in a persuasive message about safe sex practices, but were not asked to make a public speech. In addition, nonhypocrisy participants were not made mindful of their past inconsistent behaviors. After finishing their participation, the students in all conditions were given an opportunity to purchase as many condoms as they wished, in total privacy, before leaving the study. Stone et al. (1994) predicted that the participants who were in the throes of cognitive dissonance would be motivated to alleviate their dissonance by behaving more in line with their attitudes. That is, participants in all conditions approved of using condoms for safe sex practice, but it was the participants who publicly advocated it and were made mindful of past occasions in which they did not use them, who would be motivated to purchase more condoms. The results of the condom purchases showed that more than 80% of the participants in the hypocrisy condition purchased condoms before leaving the building, whereas fewer than 50% purchased the condoms in the other conditions. A similar study by Stone, Wiegand, Cooper, and Aronson (1997) confirmed these results and showed that the personal dissonance was resolved specifically in the domain implicated by the hypocrisy. If participants were given a choice between showing their moral goodness through a donation to charity, they rejected it in favor of purchasing more condoms. That is, dissonance caused by hypocritical behavior is resolved by taking action in the specific domain in which the inconstancy occurred.
392
JOEL COOPER AND MICHAEL A. HOGG
The arousal of personal dissonance using the hypocrisy paradigm has been replicated frequently and has been used to influence behavior in a number of important domains. Dickerson, Thibodeau, Aronson, and Miller (1992) successfully used the hypocrisy paradigm to have university women take shorter showers during a drought emergency, and Fried and Aronson (1995) successfully influenced students to volunteer to make phone calls for an antilittering campaign. In addition, Son‐Hing, Li, and Zanna (2002) found that, in order to alleviate their hypocrisy, aversive racists reduced their prejudiced behaviors against minority students and, in France, Fointiat’s (2004) use of the hypocrisy paradigm led to greater compliance with antispeeding laws in Provence.
B. FROM HYPOCRISY TO VICARIOUS HYPOCRISY The conflict represented in the hypocrisy paradigm is an important one because it mirrors a frequent but understudied problem in our social world. This problem arises when people are already in agreement with an issue, believe they should act accordingly, but fail to do so. There is little point in persuading people to believe in using condoms to prevent HIV and other STDs; there is little point in convincing the smoker that cigarettes are dangerous; there is no need to convince the diabetic that eating sweet desserts is bad; nor any need to convince physicians to wash their hands before and after examining patients (see Dubner & Levitt, 2006). They already agree. The problem is influencing people to act consistently with their attitudes. And, of particular importance is whether a persuasion eVort can be successful at convincing a suYcient number of people to make a diVerence from the perspective of social policy. If the dissonance arising from the perception of hypocrisy can be experienced vicariously, it would constitute a potential agent for prosocial behavior change. As we have argued, the vicarious experience stems from our participation and membership in social groups. Therefore, we may experience dissonance whenever a member of our group acts hypocritically and it may encourage anyone witnessing the behavior to resolve their vicarious dissonance through increased performance of the prosocial behavior. In a series of studies (Hogg, Cooper, & MoYtt, 2006; Stone, Fernandez, Cooper, Cascio, & Hogg, 2007), we examined whether hypocrisy can be experienced vicariously and, if so, whether it could result in the change of behaviors and behavioral intentions to act in ways conducive to improved health. Both studies involved the use of sun block to protect against the risk of skin cancer. One study was conducted at the University of Queensland, located in a subtropical Australian state with a high incidence of skin cancer
A THEORY OF VICARIOUS DISSONANCE
393
and where students are well aware of the need to protect against the sun. Participants overheard a fellow student from UQ agree to make a speech advocating the use of sun block every time a person goes outdoors. In the hypocrisy condition, the speaker was made mindful of the occasions in which she failed to practice what she advocated. Participants overheard the experimenter ask the speaker if she had ever gone outdoors without using sun block. The experimenter explained that she was interested in understanding the reasons that people sometimes do not use sun block, even though they know it is excellent protection against the risk of skin cancer. Yes, the speaker admitted, in truth she did not use sun block on every occasion and then listed some reasonable excuses for failing to apply it. In the hypocrisy condition, then, participants heard a fellow group member advocate a pro‐attitudinal position and also heard the group member become mindful of the occasions in which she had not acted in accordance with that belief. In an advocacy‐only condition, the speaker agreed to make the speech favoring the use of sunscreen, but was not explicitly made mindful of the occasions in which she had not used it. We then asked the participants about their own behavioral intentions. We asked them to respond to the statement, ‘‘I intend to start carrying sun block with me wherever I go’’ on scale ranging from 1 (not very likely) to 9 (very likely). The results showed that women participants who listened to the hypocritical speaker indicated a greater intention to carry sun block compared to those who listened in the advocacy‐only condition. As predicted, the interaction between identification and hypocrisy was also significant. Vicarious hypocrisy occurred only for students who were highly identified with their group. In a similar experiment by Fernandez, Stone, Cooper, Cascio, and Hogg (2007), students at the University of Arizona were also asked to listen to a speech made by another student that encouraged people to use sun block as a preventative measure for skin cancer. The participants were told that the University of Arizona was collaborating with rival university, Arizona State University, on a project to develop eVective public service announcements designed to convince high school graduates to use sun block as protection against skin cancer. Flattering the participants, the experimenter praised the ability of college students to serve as more persuasive role models than doctors and public health oYcials. The experimenter explained that public service announcements had already been made by college students during the previous phase of the study, and the purpose of the current session was to oVer an assessment of the messages. Participants listened to a tape recording of a female student who made a strong speech advocating the use of sun block every day to protect against the threat of melanomas. Of course, all participants heard precisely
394
JOEL COOPER AND MICHAEL A. HOGG
the same speech but the cover story allowed us to vary whether the hypocritical speech maker was allegedly from the person’s in‐group (Arizona) or from the rival out‐group (Arizona State). The speech concluded with the statement, ‘‘No matter how busy you think you are with school or work, you can and should always wear sunscreen to reduce your risk of cancer.’’ In the target‐advocacy condition, the recording ended after the statement was complete. However, in the target‐hypocrisy condition, the strong statement at the end of the speech laid the groundwork for demonstrating the target’s inability to practice what she had preached. For participants in the target‐hypocrisy condition, the tape continued with the researcher explaining that it would be helpful to know more about why college students fail to use sun block every time they spend time in the sun. He indicated that researchers in the sun block program had made a list of common reasons people use for not applying sun block. The target asked to see the list and then responded, ‘‘Yeah, it’s true for me. I can see some of the major reasons why I don’t use sun block regularly right here on the top. I sometimes forget it in my car, or in the house. . .or I’m in too much of hurry to stop and put it on before I go out.’’ In this way, the target, who had already advocated the consistent use of sun block whenever one goes out of doors, publicly admitted to behavior that contradicted the statement. From our previous research, we predicted that the participant would experience vicarious dissonance from overhearing another student confess to the hypocrisy—but only if the fellow student was from the participant’s in‐group. The hypocrisy of students from the rival institution, Arizona State University, should not lead to vicarious hypocrisy. At the conclusion of the research session, participants were asked about their attitudes toward always using sun block and the strength of their intention to use sun block in the future. We concluded with a behavioral measure: All participants were given a coupon that they could redeem for free bottles of sun block. All they needed to do was to send a confirmation of their desire to have the free sample to the e‐mail address listed on the coupon. The e‐mail address belonged to the researchers and we were able to tally the number of people who actually tried to acquire the sun block. (And we did send them their free bottle of sun block!) The results of the study were exciting (at least for the women in the sample; as in the previously presented study at the University of Queensland, men seemed to want no part of having sunscreen). Female participants’ attitudes at the end of the study became more ardent that sunscreen should always be used—provided that the target speaker was from the in‐group and the participant felt highly attracted to that in‐group. Behavior was also aVected. As shown in Fig. 8, in the condition in which the University of Arizona participant felt highly identified with her university, witnessing the
A THEORY OF VICARIOUS DISSONANCE
395
70 60 50 40 30 20 10 0 Highly identified In-group
Low identified Out-group
Fig. 8. Percentage of women redeeming their coupons for sun screen.
in‐group speaker admit to her hypocrisy resulted in 70% of the participants e‐mailing their request for their complimentary sunscreen. By contrast, only 54% of the other participants bothered to reclaim their coupons.
VIII. Summary and Conclusions Cognitive dissonance theory has been in the social psychology literature for more than 50 years (see Cooper, 2007). Its elegance and counter‐intuitive twists have always intrigued researchers, students, and even the general public—‘‘cognitive dissonance’’ is one of only a handful of scientific experimental social psychological terms that have entered into common parlance. In this chapter we draw on social identity theory, now one of social psychology’s most influential theories of the relationship between self‐conception and the social group, to add a new twist to the dissonance story. We propose that people can experience cognitive dissonance vicariously, by merely witnessing another person behaving in ways that generate personal dissonance. We argue that this happens under conditions of intersubjectivity, where observers are able to experience the actor’s inner states as their own. Drawing on social identity theory, we argue that a potent cause of this intersubjectivity is identification with the same social group as the actor. Self‐categorization as a member of the same social group as the actor leads to group prototype‐based depersonalization of one’s perceptions, feelings,
396
JOEL COOPER AND MICHAEL A. HOGG
attitudes, and conduct. Self‐categorization eVectively replaces representation and definition of self and other as idiosyncratic and separate with a representation and definition of self and in‐group other as relatively ‘‘identical’’ or ‘‘psychologically fused’’—having shared prototypical attributes. This intersubjectivity is likely to be enhanced where self and/or other are highly group prototypical, where the shared identity is important or central to self‐ conception, where conditions of self‐conceptual uncertainty are high and in need of resolution through identification, and where the group that defines the shared identity is high in entitativity. Under these circumstances dissonance‐inducing events, such as being hypocritical or engaging in counter‐attitudinal advocacy, will be experienced vicariously by in‐group observers—and will have the same attitudinal and behavioral eVects on the observer as does personal dissonance on the actor. We reported a number of studies that provide support for vicarious dissonance theory. Using well‐known cognitive dissonance paradigms, including induced compliance and hypocrisy, we found that observation of a fellow in‐group member engaging in behavior that induces personal dissonance caused many of the eVects of dissonance to occur in the observer. We also found evidence that the observer actually experiences the discomfort of dissonance vicariously. In addition, we found that the vicarious dissonance eVect was strengthened when the actor was considered to be a more highly prototypical group member—but here there was an interesting twist in which the eVect was particularly pronounced among participants who identified weakly with the group and observed a prototypically marginal member acting counter‐attitudinally. We argued that in this latter case, the observer felt a particularly strong sense of shared identity with the actor on the basis of their shared marginal membership in the group—they represented a relatively salient minority subgroup. The idea of vicarious dissonance significantly extends the reach of cognitive dissonance theory. It shows that dissonance may have its eVects on attitudes and behavior in a much wider range of indirect social contexts where people merely witness others’ behavior in, for example, the mass media. One intriguing implication is that important health‐related attitudes might be changed on a massive scale by large‐scale interventions or mass media health promotion programs that induce dissonance vicariously by manipulating perceptions of group membership and group prototypicality. Indeed, one of the paradigms we reported to test vicarious dissonance predictions focused on sun‐protective attitudes and behaviors. Vicarious dissonance clearly exists and its occurrence seems to be influenced by social identity variables. Current research is focusing on collecting further evidence for vicarious discomfort and on mapping out more precisely the role of social identity variables such as prototypicality and uncertainty.
A THEORY OF VICARIOUS DISSONANCE
397
It will also be important to examine the fluid and dynamic system put in motion by vicarious dissonance. A number of group members in a face‐to‐ face context witness one of their own behaving counter‐attitudinally—they all experience dissonance vicariously and may make social comparisons and perhaps speak with one another. How does this aVect their feelings of dissonance, the way they may resolve the dissonance, and the subsequent nature of the group and their experience of group life and social identity? Perhaps the most intriguing question raised by the theory of vicarious dissonance is the underlying motivation for group members to change their attitudes and behaviors, following the observation of the inconsistent actions of a group member. We have speculated that the basic motivation will not be found solely in the reduction of personal discomfort, although that is part of the story. It will not be found solely in the wish to support fellow group members nor solely to empathize with the group member’s experience, although those may be parts of the story. Rather, we surmise that vicarious cognitive dissonance is primarily a group‐based phenomenon whose existence upsets the equilibrium of the social group. The disturbance of equilibrium renders social identity less stable and makes the group less functional as a source of self‐concept anchoring and uncertainty reduction. The story of vicarious dissonance may be fundamentally a story about people’s need for a strong, stable, and reliable group. Attitude change following a group member’s inconsistent behavior may be a way of restoring the integrity of the social group.
Acknowledgments We would like to acknowledge grant support from the National Cancer Institute, CA097354, and the Australian Research Council that allowed us to conduct some of the research we report and helped us collaborate in the writing of the chapter.
References Abrams, D., & Hogg, M. A. (1990). Social identification, self‐categorization and social influence. European Review of Social Psychology, 1, 195–228. Abrams, D., Hogg, M. A., & Marques, J. M. (Eds.) (2005). The social psychology of inclusion and exclusion (pp. 1–23). New York: Psychology Press. Aron, A., & Aron, E. N. (1996). Self and self‐expansion in relationships. In G. J. O. Fletcher & J. Fitness (Eds.), Knowledge structures in close relationships: A social psychological approach (pp. 325–344). Mahwah, NJ: Erlbaum.
398
JOEL COOPER AND MICHAEL A. HOGG
Aron, A., & McLaughlin‐Volpe, T. (2001). Including others in the self: Extensions to own and partner’s group memberships. In C. Sedikides & M. B. Brewer (Eds.), Individual self, relational self, and collective self: Partners, opponents, or strangers? (pp. 89–108). Philadelphia, PA: Psychology Press. Aronson, E. (1968). Dissonance theory: Progress and problems. In R. P. Abelson, E. Aronson, W. J. McGuire, T. M. Newcomb, M. J. Rosenberg, & P. H. Tannenbaum (Eds.), Theories of cognitive consistency: A sourcebook (pp. 5–27). Chicago: Rand‐McNally. Aronson, E., & Carlsmith, J. M. (1962). Performance expectancy as a determinant of actual performance. Journal of Abnormal and Social Psychology, 65, 178–182. Aronson, E., & Mills, J. (1959). The eVect of severity of initiation on liking for a group. Journal of Abnormal and Social Psychology, 59, 177–181. Asch, S. E. (1952). Social psychology. Englewood CliVs, NJ: Prentice‐Hall. Batson, C. D., Duncan, B., Ackerman, P., Buckley, T., & Birch, K. (1981). Is empathic emotion a source of altruistic motivation? Journal of Personality and Social Psychology, 40, 290–302. Batson, C. D., Early, S., & Salvarani, G. (1997). Perspective taking: Imagining how another feels versus imagining how you would feel. Personality and Social Psychology Bulletin, 23, 751–758. Batson, C. D., & Oleson, K. C. (1991). Current status of the empathy–altruism hypothesis. In M. S. Clark (Ed.), Prosocial behavior (pp. 62–85). Newbury Park, CA: Sage. Baumeister, R. F. (1998). The self. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th ed., Vol. 1, pp. 680–740). New York: McGraw‐Hill. Baumeister, R. F. (Ed.) (1999). The self in social psychology. Philadelphia, PA: Psychology Press. Beauvois, J. L., & Joule, R. V. (1999). A radical point of view on dissonance theory. In E. Harmon‐Jones & J. Mills (Eds.), Cognitive dissonance: Progress on a pivotal theory in social psychology (pp. 43–70). Washington, DC: American Psychological Association. Brehm, J. W. (1956). Postdecision changes in the desirability of alternatives. Journal of Abnormal and Social Psychology, 52, 384–389. Brewer, M. B., & Gardner, W. (1996). Who is this ‘We’? Levels of collective identity and self representation. Journal of Personality and Social Psychology, 71, 83–93. Cadinu, M. R., & Rothbart, M. (1996). Self‐anchoring and diVerentiation processes in the minimal group setting. Journal of Personality and Social Psychology, 70, 661–677. Chong, J., & Cooper, J. (2007). Cognitive dissonance and vicarious cognitive dissonance in East Asia: Can I feel your discomfort but not my own? Poster presented at the Society of Personality and Social Psychology meeting, January, 2007. Memphis, TN. Cooper, J. (1971). Personal responsibility and dissonance: The role of foreseen consequences. Journal of Personality and Social Psychology, 18, 354–363. Cooper, J. (2003). Cultural aspects of cognitive dissonance. Paper presented at Social Cognition Conference in Budapest, Hungary, July, 2003. Cooper, J. (2007). Cognitive dissonance: 50 years of a classic theory. Thousand Oaks, CA: Sage. Cooper, J., & Fazio, R. H. (1984). A new look at dissonance theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 17, pp. 229–266). San Diego, CA: Academic Press. Cooper, J., & Hogg, M. A. (2002). Dissonance arousal and the collective self: Vicarious experience of dissonance based on shared group membership. In J. P. Forgas & K. D. Williams (Eds.), The social self: Cognitive, interpersonal, and intergroup perspectives (pp. 327–341). New York: Psychology Press. Cooper, J., & Mackie, D. (1983). Cognitive dissonance in an intergroup context. Journal of Personality and Social Psychology, 44, 536–544.
A THEORY OF VICARIOUS DISSONANCE
399
Cooper, J., & Stone, J. (2000). Cognitive dissonance and the social group. In D. J. Terry & M. A. Hogg (Eds.), Attitudes, behavior, and social context: The role of norms and group membership (pp. 227–244). Mahwah, NJ: Lawrence Erlbaum Associates Publishers. Cooper, J., & Worchel, S. (1970). Role of undesired consequences in arousing cognitive dissonance. Journal of Personality and Social Psychology, 16, 199–206. Cooper, J., Zanna, M. P., & Taves, P. A. (1978). Arousal as a necessary condition for attitude change following induced compliance. Journal of Personality and Social Psychology, 36, 110–1106. Croyle, R. T., & Cooper, J. (1983). Dissonance arousal: Physiological evidence. Journal of Personality and Social Psychology, 45, 782–791. Davis, M. H. (1983). Measuring individual diVerences in empathy: Evidence for a multidimensional approach. Journal of Personality and Social Psychology, 44, 113–126. Davis, K. E., & Jones, E. E. (1960). Changes in interpersonal perception as a means of reducing cognitive dissonance. Journal of Abnormal and Social Psychology, 61, 402–410. Dickerson, C. A., Thibodeau, R., Aronson, E., & Miller, D. (1992). Using cognitive dissonance to encourage water conservation. Journal of Applied Social Psychology, 22(11), 841–854. Dubner, S. J., & Levitt, S. D. (2006). Selling soap: How do you get doctors to wash their hands? New York Times Magazine, September 24, 2006, 22–23. Elkin, R. A., & Leippe, M. R., (1986). Physiological arousal, dissonance, and attitude change: Evidence for a dissonance‐arousal link and a ‘‘Don’t remind me’’ eVect. Journal of Personality and Social Psychology, 51(1), 5–65. Elliot, A. J., & Devine, P. G. (1994). On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort. Journal of Personality and Social Psychology, 67, 382–394. Erlich, D., Guttman, J., Schonbach, P., & Mills, J. (1957). Postdecision exposure to relevant information. Journal of Abnormal and Social Psychology, 54, 98–102. Fernandez, N., Stone, J., Cooper, J., Cassio, E., & Hogg, M. A. (2007). Vicarious hypocrisy: The use of attitude bolstering after exposure to a hypocritical ingroup member. Poster session at Society of Personality and Social Psychology. Memphis, TN, January, 2007. Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203–210. Festinger, L., Riecken, H. W., & Schachter, S. (1956). When prophecy fails. Minneapolis: University of Minnesota Press. Fointiat, V. (2004). ‘‘I know what I have to do, but . . . .’’ When hypocrisy leads to behavioral change. Social Behavior and Personality, 32, 741–746. Fried, C. B., & Aronson, E. (1995). Hypocrisy, misattribution, and dissonance reduction. Personality and Social Psychology Bulletin, 21, 925–933. Galinsky, A. D., & Ku, G. (2004). The eVects of perspective‐taking on prejudice: The moderating role of self‐evaluation. Personality and Social Psychology Bulletin, 30, 594–604. Galinsky, A. D., Ku, G., & Wang, C. S. (2005). Perspective‐taking and self‐other overlap: Fostering social bonds and facilitating social coordination. Group Processes and Intergroup Relations, 8, 109–124. Galinsky, A. D., & Moskowitz, G. B. (2000). Perspective‐taking: Decreasing stereotype expression, stereotype accessibility, and in‐group favoritism. Journal of Personality and Social Psychology, 78, 708–724. Galinsky, A. D., Stone, J., & Cooper, J. (2000). The reinstatement of dissonance and psychological discomfort following failed aYrmations. European Journal of Social Psychology, 30, 123–147.
400
JOEL COOPER AND MICHAEL A. HOGG
Goethals, G. R., Cooper, J., & Naficy, A. (1979). Role of foreseen, foreseeable, and unforeseeable behavioral consequences in the arousal of cognitive dissonance. Journal of Personality and Social Psychology, 37, 1179–1185. Gordijn, E. H., Yzerbyt, V., Wigboldus, D., & Dumont, M. (2006). Emotional reactions to harmful intergroup behavior. European Journal of Social Psychology, 36, 15–30. Harmon‐Jones, E. (1999). Toward an understanding of the motivation underlying dissonance eVects: Is the production of aversive consequences necessary? In E. Harmon‐Jones & J. Mills (Eds.), Cognitive dissonance: Progress on a pivotal theory in social psychology (pp. 71–99). Washington, DC: American Psychological Association. Harmon‐Jones, E., Brehm, J. W., Greenberg, J., Simon, L., & Nelson, D. E. (1996). Evidence that the production of aversive consequences is not necessary to create cognitive dissonance. Journal of Personality and Social Psychology, 70, 5–16. Harmon‐Jones, E. & Mills, J. (Eds.) (1999). Cognitive dissonance: Progress on a pivotal theory in social psychology. Washington, DC: American Psychological Association. Haslam, S. A., Oakes, P. J., McGarty, C., Turner, J. C., & Onorato, S. (1995). Contextual changes in the prototypicality of extreme and moderate outgroup members. European Journal of Social Psychology, 25, 509–530. Hatfield, E., Cacioppo, J. T., & Rapson, R. L. (Eds.) (1994). Emotional contagion. Studies in emotion and social interaction. New York: Cambridge University Press. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Heine, S. J., & Lehman, D. R. (1997). Culture, dissonance, and self‐aYrmation. Personality and Social Psychology Bulletin, 23, 389–400. Hofstede, G. (1980). Culture’s consequences: International diVerences in work‐related values. Beverly Hills, CA: Sage. Hogg, M. A. (2000). Subjective uncertainty reduction through self‐categorization: A motivational theory of social identity processes. European Review of Social Psychology, 11, 223–255. Hogg, M. A. (2003). Social identity. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 462–479). New York: Guilford. Hogg, M. A. (2005a). The social identity perspective. In S. A. Wheelan (Ed.), The handbook of group research and practice (pp. 33–157). Thousand Oaks, CA: Sage. Hogg, M. A. (2005b). All animals are equal but some animals are more equal than others: Social identity and marginal membership. In K. D. Williams, J. P. Forgas, & W. von Hippel (Eds.), The social outcast: Ostracism, social exclusion, rejection, and bullying (pp. 243–261). New York: Psychology Press. Hogg, M. A. (2006). Social identity theory. In P. J. Burke (Ed.), Contemporary social psychological theories (pp. 111–136). Palo Alto, CA: Stanford University Press. Hogg, M. A. (2007). Uncertainty‐identity theory. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 39, pp. 69–126). San Diego, CA: Academic Press. Hogg, M. A., & Abrams, D. (1988). Social identifications: A social psychology of intergroup relations and group processes. London: Routledge. Hogg, M. A., & Cooper, J. (2006). Group prototypicality, social identity and vicarious dissonance. Unpublished manuscript, Claremont Graduate University. Hogg, M. A., Cooper, J., & MoYtt, G. (2006). Sun‐Smart: The eVect of vicarious dissonance on decisions to use sunscreen. Unpublished manuscript, Claremont Graduate University. Hogg, M. A., Cooper‐Shaw, L., & Holzworth, D. W. (1993). Group prototypicality and depersonalized attraction in small interactive groups. Personality and Social Psychology Bulletin, 19, 452–465. Hogg, M. A., & Hains, S. C. (1996). Intergroup relations and group solidarity: EVects of group identification and social beliefs on depersonalized attraction. Journal of Personality and Social Psychology, 70, 295–309.
A THEORY OF VICARIOUS DISSONANCE
401
Hogg, M. A., Hains, S. C., & Mason, I. (1998). Identification and leadership in small groups: Salience, frame of reference, and leader stereotypicality eVects on leader evaluations. Journal of Personality and Social Psychology, 75, 1248–1263. Hogg, M. A., & Smith, J. R. (in press). Attitudes in social context: A social identity perspectives. European Review of Social Psychology. Hogg, M. A., & van Knippenberg, D. (2003). Social identity and leadership processes in groups. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 35, pp. 1–52). San Diego, CA: Academic Press. Hoshino‐Browne, E., Zanna, A. S., Spencer, S. J., Zanna, M. P., Kitayama, S., & Lackenbauer, S. (2005). On the cultural guises of cognitive dissonance: The case of Easterners and Westerners. Journal of Personality and Social Psychology, 89, 294–310. Kitayama, S., Ishii, K., Imada, T., Takemura, K., & Ramaswamy, J. (2006). Voluntary settlement and the spirit of independence: Evidence from Japan’s ‘‘northern frontier.’’ Journal of Personality and Social Psychology, 91, 369–384. Kitayama, S., Snibbe, A. C., Markus, H. R., & Suzuki, T. (2004). Is there any ‘‘free’’ choice? Self and dissonance in two cultures. Psychological Science, 15, 527–533. La Pierre, R. T., & Farnsworth, P. R. (1949). Social psychology. New York: McGraw‐Hill. Leary, M. R. & Tangney, J. P. (Eds.) (2003). Handbook of self and identity. New York: Guilford. Lickel, B., Schmader, T., Curtis, M., Scarnier, M., & Ames, D. R. (2005). Vicarious shame and guilt. Group Processes and Intergroup Relations, 8, 145–157. Linder, D. E., Cooper, J., & Jones, E. E. (1967). Decision freedom as a determinant of the role of incentive magnitude in attitude change. Journal of Personality and Social Psychology, 6, 245–254. Linder, D. E., Cooper, J., & Wicklund, R. A. (1968). Pre‐exposure persuasion as a result of commitment to pre‐exposure eVort. Journal of Experimental Social Psychology, 4, 470–482. Lipps, T. (1926). Psychological studies. Baltimore, MD: Wiliams & Wilkins. Losch, M. E., & Cacioppo, J. T. (1990). Cognitive dissonance may enhance sympathetic tonus, but attitudes are changed to reduce negative aVect rather than arousal. Journal of Experimental Social Psychology, 26, 289–304. Mackie, D. M., Devos, T., & Smith, E. R. (1999). Intergroup emotions: Explaining oVensive action tendencies in an intergroup context. Journal of Personality and Social Psychology, 79, 602–616. Mackie, D. M., Silver, L. A., & Smith, E. R. (2004). Intergroup emotions: Emotion as an intergroup phenomenon. In L. Z. Tiedens & C. W. Leach (Eds.), The social life of emotions (pp. 227–245). New York: Cambridge University Press. Mackie, D. M., & Smith, E. R. (2002). Intergroup emotions and the social self: Prejudice reconceptualized as diVerentiated reactions to out groups. In J. P. Forgas & K. D. Williams (Eds.), The social self: Cognitive, interpersonal, and intergroup perspectives (pp. 309–326). New York: Psychology Press. Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in‐group messages. Journal of Personality and Social Psychology, 58, 812–822. Maitner, A. T., Mackie, D. M., & Smith, E. R. (2007). Antecedents and consequences of satisfaction and guilt following in‐group aggression. Group Processes and Intergroup Relations. (In press). Maner, J. K., Luce, C. L., Neuberg, S. L., Cialdini, R. B., Brown, S., & Sagarin, B. J. (2002). The eVects of perspective taking on motivations for helping: Still no evidence for altruism. Personality and Social Psychology Bulletin, 28, 1601–1610. Marks, G., & Miller, N. (1987). Ten years of research on the false‐consensus eVect: An empirical and theoretical review. Psychological Bulletin, 102, 72–90. Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253.
402
JOEL COOPER AND MICHAEL A. HOGG
McDougall, W. (1908). An introduction to social psychology. New York: Methuen. Miller, J. G. (1984). Culture and the development of everyday social explanation. Journal of Personality and Social Psychology, 46, 961–978. Monin, B., Norton, M. I., Cooper, J., & Hogg, M. A. (2004). Reacting to an assumed situation vs. conforming to an assumed reaction: The role of perceived speaker attitude in vicarious dissonance. Group Processes and Intergroup Relations, 7, 207–220. Mugny, G., & Pe´rez, J. A. (1991). The social psychology of minority influence. Cambridge, UK: Cambridge University Press. Norton, M. I., Monin, B., Cooper, J., & Hogg, M. A. (2003). Vicarious dissonance: Attitude change from the inconsistency of others. Journal of Personality and Social Psychology, 85, 47–62. Oakes, P. J., Haslam, S. A., & Turner, J. C. (1994). Stereotyping and social reality. Oxford, UK: Blackwell. Oswald, P. A. (1996). The eVects of cognitive and aVective perspective taking on empathic concern and altruistic helping. Journal of Social Psychology, 136, 613–623. Otten, S. (2002). ‘‘Me and us’’ or ‘‘us and them’’? The self as a heuristic for defining minimal ingroups. European Review of Social Psychology, 13, 1–33. Oyserman, D., Kemmelmeier, M., & Coon, H. M. (2002). Cultural psychology, a new look: Reply to Bond (2002), Fiske (2002), Kitayama (2002), and Miller (2002). Psychological Bulletin, 128, 110–117. Reicher, S. D. (1984). Social influence in the crowd: Attitudinal and behavioural eVects of de‐ individuation in conditions of high and low group salience. British Journal of Social psychology, 23, 341–350. Reicher, S. D., & Hopkins, N. (2003). On the science of the art of leadership. In D. van Knippenberg & M. A. Hogg (Eds.), Leadership and power: Identity processes in groups and organizations (pp. 197–209). London: Sage. Robbins, J. M., & Krueger, J. L. (2005). Social projection to ingroups and out groups: A review and meta‐analysis. Personality and Social Psychology Review, 9, 32–47. Sakai, H. (1999). A multiplicative power‐function model of cognitive dissonance: Toward an integrated theory of cognition, emotion and behavior after Leon Festinger. In E. Harmon‐ Jones & J. Mills (Eds.), Cognitive dissonance theory: Progress on a pivotal theory in social psychology (pp. 267–294). Washington, DC: American Psychological Association. Scher, S. J., & Cooper, J. (1989). Motivational basis of dissonance: The singular role of behavioral consequences. Journal of Personality and Social Psychology, 56, 899–906. Sedikides, C., & Gregg, A. P. (2003). Portraits of the self. In M. A. Hogg & J. Cooper (Eds.), The Sage handbook of social psychology (pp. 110–138). London: Sage. Sherman, D. K., & Cohen, G. L. (2006). The psychology of self‐defense: Self‐aYrmation theory. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 38, pp. 183–242). San Diego, CA: Academic Press. Son‐Hing, L. S., Li, W., & Zanna, M. P. (2002). Inducing hypocrisy to reduce prejudicial responses among aversive racists. Journal of Experimental Social Psychology, 38, 71–78. Steele, C. M. (1988). The psychology of self‐aYrmation: Sustaining the integrity of the self. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp. 261–302). New York: Academic Press. Stone, J., Aronson, E., Crain, A. L., Winslow, M. P., & Fried, C. B. (1994). Inducing hypocrisy as a means of encouraging young adults to use condoms. Personality and Social Psychology Bulletin, 20, 116–128. Stone, J., & Cooper, J. (2001). A self standards model of cognitive dissonance. Journal of Experimental Social Psychology, 37, 228–243.
A THEORY OF VICARIOUS DISSONANCE
403
Stone, J., Wiegand, A. W., Cooper, J., & Aronson, E. (1997). When exemplification fails: Hypocrisy and the motive for self‐integrity. Journal of Personality and Social Psychology, 72, 54–65. Tajfel, H. (1972). Social categorization. English manuscript of ‘La cate´gorisation sociale’. In S. Moscovici (Ed.), Introduction a` la psychologie sociale (Vol. 1, pp. 272–302). Paris: Larousse. Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole. Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude‐behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22, 776–793. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self‐categorization theory. Oxford, UK: Blackwell. Turner, J. C., & Oakes, P. J. (1989). Self‐categorization and social influence. In P. B. Paulus (Ed.), The psychology of group influence (2nd ed., pp. 233–275). Hillsdale, NJ: Erlbaum. Turner, J. C., Oakes, P. J., Haslam, S. A., & McGarty, C. A. (1994). Self and collective: Cognition and social context. Personality and Social Psychology Bulletin, 20, 454–463. Vescio, T. K., Sechrist, G. B., & Paolucci, M. P. (2003). Perspective taking and prejudice reduction: The mediational role of empathy arousal and situational attributions. European Journal of Social Psychology, 33, 455–472. Waterman, C. K., & Katkin, E. S. (1967). Energizing (dynamogenic) eVect of cognitive dissonance on task performance. Journal of Personality and Social Psychology, 6, 126–131. Wright, S. C., Aron, A., & Tropp, L. R. (2002). Including others (and groups) in the self: Self‐expansion and intergroup relations. In J. P. Forgas & K. D. Williams (Eds.), The social self: Cognitive, interpersonal, and intergroup perspectives (pp. 343–363). New York: Psychology Press. Zanna, M. P., & Cooper, J. (1974). Dissonance and the pill: An attribution to studying the arousal properties of dissonance. Journal of Personality and Social Psychology, 29, 703–709. Zanna, M. P., & Cooper, J. (1976). Dissonance and the attribution process. In J. Harvey, W. Ickes, & R. Kidd (Eds.), New directions in attribution research (pp. 199–217). Hillsdale, NJ: Lawrence Erlbaum Associates. Zanna, M. P., & Sande, G. N. (1987). The eVects of collective actions on the attitudes of individual group members: A dissonance analysis. In M. P. Zanna, J. M. Olson, & C. P. Herman (Eds.), Social influence: The Ontario Symposium (Vol. 5, pp. 151–163). Hillsdale, NJ: Lawrence Erlbaum Associates.
CONTENTS Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ix
Culture and the Structure of Personal Experience: Insider and Outsider Phenomenologies of the Self and Social World Dov Cohen, Etsuko Hoshino‐Browne, and Angela K.‐y. Leung I. II. III. IV. V. VI. VII. VIII.
IX. X. XI.
Introduction.. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . A Sociofunctional Account of Perception. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Relational Versus Egocentric Projection .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Memory Imagery . .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Online Imagery . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Mental Models of the Self and Others in Narrative . . . . . .. . .. . . .. . .. . . .. . .. . . .. . Confusing What Is in One’s Own Head and What Is Out There. . . . . .. . .. . . .. . Confusing What Is in One’s Own Head with What Is in Other People’s Heads: The Illusion of One’s Own Transparency and Empathy‐as‐Projection . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Projection‐as‐Empathy in a Group Setting . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Characterizing the World . . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . General Discussion . . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .
1 3 13 16 18 21 28
31 36 38 41 59
Uncertainty–Identity Theory Michael A. Hogg I. II. III. IV. V. VI.
Historical Background . . .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Uncertainty. .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Social Identity. . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Entitativity . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Social Extremism and Totalistic Groups .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Extensions, Applications, and Implications of Uncertainty–Identity Theory . . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .
v
70 73 79 87 91 97
vi
CONTENTS
VII. Uncertainty–Identity Theory in Relation to Other Ideas . . . .. . .. . . .. . .. . . .. . . .. . VIII. Concluding Comments . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .
105 111 114
Metacognitive Experiences and the Intricacies of Setting People Straight: Implications for Debiasing and Public Information Campaigns Norbert Schwarz, Lawrence J. Sanna, Ian Skurnik, and Carolyn Yoon I. II. III. IV.
Introduction . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Metacognitive Experiences . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Accessibility Experiences and the Emergence and Attenuation of Bias. . . . . . .. . Fluency, Familiarity, and Truth: Implications for Public Information Campaigns . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . V. Implications and Future Directions . . .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .
127 129 132 143 153 157
Multiple Social Categorization Richard J. Crisp and Miles Hewstone I. II. III. IV. V.
Introduction . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . First Principles and Assumptions. . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . DiVerentiation–Discrimination. .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Decategorization . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Conclusions . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .
164 170 177 218 234 239
On the Parameters of Human Judgment Arie W. Kruglanski, Antonio Pierro, Lucia Mannetti, Hans‐Peter Erb, and Woo Young Chun I. II. III. IV. V.
Introduction . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . The Role of Rules in Judgment Formation .. . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . The Parameters of Human Judgment .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . A Parametric Model of Social Judgment . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . Recapitulation and Conclusions . . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . References. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. .
255 257 265 275 291 297
CONTENTS
vii
Panglossian Ideology in the Service of System Justification: How Complementary Stereotypes Help Us to Rationalize Inequality Aaron C. Kay, John T. Jost, Anesu N. Mandisodza, Steven J. Sherman, John V. Petrocelli, and Amy L. Johnson I. II. III. IV.
Introduction.. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . System Justification Theory . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . The System‐Justifying Function of Complementary Stereotypes . . . . . .. . .. . . .. . Moderators of the EVect of Complementary Stereotypes on System Justification. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . V. Implicit Complementary Versus Noncomplementary Stereotypical Associations . . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . VI. Concluding Remarks: SJT and Stereotyping as Rationalization. . . . . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .
306 308 312 335 346 350 352
Feeling the Anguish of Others: A Theory of Vicarious Dissonance Joel Cooper and Michael A. Hogg I. II. III. IV. V. VI. VII.
Introduction.. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Personal Cognitive Dissonance. . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Social Identity and Vicarious Dissonance. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Evidence for Vicarious Dissonance . . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . The Process of Vicarious Dissonance. . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Vicarious Dissonance and the Collective .. . . . . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . Vicarious Hypocrisy: Translating Dissonance into Actions and Attitudes to Improve Health . . . . . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . VIII. Summary and Conclusions . . . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . References . . .. . . . . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . .. . . .. . .. . . .. . .. . . .. .
360 361 365 368 381 384
Index ........................................................................................... Contents of Other Volumes ...............................................................
405 419
390 395 397
CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin.
WOO YOUNG CHUN (255), Department of Psychology, Hallym University, Chuncheon, Gangwon-do, 200-702 Korea DOV COHEN (1), Department of Psychology, University of Illinois, Champaign, IL 61820 JOEL COOPER (359), Department of Psychology, Princeton University, Princeton, NJ 08540 RICHARD J. CRISP (163), School of Psychology, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom HANS-PETER ERB (255), Wirtschafts-, Organisations- und Sozialpsychologie, Chemnitz University of Technology, 09120 Chemnitz, Germany MILES HEWSTONE (163), Department of Experimental Psychology, University of Oxford, Oxford OX1 3UD, United Kingdom MICHAEL A. HOGG (69, 359), School of Behavioral and Organizational Sciences, Claremont Graduate University, Claremont, CA 91711 ETSUKO HOSHINO-BROWNE (1), Department of Psychology, Swarthmore College, Swarthmore, PA 19081 AMY L. JOHNSON (305), International Survey Research, Chicago, IL 60611 JOHN T. JOST (305), Department of Psychology, New York University, New York, NY 10003 AARON C. KAY (305), Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1 ARIE W. KRUGLANSKI (255), Department of Psychology, University of Maryland, MD 20742 ANGELA K.-y. LEUNG (1), Department of Psychology, University of Illinois, Champaign, IL 61820
ix
x
CONTRIBUTORS
ANESU N. MANDISODZA (305), Department of Psychology, New York University, New York, NY 10003 LUCIA MANNETTI (255), Dipartimento di Psicologia dei Processi di Sviluppo e Socializzazione, Universita` di Roma La Sapienza, 00185 Roma, Italy JOHN V. PETROCELLI (305), Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN 47405 ANTONIO PIERRO (255), Dipartimento di Psicologia dei Processi di Sviluppo e Socializzazione, Universita` di Roma La Sapienza, 00185 Roma, Italy LAWRENCE J. SANNA (127), Department of Psychology, University of North Carolina, Chapel Hill, NC 27599 NORBERT SCHWARZ (127), Institute for Social Research, University of Michigan, Ann Arbor, MI 48106 STEVEN J. SHERMAN (305), Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN 47405 IAN SKURNIK (127), Darden School of Business, University of Virginia, Charlottesville, VA 22906 CAROLYN YOON (127), Stephen M. Ross School of Business, University of Michigan, Ann Arbor, MI 48109
CONTENTS OF OTHER VOLUMES
Volume 1 Cultural Influences upon Cognitive Processes Harry C. Triendis The Interaction of Cognitive and Physiological Determinants of Emotional State Stanley Schachter Experimental Studies of Coalirion Formation William A. Gamson Communication Networks Marvin E. Shaw A Contingency Model of Leadership Effectiveness Fred E. Fiedler Inducing Resistance to Persuasion: Some Contemporary Approaches William J. McGuire Social Motivation, Dependency, and Susceptibility to Social Influence Richard H. Walters and Ross D. Purke Sociability and Social Organization in Monkeys and Apes William A. Mason Author Index—Subject Index Volume 2 Vicarious Processes: A Case of No-Trial Learning Albert Bandura Selective Exposure Jonathan L. Freedman and David O. Sears Group Problem Solving L. Richard Hoffman Situational Factors in Conformity Vernon L. Allen Social Power John Schopler
From Acts to Dispositions: The Attribution Process in Person Perception Edward E. Jones and Keith E. Davis Inequality in Social Exchange J. Stacy Adams The Concept of Aggressive Drive: Some Additional Considerations Leonard Berkowitz Author Index—Subject Index Volume 3 Mathematical Models in Social Psychology Robert P. Abelson The Experimental Analysis of Social Performance Michael Argyle and Adam Kendon A Structural Balance Approach to the Analysis of Communication Effects N. T. Feather Effects of Fear Arousal on Attitude Change: Recent Developments in Theory and Experimental Research Irving L. Janis Communication Processes and the Properties of Language Serge Moscovici The Congruity Principle Revisited: Studies in the Reduction, Induction, and Generalization of Persuasion Percy H. Tannenbaum Author Index—Subject Index Volume 4 The Theory of Cognitive Dissonance: A Current Perspective Elliot Aronson
419
420
CONTENTS OF OTHER VOLUMES
Attitudes and Attraction Donn Byrne Sociolinguistics Susan M. Ervin-Tripp Recognition of Emotion Nico H. Frijda Studies of Status Congruence Edward E. Sampson Exploratory Investigations of Empathy Ezra Stotland The Personal Reference Scale: An Approach to Social Judgment Harry S. Upshaw Author Index—Subject Index Volume 5 Media Violence and Aggressive Behavior: A Review of Experimental Research Richard E. Goranson Studies in Leader Legitimacy, Influence, and Innovation Edwin P. Hollander and James W. Julian Experimental Studies of Negro-White Relationships Irwin Katz Findings and Theory in the Study of Fear Communications Howard Leventhal Perceived Freedom Ivan D. Steiner Experimental Studies of Families Nancy E. Waxler and Elliot G. Mishler Why Do Groups Make Riskier Decisions than Individuals? Kenneth L. Dion, Robert S. Baron, and Norman Miller Author Index—Subject Index Volume 6 Self-Perception Theory Daryl J. Bem Social Norms, Feelings, and Other Factors Affecting Helping and Altruism Leonard Berkowitz The Power of Liking: Consequence of Interpersonal Attitudes Derived from a Liberalized View of Secondary Reinforcement Albert J. Lott and Bernice E. Lott
Social Influence, Conformity Bias, and the Study of Active Minorities Serge Moscovici and Claude Faucheux A Critical Analysis of Research Utilizing the Prisoner’s Dilemma Paradigm for the Study of Bargaining Charlan Nemeth Structural Representations of Implicit Personality Theory Seymour Rosenberg and Andrea Sedlak Author Index—Subject Index Volume 7 Cognitive Algebra: Integration Theory Applied to Social Attribution Norman A. Anderson On Conflicts and Bargaining Erika Apfelbaum Physical Attractiveness Ellen Bersheid and Elaine Walster Compliance, Justification, and Cognitive Change Harold B. Gerard, Edward S. Connolley, and Roland A. Wilhelmy Processes in Delay of Gratification Walter Mischel Helping a Distressed Person: Social, Personality, and Stimulus Determinants Ervin Staub Author Index—Subject Index Volume 8 Social Support for Nonconformity Vernon L. Allen Group Tasks, Group Interaction Process, and Group Performance Effectiveness: A Review and Proposed Integration J. Richard Hackman and Charles G. Morris The Human Subject in the Psychology Experiment: Fact and Artifact Arie W. Kruglanski Emotional Arousal in the Facilitation of Aggression Through Communication Percy H. Tannenbaum and Dolf Zillman The Reluctance to Transmit Bad News Abraham Tesser and Sidney Rosen Objective Self-Awareness Robert A. Wicklund
CONTENTS OF OTHER VOLUMES Responses to Uncontrollable Outcomes: An Integration of Reactance Theory and the Learned Helplessness Model Camille B. Wortman and Jack W. Brehm Subject Index Volume 9 New Directions in Equity Research Elaine Walster, Ellen Berscheid, and G. William Walster Equity Theory Revisited: Comments and Annotated Bibliography J. Stacy Adams and Sara Freedman The Distribution of Rewards and Resources in Groups and Organizations Gerald S. Leventhal Deserving and the Emergence of Forms of Justice Melvin J. Lerner, Dale T. Miller, and John G. Holmes Equity and the Law: The Effect of a Harmdoer’s ‘‘Suffering in the Act’’ on Liking and Assigned Punishment William Austin, Elaine Walster, and Mary Kristine Utne Incremental Exchange Theory: A Formal Model for Progression in Dyadic Social Interaction L. Lowell Huesmann and George Levinger Commentary George C. Homans Subject Index Volume 10 The Catharsis of Aggression: An Evaluation of a Hypothesis Russell G. Geen and Michael B. Quanty Mere Exposure Albert A. Harrison Moral Internalization: Current Theory and Research Martin L. Hoffman Some Effects of Violent and Nonviolent Movies on the Behavior of Juvenile Delinquents Ross D. Parke, Leonard Berkowitz, Jacques P. Leyens, Stephen G. West, and Richard Sebastian
421
The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process Less Ross Normative Influences on Altruism Shalom H. Schwartz A Discussion of the Domain and Methods of Social Psychology: Two Papers by Ron Harre and Barry R. Schlenker Leonard Berkowitz The Ethogenic Approach: Theory and Practice R. Harre On the Ethogenic Approach: Etiquette and Revolution Barry R. Schlenker Automatisms and Autonomies: In Reply to Professor Schlenker R. Harre Subject Index Volume 11 The Persistence of Experimentally Induced Attitude Change Thomas D. Cook and Brian F. Flay The Contingency Model and the Dynamics of the Leadership Process Fred E. Fiedler An Attributional Theory of Choice Andy Kukla Group-Induced Polarization of Attitudes and Behavior Helmut Lamm and David G. Myers Crowding: Determinants and Effects Janet E. Stockdale Salience: Attention, and Attribution: Top of the Head Phenomena Shelley E. Taylor and Susan T. Fiske Self-Generated Attitude Change Abraham Tesser Subject Index Volume 12 Part I. Studies in Social Cognition Prototypes in Person Perception Nancy Cantor and Walter Mischel A Cognitive-Attributional Analysis of Stereotyping David L. Hamilton
422
CONTENTS OF OTHER VOLUMES
Self-Monitoring Processes Mark Snyder Part II. Social Influences and Social Interaction Architectural Mediation of Residential Density and Control: Crowding and the Regulation of Social Contact Andrew Baum and Stuart Valins A Cultural Ecology of Social Behavior J. W. Berry Experiments on Deviance with Special Reference to Dishonesty David P. Farrington From the Early Window to the Late Night Show: International Trends in the Study of Television’s Impact on Children and Adults John P. Murray and Susan Kippax Effects of Prosocial Television and Film Material on the Behavior of Viewers J. Phillipe Rushton Subject Index
Cognitive, Social, and Personality Processes in the Physiological Detection of Deception William M. Waid and Martin T. Orne Dialectic Conceptions in Social Psychology: An Application to Social Penetration and Privacy Regulation Irwin Altman, Anne Vinsel, and Barbara B. Brown Direct Experience and Attitude–Behavior Consistency Russell H. Fazio and Mark P. Zanna Predictability and Human Stress: Toward a Clarification of Evidence and Theory Suzanne M. Miller Perceptual and Judgmental Processes in Social Contexts Arnold Upmeyer Jury Trials: Psychology and Law Charlan Jeanne Nemeth Index Volume 15
Volume 13 People’s Analyses of the Causes of AbilityLinked Performances John M. Darley and George R. Goethals The Empirical Exploration of Intrinsic Motivational Processes Edward I. Deci and Richard M. Ryan Attribution of Responsibility: From Man the Scientist to Man as Lawyer Frank D. Fincham and Joseph M. Jaspars Toward a Comprehensive Theory of Emotion Howard Leventhal Toward a Theory of Conversion Behavior Serge Moscovici The Role of Information Retrieval and Conditional Inference Processes in Belief Formation and Change Robert S. Wyer, Jr. and Jon Hartwick Index
Balance, Agreement, and Positivity in the Cognition of Small Social Structures Walter H. Crockett Episode Cognition: Internal Representations of Interaction Routines Joseph P. Forgas The Effects of Aggressive-Pornographic Mass Media Stimuli Neil M. Malamuth and Ed Donnerstein Socialization in Small Groups: Temporal Changes in Individual–Group Relations Richard L. Moreland and John M. Levine Translating Actions into Attitudes: An Identity-Analytic Approach to the Explanation of Social Conduct Barry R. Schlenker Aversive Conditions as Stimuli to Aggression Leonard Berkowitz Index
Volume 14
Volume 16
Verbal and Nonverbal Communication of Deception Miron Zuckerman, Bella M. DePaulo, and Robert Rosenthal
A Contextualist Theory of Knowledge: Its Implications for Innovation and Reform in Psychological Research William J. McGuire
CONTENTS OF OTHER VOLUMES Social Cognition: Some Historical and Theoretical Perspectives Janet Landman and Melvin Manis Paradigmatic Behaviorism: Unified Theory for Social-Personality Psychology Arthur W. Staats Social PsychologyfromtheStandpointofa StructuralSymbolicInteractionism: Toward an InterdisciplinarySocialPsychology Sheldon Stryker Toward an Interdisciplinary Social Psychology Carl W. Backman Index Volume 17 Mental Representations of Self John F. Kihlstrom and Nancy Cantor Theory of the Self: Impasse and Evolution Kenneth J. Gergen A Perceptual-Motor Theory of Emotion Howard Leventhal Equity and Social Change in Human Relationships Charles G. McClintock, Roderick M. Kramer, and Linda J. Keil A New Look at Dissonance Theory Joel Cooper and Russell H. Fazio Cognitive Theories of Persuasion Alice H. Eagly and Shelly Chaiken Helping Behavior and Altruism: An Empirical and Conceptual Overview John F. Dovidio Index Volume 18 A Typological Approach to Marital Interaction: Recent Theory and Research Mary Anne Fitzpatrick Groups in Exotic Environments Albert A. Harrison and Mary M. Connors Balance Theory, the Jordan Paradigm, and the Wiest Tetrahedon Chester A. Insko The Social Relations Model David A. Kenny and Lawrence La Voie Coalition Bargaining S. S. Komorita
423
When Belief Creates Reality Mark Snyder Index Volume 19 Distraction–Conflict Theory: Progress and Problems Robert S. Baron Recent Research on Selective Exposure to Information Dieter Frey The Role of Threat to Self-Esteem and Perceived Control in Recipient Reaction to Help: Theory Development and Empirical Validation Arie Nadler and Jeffrey D. Fisher The Elaboration Likelihood Model of Persuasion Richard E. Petty and John T. Cacioppo Natural Experiments on the Effects of Mass Media Violence on Fatal Aggression: Strengths and Weaknesses of a New Approach David P. Phillips Paradigms and Groups Ivan D. Steiner Social Categorization: Implications for Creation and Reduction of Intergroup Bias David A. Wilder Index Volume 20 Attitudes, Traits, and Actions: Dispositional Prediction of Behavior in Personality and Social Psychology Icek Ajzen Prosocial Motivation: Is It Ever Truly Altruistic? C. Daniel Batson Dimensions of Group Process: Amount and Structure of Vocal Interaction James M. Dabbs, Jr. and R. Barry Ruback The Dynamics of Opinion Formation Harold B. Gerard and Ruben Orive Positive Affect, Cognitive Processes, and Social Behavior Alice M. Isen
424
CONTENTS OF OTHER VOLUMES
Between Hope and Fear: The Psychology of Risk Lola L. Lopes Toward an Integration of Cognitive and Motivational Perspectives on Social Inference: A Biased Hypothesis-Testing Model Tom Pyszczynski and Jeff Greenberg Index Volume 21 Introduction Leonard Berkowitz Part I. The Self as Known Narrative and the Self as Relationship Kenneth J. Gergen and Mary M. Gergen Self and Others: Studies in Social Personality and Autobiography Seymour Rosenberg Content and Process in the Experience of Self William J. McGuire and Claire V. McGuire Information Processing and the Study of the Self John F. Kihlstrom, Nancy Cantor, Jeanne Sumi Albright, Beverly R. Chew, Stanley B. Klein, and Paula M. Niedenthal Part II. Self-Motives Toward a Self-Evaluation Maintenance Model of Social Behavior Abraham Tesser The Self: A Dialectical Approach Carl W. Backman The Psychology of Self-Affirmation: Sustaining the Integrity of the Self Claude M. Steele A Model of Behavioral Self-Regulation: Translating Intention into Action Michael F. Scheier and Charles S. Carver Index Volume 22 On the Construction of the Anger Experience: Aversive Events and Negative Priming in the Formation of Feelings Leonard Berkowitz and Karen Heimer Social Psychophysiology: A New Look John T. Cacioppo, Richard E. Petty, and Louis G. Tassinary
Self-Discrepancy Theory: What Patterns of Self-Beliefs Cause People to Suffer? E. Tory Higgins Minding Matters: The Consequences of Mindlessness-Mindfulness Ellen J. Langer The Tradeoffs of Social Control and Innovation in Groups and Organizations Charlan Jeanne Nemeth and Barry M. Staw Confession, Inhibition, and Disease James W. Pennebaker A Sociocognitive Model of Attitude Structure and Function Anthony R. Pratkanis and Anthony G. Greenwald Introspection, Attitude Change, and Attitude– Behavior Consistency: The Disruptive Effects of Explaining Why We Feel the Way We Do Timothy D. Wilson, Dana S. Dunn, Dolores Kraft, and Douglas J. Lisle Index Volume 23 A Continuum of Impression Formation, from Category-Based to Individuating Processes: Influences of Information and Motivation on Attention and Interpretation Susan T. Fiske and Steven L. Neuberg Multiple Processes by Which Attitudes Guide Behavior: The MODE Model as an Integrative Framework Russell H. Fazio PEAT: An Integrative Model of Attribution Processes John W. Medcof Reading People’s Minds: A Transformation Rule Model for Predicting Others’ Thoughts and Feelings Rachel Karniol Self-Attention and Behavior: A Review and Theoretical Update Frederick X. Gibbons Counterfactual Thinking and Social Perception: Thinking about What Might Have Been Dale T. Miller, William Turnbull, and Cathy McFarland Index
CONTENTS OF OTHER VOLUMES Volume 24 The Role of Self-Interest in Social and Political Attitudes David O. Sears and Carolyn L. Funk A Terror Management Theory of Social Behavior: The Psychological Functions of Self-Esteem and Cultural Worldviews Sheldon Solomon, Jeff Greenberg, and Tom Pyszczynski Mood and Persuasion: Affective States Influence the Processing of Persuasive Communications Norbert Schwarz, Herbert Bless, and Gerd Bohner A Focus Theory of Normative Conduct: A Theoretical Refinement and Reevaluation of the Role of Norms in Human Behavior Robert B. Cialdini, Carl A. Kallgren, and Raymond R. Reno The Effects of Interaction Goals on Person Perception James L. Hilton and John M. Darley Studying Social Interaction with the Rochester Interaction Record Harry T. Reis and Ladd Wheeler Subjective Construal, Social Inference, and Human Misunderstanding Dale W. Griffin and Lee Ross Index Volume 25 Universals in the Content and Structure of Values: Theoretical Advances and Empirical Tests in 20 Countries Shalom H. Schwartz Motivational Foundations of Behavioral Confirmation Mark Snyder A Relational Model of Authority in Groups Tom R. Tyler and E. Allan Lind You Can’t Always Think What You Want: Problems in the Suppression of Unwanted Thoughts Daniel M. Wegner Affect in Social Judgments and Decisions: A Multiprocess Model Joseph Paul Forgas
425
The Social Psychology of Stanley Milgram Thomas Blass The Impact of Accountability on Judgment and Choice: Toward a Social Contingency Model Philip E. Tetlock Index Volume 26 Attitudes Toward High Achievers and Reactions to Their Fall: Theory and Research Concerning Tall Poppies N. T. Feather Evolutionary Social Psychology: From Sexual Selection to Social Cognition Douglas T. Kenrick Judgment in a Social Context: Biases, Short-comings, and the Logic of Conversation Norbert Schwarz A Phase Model of Transitions: Cognitive and Motivational Consequences Diane N. Ruble Multiple-Audience Problems, Tactical Communication, and Social Interaction: A Relational-Regulation Perspective John H. Fleming From Social Inequality to Personal Entitlement: The Role of Social Comparisons, Legitimacy Appraisals, and Group Membership Brenda Major Mental Representations of Social Groups: Advances in Understanding Stereotypes and Stereotyping Charles Stangor and James E. Lange Index Volume 27 Inferences of Responsibility and Social Motivation Bernard Weiner Information Processing in Social Contexts: Implications for Social Memory and Judgment Robert S. Wyer, Jr. and Deborah H. Gruenfeld
426
CONTENTS OF OTHER VOLUMES
The Interactive Roles of Stability and Level of Self-Esteem: Research and Theory Michael H. Kernis and Stephanie B. Waschull Gender Differences in Perceiving Internal State: Toward a His-and-Hers Model of Perceptual Cue Use Tomi-Ann Roberts and James W. Pennebaker On the Role of Encoding Processes in Stereotype Maintenance William von Hippel, Denise Sekaquaptewa, and Patrick Vargas Psychological Barriers to Dispute Resolution Lee Ross and Andrew Ward Index Volume 28 The Biopsychosocial Model of Arousal Regulation Jim Blascovich and Joe Tomaka Outcome Biases in Social Perception: Implications for Dispositional Inference, Attitude Change, Stereotyping, and Social Behavior Scott T. Allison, Diane M. Mackie, and David M. Messick Principles of Judging Valence: What Makes Events Positive or Negative? C. Miguel Brendl and E. Tory Higgins Pluralistic Ignorance and the Perpetuation of Social Norms by Unwitting Actors Deborah A. Prentice and Dale T. Miller People as Flexible Interpreters: Evidence and Issues from Spontaneous Trait Inference James S. Uleman, Leonard S. Newman, and Gordon B. Moskowitz Social Perception, Social Stereotypes, and Teacher Expectations: Accuracy and the Quest for the Powerful Self-Fulfilling Prophecy Lee Jussim Jacquelynne Eccles, and Stephanie Madon Nonverbal Behavior and Nonverbal Communication: What do Conversational Hand Gestures Tell Us? Robert M. Krauss, Yihsiu Chen, and Purnima Chawla Index
Volume 29 Counterfactual Thinking: The Intersection of Affect and Function Neal J. Roese and James M. Olson Terror Management Theory of Self-Esteem and Cultural Worldviews: Empirical Assessments and Conceptual Refinements Jeff Greenberg, Sheldon Solomon, and Tom Pyszczynski The Flexible Correction Model: The Role of Naı¨ve Theories of Bias in Bias Correction Duane T. Wegener and Richard E. Petty Self-Evaluation: To Thine Own Self Be Good, to Thine Own Self Be Sure, to Thine Own Self Be True, and to Thine Own Self Be Better Constantine Sedikides and Michael J. Strube Toward a Hierarchical Model of Intrinsic and Extrinsic Motivation Robert J. Vallerand Volume 30 Promotion and Prevention: Regulatory Focus as a Motivational Principle E. Tory Higgins The Other ‘‘Authoritarian Personality’’ Bob Altemeyer Person Preception Comes of Age: The Salience and Significance of Age in Social Adjustments Joann M. Montepare and Leslie A. Zebrowitz On the Perception of Social Consensus Joachim Krueger Prejudice and Stereotyping in Everyday Communication Janet B. Ruscher Situated Optimism: Specific Outcome Expectancies and Self-Regulation David A. Armor and Shelley E. Taylor Volume 31 Affect and Information Processing Robert S. Wyer, Jr., Gerald L. Clore, and Linda M. Isbell Linguistic Intergroup Bias: Stereotype Perpetuation through Language Anne Maass
CONTENTS OF OTHER VOLUMES Relationships from the Past in the Present: Significant-Other Representations and Transference in Interpersonal Life Serena Chen and Susan M. Anderson The Puzzle of Continuing Group Inequality: Piecing Together Psychological, Social, and Cultural Forces in Social Dominance Theory Felicia Prano Attitude Representation Theory Charles G. Lord and Mark R. Lepper Discontinuity Theory: Cognitive and Social Searches for Rationality and Normality— May Lead to Madness Philip G. Zimbardo
427
Ambivalent Sexism Peter Glick and Susan T. Fiske Videotaped Confessions: Is Guilt in the Eye of the Camera? G. Daniel Lassiter, Andrew L. Geers, Patrick J. Munhall, Ian M. Handley, and Melissa J. Beers Effort Determination of Cardiovascular Response: An Integrative Analysis with Applications in Social Psychology Rex A. Wright and Leslie D. Kirby Index Volume 34
Volume 32 The Nature and Function of Self-Esteem: Sociometer Theory Mark R. Leary and Roy F. Baumeister Temperature and Aggression Craig A. Anderson, Kathryn B. Anderson, Nancy Dorr, Kristina M. DeNeve, and Mindy Flanagan The Importance of Being Selective: Weighing the Role of Attribute Importance in Attitudinal Judgment Joop van der Pligt, Nanne K. de Vries, Antony S. R. Manstead, and Frank van Harreveld Toward a History of Social Behavior: Judgmental Accuracy from Thin Slices of the Behavioral Stream Nalini Amabady, Frank J. Bernieri, and Jennifer A. Richeson Attractiveness, Attraction, and Sexual Selection: Evolutionary Perspectives on the Form and Function of Physical Attractiveness Dianne S. Berry Index Volume 33 The Perception–Behavior Expressway: Automatic Effects of Social Perception on Social Behavior Ap Dijksterhuis and John A. Bargh A Dual-Process Cognitive-Motivational Theory of Ideology and Prejudice John Duckitt
Uncertainty Management by Means of Fairness Judgments Kees van den Bos and E. Allan Lind Cognition in Persuasion: An Analysis of Information Processing in Response to Persuasive Communications Dolores Albarracin Narrative-Based Representations of Social Knowledge: Their Construction and Use in Comprehension, Memory, and Judgment Robert S. Wyer, Jr., Rashmi Adaval and Stanley J. Colcombe Reflexion and Reflection: A Social Cognitive Neuroscience Approach to Attributional Inference Matthew D. Lieberman, Ruth Gaunt, Daniel T. Gilbert, and Yaacov Trope Antecedents and Consequences of Attributions to Discrimination: Theoretical and Empirical Advances Brenda Major, Wendy J. Quinton, and Shannon K. McCoy A Theory of Goal Systems Arie W. Kruglanski, James Y. Shah, Ayelet Fishbach, Ron Friedman, Woo Young Chun, and David Sleeth-Keppler Contending with Group Image: The Psychology of Stereotype and Social Identity Threat Claude M. Steele, Steven J. Spencer, and Joshua Aronson Index
428
CONTENTS OF OTHER VOLUMES
Volume 35 Social Identity and Leadership Processes in Groups Michael A. Hogg and Daan van Knippenberg The Attachment Behavioral System in Adulthood: Activation, Psychodynamics, and Interpersonal Processes Mario Mikulincer and Phillip R. Shaver Stereotypes and Behavioral Confirmation: From Interpersonal to Intergroup Perspectives Olivier Klein and Mark Snyder Motivational Bases of Information Processing and Strategy in Conflict and Negotiation Carsten K. W. De Dreu and Peter J. Carnevale Regulatory Mode: Locomotion and Assessment as Distinct Orientations E. Tory Higgins, Arie W. Kruglanski, and Antonio Pierro Affective Forecasting Timothy D. Wilson and Daniel T. Gilbert Index Volume 36 Aversive Racism John F. Dovidio and Samuel L. Gaertner Socially Situated Cognition: Cognition in its Social Context Eliot R. Smith and Gu¨n R. Semin Social Axioms: A Model for Social Beliefs in Multicultural Perspective Kwok Leung and Michael Harris Bond Violent Video Games: Specific Effects of Violent Content on Aggressive Thoughts and Behavior Craig A. Anderson, Nicholas L. Carnagey, Mindy Flanagan, Arlin J. Benjamin, Jr., Janie Eubanks, and Jeffery C. Valentine Survival and Change in Judgments: A Model of Activation and Comparison Dolores Albarracı´n, Harry M. Wallace, and Laura R. Glasman
The Implicit Volition Model: On the Preconscious Regulation of Temporarily Adopted Goals Gordon B. Moskowitz, Peizhong Li, and Elizabeth R. Kirk Index Volume 37 Accuracy in Social Perception: Criticisms, Controversies, Criteria, Components, and Cognitive Processes Lee Jussim Over Thirty Years Later: A Contemporary Look at Symbolic Racism David O. Sears and P. J. Henry Managing Group Behavior: The Interplay Between Procedural Justice, Sense of Self, and Cooperation David De Cremer and Tom R. Tyler So Right it’s Wrong: Groupthink and the Ubiquitous Nature of Polarized Group Decision Making Robert S. Baron An Integrative Theory of Intergroup Contact Rupert Brown and Miles Hewstone Says Who?: Epistemic Authority Effects in Social Judgment Arie W. Kruglanski, Amiram Raviv, Daniel Bar-Tal, Alona Raviv, Keren Sharvit, Shmuel Ellis, Ruth Bar, Antonio Pierro, and Lucia Mannetti Index Volume 38 Exploring the Latent Structure of Strength-Related Attitude Attributes Penny S. Visser, George Y. Bizer, and Jon A. Krosnick Implementation Intentions and Goal Achievement: A Meta-Analysis of Effects and Processes Peter M. Gollwitzer and Paschal Sheeran Interracial Interactions: A Relational Approach J. Nicole Shelton and Jennifer A. Richeson
CONTENTS OF OTHER VOLUMES The Psychology of Self-Defense: Self-Affirmation Theory David K. Sherman and Geoffrey L. Cohen Intergroup Beliefs: Investigations from the Social Side Charles Stangor and Scott P. Leary
429
A Multicomponent Conceptualization of Authenticity: Theory and Research Michael H. Kernis and Brian M. Goldman Index
INDEX
A
ideal type, 47–48 ideology and, 45 insider perspective, lessons from, 47–51 interventionist and do unto others morality, 49–51 outsider perspective, lessons from, 44–47 private self‐consciousness, 34–35 reaction times for stories, 22–24 reciprocal projection, 37 self constructed and experienced for, 5, 9 spontaneous construction of sentences, 25 third party projection, 37 third‐person imagery use, during self‐control task by, 19–20 Asian holistic philosophy, 52 Associative learning, 258 Assuredness, 4 Attentional capacity, in human judgment, 270 Attitude‐consistent behavior, uncertainty and normative support eVect on, 105 Attitudes, 2–4, 42–43 uncertainty about, 73–74 Attributional complexity, 78 Authoritarian personality, 77 Automatically triggered motivation, automotive model of, 315 Awareness, in judgment formation, 262–263
Absolute certainty, 73 Accessibility experiences, 129–131 bias, emergence and attenuation of, 132–143 declarative and experiential information interplay, 135–140 discrediting informational value of experience, 140–143 Accessible declarative information, 133–134 Adjustment process, 10 Alliances, 48–49 Alternatives attenuate bias, 133 Anchoring process, 10 Anger, 14–15, 42, 47 Animal‐learning, 258 Arm‐in‐arm alliance, 48 Asian‐Americans accuracy under cognitive load and no load conditions, 29–31 characterization of world, 38–40, 42 collectivistic ideologies, 5 culture, 12–13 egocentric and relational projection scores among, 15 embodied cognition, 26–27 empathy‐as‐projection, 33–36, 38, 42 experiencing self and the world, 39, 42–43 illusion of transparency for, 32–36, 42 insider perspective, 10, 28 memory imagery, 16–18 online imagery, 19–20 outsider perspective, 9–10, 27 insider bias, 28 phenomenological experiences, 4, 41–43, 58–59 alliances and side‐by‐side relationships, 48–49
B Behavior engulfs field, concept of, 52 Behaviors, 2, 77 Beliefs, 2–4, 42–43 uncertainty about, 74 Bias emergence and attenuation, accessibility experiences and, 132–143 Bidirectional causation, between behavioral mimicry and holistic thinking style, 52–55 405
406
INDEX
Blind patriotism, 92 Blocking experiments, 259 C Categorization aVects intergroup diVerentiation, 176–177 along multiple dimensions simultaneously, 172–176 category selection, 172–174 simultaneous categorization, 174–176 category diVerentiation models, 178–179 common in‐group identity model, 179–181, 183, 211 complexity in, 215 convergent categorization, 169 crossed categorization model, 169, 170, 181–187 definition, 170 diVerentiation–discrimination models, 177–218 of multiple categorization eVects, 212–15, 235–238 diVerentiation process, 176 –177 distinctiveness threat, 204 –205 divergent categorization, 169 dual identification, 211–215 evaluation, patterns of, 187–190 hierarchical categorization, 170 identification and multiple categorization, 207–211 in‐group distinctiveness, 205 identification, 205–206 motivational processes, 203–204 motivation and multiple categorization, 206–207 multiple categorization, 166–167, 169–170 perceived group variability, 215–216 principles and assumptions, 170–177 process, 176–177 self‐esteem in, 203–204 social categorization, 165, 169 social identity complexity in, 217–218 subgrouping variability, 215–216 subordinate categorization, 170 superordinate categorization, 170 two moderators, one route, 199–203 Category combinations, in decategorization, 228–230
Category confusion paradigm, 173–174 Category conjunctions, in decategorization, 228 Category diVerentiation concept of, 168 evaluation‐relationship, 168 mechanism, 168 models, 178–179 minimal group paradigm, 178 Category selection, 172–174 category confusion paradigm, 173 contextual reason, 172–174 motivational reasons, 173–174 stereotype threat model, 172 Central member, in uncertainty‐identity theory, 99–102 Certainty and meaning, 76 Certainty‐oriented people, 78 Classical conditioning. See Pavlovian conditioning Classical projection. See Egocentric projection Closed‐minded personality, 77 Cognitive complexity, 77–78 Cognitive‐developmental theory, 231 Cognitive dissonance concept of, 359 process within groups, 364–365 theory, 362–364, 395–397 Cognitive resources attentional capacity, 270 in human judgment, 269–270, 274 rule accessibility, 269–270 Collective memories, 17–18 Collective self, 365 Common in‐group identity model, 169, 170, 179–181, 211 crossed categorization model and, 183–185 shifting identity model and, 183 Communication perspective, on uncertainty, 78 Compensatory conviction, uncertainty‐identity theory implications and, 108–111, 113 Complementary gender stereotypes, 313 exposure eVects to on diVuse system justification, 318–321 on gender‐specific system justification, 315–318 gender inequality, altering presumptive context of, 321–323
407
INDEX Complementary stereotypes exposure eVects to on diVuse system justification, 318–321 on gender‐specific system justification, 315–318 function of, system‐justifying, 312–335 gender inequality, altering presumptive context of, 321–323 moderators’ eVect, on system justification, 335–346 causal link perception, between trait and outcome, 336–341 political orientation, 341–344 Protestant Work Ethic (PWE), 345–346 poor but happy and poor but honest stereotype eVects of exposure on diVuse system justification, 329–332 on implicit activation of justice motive, 332–335 potential of, system‐justifying, 335–346 perception of causal link between trait and outcome, 336–341 racial attitudes, role in, 318 rationalize status diVerences, 324–335 ethnic status stereotypes, regional and, 324–327 virtues of oppressed, 327–328 Complexity, in categorization, 215 Conceptual fluency, 131 Conditioned inhibition eVects, 259 Conditioned stimulus (CS), learning by repeated pairing of, 258–259 Conditioning phenomena in judgment formation, 258–261 Pavlovian versus evaluative conditioning, 260–261 Confidence changes, in declarative and experiential information interplay, 135–137 Conscious awareness, 261 Consensus and truth, fluency implications for public information campaigns, 145–146 Consider‐the‐opposite strategy, 128, 140 Contempt, 14 Content‐based models, 130, 133–135, 138, 140 Contiguity, learning by, 258 Continuum model of impression formation, 219 Converging categorizations, 169, 186–187 Cross‐cultural psychology, 2
Crossed categorization model, 169, 170, 181–187 categories formed from crossing gender and age dimensions, 182 common in‐group identity model and, 183–185 Cue‐configuration, concept of, 263 Culture beliefs, 3 characterizing world, 11, 38–40, 42 confusion in what is in one’s own head empathy as projection, 31–38 illusion of one’s own transparency, 31–36 and what is out there, 28–31 content of, 2 dynamic constructivism, 44 empathy as projection, 31–36 face of, 45, 51, 57 ideology and psychology, 2 internal experience, 4 invisibility and strength of, 55–57 lessons of experience from insider perspective, 47–51 outsider perspective, 44–47 loose, 5–7, 12 memory imagery and, 16–18 mental models of self and others, 21–28 mutual constitution, 44 online imagery and, 18–20 of personal experience, 1–59 process of, 2 projection as empathy in group setting, 36–38 and psyches, 2, 44 relational versus egocentric projection and, 13–15 role, in structuring people’s personal phenomenological experience, 1 sociofunctional account perception and, 3–13 soft embodiment of, 55–57 tight, 5–7, 12 D Debiasing theories, 129 Decategorization, 218–233 category combinations, 228–230 category function, 218–219 impression formation, dual process models of, 219–220
408
INDEX
Decategorization (cont.) multiple categorization and, 220–227 process, 169 skills development, 230–233 Declarative and experiential information interplay accessibility experiences, 135–140 confidence changes, 135–137 hindsight bias, 136–139 planning fallacy, 139–140 success ratings, by thought‐listing and time, 136 Default phenomenology, 58–59 Deictic words, 21 Democracy in America, 48 Depersonalization, 365–366, 395 process, 79, 112 uncertainty‐identity theory, 98–99 Deviance social identity analyses of, 99–100 in uncertainty‐identity theory, 99–102 DiVerentiation–discrimination models of categorization, 177–218 aVect and cognition, 190–198 aVective state, 190–195 category primes, 195–198 category diVerentiation models, 178–179 common in‐group identity models, 179–181, 183, 211 complexity, 215 crossed categorization model, 181–187 distinctiveness threat, 204–205 dual identification, 211–215 dual‐route model, 199 evaluation, patterns of, 187–190 additive pattern, 188–190 dominance pattern, 188–189 equivalence pattern, 188, 190 hierarchical acceptance pattern, 188–189 hierarchical rejection pattern, 188–190 social exclusion pattern, 188–189 social inclusion pattern, 188–190 identification and multiple categorization, 207–211 in‐group distinctiveness, 205 identification, 205–206 motivational processes, 203–204 motivation and multiple categorization, 206–207
multiple categorization eVects, 212–15, 235–238 perceived group variability, 215–216 self‐esteem in, 203–204 social identity complexity in, 217–218 subgrouping variability, 215–216 two moderators, one route, 199–203 DiVuse system justification exposure eVects to complementary stereotypes on, 318–321 poor but happy and poor but honest stereotype on, 329–332 poor but honest and rich but dishonest stereotype on, 330–332 victim‐derogation and victim‐ enhancement on, 337–340 item for measuring, 318–319 Directional motivation, in human judgment, 271–272 Dissonance eVects, 9 Distinctiveness threat, in categorization, 204–205 Diverging categorizations, 169, 185 Dogmatic personality, 77 Do unto others morality, 49–51 Dual identification, in categorization, 211–215 Dual process models of impression formation, in decategorization, 219–220 Dual process theory, 219 E Egalitarian veneer, 350 Egocentric biases, 1, 10, 28 Egocentric projection phenomenon, 14–15 Egocentrism, 6 Emotions, 14–15 facial feedback hypothesis of, 46–47 influence on perceptions, 14 phenomenological experience of, 46–47 Empathy‐as‐projection, 33–38, 42, 48–50 Entitativity role in uncertainty induced identification, 89–91 role of, 72 uncertainty‐identity theory and, 88–91 Environmental contingencies, 259 Euro‐Americans accuracy under cognitive load and no load conditions, 29–31
409
INDEX characterization of world, 38–40, 42 collectivistic ideologies, 5 culture, 12–13 default phenomenology, 58–59 egocentric and relational projection scores among, 15 embodied cognition, 26–27 empathy‐as‐projection, 33–38, 42, 48 experiencing self and the world, 39, 42–43 illusion of transparency for, 32–36, 42 insider perspective, 10 memory imagery, 16–18 online imagery, 19–20 outsider perspective, 9–10, 27 insider bias, 28 phenomenological experiences, 4, 41–43, 58–59 alliances and side‐by‐side relationships, 48–49 ideal type, 47–48 ideology and, 45–51 insider perspective, lessons from, 47–51 interventionist and do unto others morality, 49–51 outsider perspective, lessons from, 44–47 private self‐consciousness, 34–35 reaction times for stories, 22–24 reciprocal projection, 37 self, 5 spontaneous construction of sentences, 25 third party projection, 37 third‐person imagery use, during self‐control task by, 19–20 Evaluative conditioning, in judgment formation, 260–261 Existential anxiety, 108–109 Experiential and declarative information interplay accessibility experiences, 135–140 confidence changes, 135–137 hindsight bias, 136–139 planning fallacy, 139–140 success ratings by thought‐listing and time, 136 Extreme‐totalistic groups, in uncertainty‐ identity theory, 91–97, 113 Extremism, studies of, 94–97, 113 Extremist groups, hierarchical leadership, 101–102
F Face culture of, 45, 51, 57 East Asian conception of, 6, 57 metaphor of, 45–46 Face‐to‐face relationships, 48–49 Familiarity and bias, fluency implications for public information campaigns, 143–145 implications for public information campaigns, 143–153 Fanaticism, 92 Fear, 14, 42, 47 Feelings, 6 as information model, 261 uncertainty about, 69, 73–74, 77, 79, 111 Fluency‐familiarity‐truth link implications for public information campaigns, 146–152 lending credibility to myths, 151–152 spreading myths by debunking them, 146–149 turning warnings into recommendations, 149–151 Fluency implications, for public information campaigns, 143–153 and detection of distorted questions, 144–145 fluency‐consensus and truth, 145–146 fluency‐familiarity and bias, 143–145 and hindsight bias, 144 lending credibility to myths, 151–152 spreading myths by debunking them, 146–149 turning warnings into recommendations, 149–151 Form as content, 3–4 of third‐person experience, 3 Functionalist thought experiment, 9 G Gender inequality, altering presumptive context of, 321–323 Gender‐specific system justification exposure eVects to complementary stereotypes on, 315–318 measuring scale for, 316–317 Generalized other, notion of, 8–9
410
INDEX
Genocide, 92 Group‐centrism syndrome, 106–107 Group extremism, 72 Group identification categorization and uncertainty eVect on, 82–84 entitativity and uncertainty eVect on, 90–91 as function of group extremism, 95–96 perceived entitativity, 89–90 group extremism and uncertainty eVect on, 97 measurement, 82–83 prototype of, 79–80, 112 by self‐categorization, 79–80 for social identity, 79–81 status and uncertainty eVect on, 86 uncertainty and, 79–80 Group prototypicality, vicarious dissonance and, 378–381 Guilt, 57
gleaning relevance of information given, 267–268 independence of informational contents, 273–274 informational relevance, 266–267, 274 motivation, 271–272, 294 multiple determination of values, 272 nondirectional motivation, 271–272, 274 orthogonality of, 273 potential and perceived relevance, 267 properties of, 272–274 rule accessibility, 269–270 task demands, 268–269, 274, 294 process‐awareness dimension, 256 strategies to debias, 127 (See also Judgment and decision making) Human motivation meaning‐making, 76 models of, 76 uncertainty as, 74–75 Hypocrisy, 390–395
H
I
Habits of heart perspective, 49 Health and consumer behavior, 127 Hindsight bias in declarative and experiential information interplay, 136–139 and fluency implications for public information campaigns, 144 Homeostasis macro‐level ideology in, 52–55 phenomenological experience and cultural ideas in, 53 Human behavior, motivation by uncertainty, 73–74 Human consciousness, 43 Human identity. See also Categorization solitarist and pluralist approaches to, 234 Human judgment. See also Judgment formation; Social judgment domain of, 256, 294 dual‐mode perspective, 291–292, 296 motivation, 271–272, 294 multidimensional parametric space, 274 mysteries of, 256 parameters, 265–274, 291–297 attentional capacity, 270 cognitive resources, 269–270, 274, 294 directional motivation, 271–272, 274
Identification, and multiple categorization, 207–211 Identity, studies of, 94–97 Ideological thinking, 92 Ideology, uncertainty‐identity theory and, 72, 103 Ideology and phenomenology lessons of experience, 44–51, 56, 59 alliances and side‐by‐side relationships, 48–49 ideal type, 47–48 insider form of experience, 47–51 interventionist and do unto others morality, 49–51 outsider form of experience, 44–47 Illusion‐of‐truth eVects pattern, 147 Imagined feeling of being, 2 Impartial spectator, notion of, 8–9 Implicit complementary versus noncomplementary stereotypical associations, 346–349 concerning rich/poor and happiness/ honesty, 329–332, 347–348 concerning white/black and happiness/ honesty, 348–349 Impression formation, dual process models of, 219–220
411
INDEX Independent cultures, 6, 12 Independent self, 2, 365. See also Self Individualistic culture, 4 Individual self, 364–365 Informational relevance gleaning to judgmental task, 267–268 in human judgment, 266–267, 274 potential and perceived relevance, 267 In‐group distinctiveness, in categorization, 205 identification, in categorization, 205–206 In‐group Projection Model, 217 Insider perspective, 6–7 lessons of experiences from, 47–51 alliances and side‐by‐side relationships, 48–49 ideal type, 47–48 interventionist and do unto others morality, 49–51 note on, 11 perceptions on world, 7, 10, 15 phenomenological gestalt of, 7, 11 self projected onto world, 7 Interdependent culture, 12 Interdependent self, 2, 365 Intergroup bias, 168–169, 176, 179, 184–187 diVerentiation and categorization, 176–177, 179, 184 –187 relations, 101, 165, 176 Internal experience, and perceptions, 4 Internal thoughts, 2, 6 Interpersonal harmony, 3 Interventionist morality, 49–51 Introspection, phenomenology and, 6, 55–57 J Japanese culture, emotions in, 46 Judgmental parameters. See also Human judgment independence of informational contents, 273–274 multiple determination of values, 272 orthogonality of, 273 properties of, 272–274, 291–297 Judgment and decision making content‐focused models of, 130, 133–135, 138, 140 metacognitive experiences role, 128
role of subjective accessibility experiences in, 143 theories of, 129 Judgment formation. See also Human judgment associative versus rule‐based, 263 awareness and, 262–263 conditioning phenomena, 258–261 implicational structure of reasoning, 258 multidimensional parametric space, 274 Pavlovian versus evaluative conditioning, 260–261 perception and, 261–262 phenomenon of pattern recognition, 263–265 psychophysics and, 261–262, 291 role of rules in, 257–265
K Knowledge, illusions of, 131
L Leaders, in uncertainty‐identity theory, 99–102 Leadership, social identity theory of, 99 Lending credibility to myths, fluency‐ familiarity‐truth link implications for, 151–152 Looking glass self, notion of, 8–9 Loose cultures, 5–7, 12 self‐esteem, 6
M Macro‐level ideology, in homeostasis, 52–55 Marginal member, in uncertainty‐identity theory, 99–102 Meaning certainty and, 76 human motive, 76 maintenance model, 76 pursuit of, 75–76 Memory study, 15–16 Mental functioning, naive theories of, 129–131, 134, 143, 154–155
412
INDEX
Mental models of self and others in narrative, 21–28 reaction times for stories, 22–24 research on, 24–27 embodiment of self and other’s perspective, 25–27 spontaneous construction of sentences, 24–25 Mental toughness, 18, 41 Metacognitive experiences accessibility experiences, 129–131 declarative and experiential information interplay, 135–140 discrediting informational value of experience, 140–143 emergence and attenuation of bias and, 132–143 content‐based models, 130, 133–135, 138, 140 declarative and experiential information interplay, 135–140 fluency‐familiarity‐truth link implications for lending credibility to myths, 151–152 public information campaigns, 146–152 spreading myths by debunking them, 146–149 turning warnings into recommendations, 149–151 in groups, 156 implications and future directions, 153–156 naive theories, 129–131, 134, 143, 154–155 temporal changes, 155–156 interplay of thought content and, 134–140 processing fluency, 131–132, 143–146, 155 role in judgment and decision making, 128 Micro‐level experience, in homeostasis, 52–55 Micro‐level phenomenology, 52 Minangkabau culture, emotions in, 47 Moderators of eVect of complementary stereotypes on system justification, 335–346 perception of causal link between trait and outcome, 336–341 political orientation, 341–344 Protestant Work Ethic (PWE), 345–346 Mortality salience, 109 Motivation in human judgment, 271–272 multiple categorization and, 206–207
Motivational processes, in categorization, 203–204 Multiple categorization, 166–167. See also Categorization decategorization and, 220–227 diVerentiation–discrimination models of, 212–15, 235–238 eVects diVerentiation–discrimination models of, 212–15, 235–238 identification and, 207–211 motivation and, 206–207 processes involved in, 167, 169 psychological and behavioral impact of, 169–170 in social contexts, 167 Multiple classification skills, 169 N Naive theories, of mental functioning, 129–131, 134, 143, 154–155 Natural systems, 74 Nondirectional motivation, in human judgment, 271–272, 274 O Objective self‐awareness phenomena, 8 Oneself, uncertainty about, 74, 80 Organizational socialization theories, uncertainty in, 73 Outsider phenomenology, 3 Outsider’s perspective, 9–10 lessons of experiences from, 44–47 note on, 11 phenomenological gestalt of, 7, 11 on self and self as seen by others, 7, 15 P Panglossian ideology, in system justification, 305–353 Participants, note on, 11–13 Pattern recognition phenomenon, in judgment formation, 263–265 Pavlovian conditioning, in judgment formation, 260–261 People’s personal phenomenological experience, culture role in structuring, 1 Perceived group variability, in categorization, 215–216
413
INDEX Perception, 2 internal experience and, 4 in judgment formation, 261–262 sociofunctional account, 3–13 characterizing world, 11 generalized other, 8–9 impartial spectator, 8–9 insider’s perspective, 10–11 looking glass self, 8–9 outsider’s perspective, 9–11 participants in, 11–13 phenomenological gestalts, 11 tight and loose cultures, 5–7 uncertainty about, 73–74, 77 Perceptual fluency, 131 Personal cognitive dissonance, 361–365 in east and west, 384–386 Personal experience, culture and structure of, 1–59 Personal identity, 365 Phenomenological experiences Asian‐Americans, 4, 41–43 Euro‐Americans, 4, 41–43 form of, 3–4 ideology and, 44–51, 53 lessons of experience from insider perspective, 47–51 outsider perspective, 44–47 structure of, 3 Phenomenology for Asian‐Americans, 2 of being a self, 2 of embodiment of narratives, 1 for Euro‐Americans, 2 gestalt of insider’s perspective, 7, 11 outsider’s perspective, 7, 11 ideology and, 44–51, 56, 59 insider form of experience alliances and side‐by‐side relationships, 48–49 ideal type, 47–48 interventionist and do unto others morality, 49–51 of memory imagery, 1 methodology and introspection, 55–57 of online imagery, 1 outsider form of experience, 2–3, 44–47 of relational versus egocentric projection, 1 self‐as‐object, 54
thoughts and feelings, 6 of visualization, 1 Planning fallacy, in declarative and experiential information interplay, 139–140 Poor but happy and poor but honest stereotype exposure eVects on diVuse system justification, 329–332 implicit activation of justice motive, 332–335 Priming cultural identities, 388–390 Processing fluency, 131–132, 143–146, 155 Protestant work ethic (PWE), 345–346 Psyches and cultures, 2, 44 Psychological process and social structures, 167–168 Psychophysics, in judgment formation, 261–262, 291 Public information campaigns designed to dispel erroneous beliefs, 127 familiarity implications for, 143–153 fluency‐familiarity‐truth link implications for, 146–152 lending credibility to myths, 151–152 spreading myths by debunking them, 146–149 turning warnings into recommendations, 149–151 fluency implications for, 143–153 detection of distorted questions, 144–145 fluency‐consensus and truth, 145–146 fluency‐familiarity and bias, 143–145 hindsight bias, 144 truth implications for, 143–153 R Rationalize status diVerences among complementary stereotypes, 324–335 regional and ethnic status stereotypes, 324–327 in England, 325–326 in Israel, 325–327 in Italy, 324–326 in United States, 324–325 virtues of oppressed, 327–328 Reasoning implicational structure, in judgment formation, 258
414
INDEX
Regional and ethnic status stereotypes, 324–327 Regulatory focus theory, 100 Relational projection phenomenon, 14–15 Relational self, 364–365 Relationships engulfs field, concept of, 52 Relative informational impact, in social judgment, 289–291 Religious fundamentalism, 92 Repeated pairing, learning by, 258 Rule accessibility, in human judgment, 269–270 Rules role, in judgment formation, 257–265 S Sadness, 14–15, 47 Self as an outsider, 6 by attitudes, 5 categorization, 79–80, 112 categorization theory, 71–72, 178, 217, 365, 395–396 conceptual uncertainty, 396 constructed and experienced for Asian‐ Americans, 9 empirical study of, 364–365 enhancement, 71–72, 112 esteem, 71, 85–87, 112 failure to suppress (See Transparency illusion) feeling of, 2 independent, 2 individual‐level phenomenology of, 52 by in‐group–out‐group distinctions, 5 insider phenomenology of, 2 insider’s perspective on projected onto world, 7 interdependent, 2–3 by interpersonal duties, 5 by intrapsychic traits, 5 mental models of, 21–28 as object phenomenology, 54 outsider phenomenology of, 2, 7 outsider’s perspective on, 7, 10 by personal expressivity, 5 on projected onto world, 7 relevance, 73 as seen by others, 7, 58 service of, 75
by social roles, 5 tight cultures, 5 uncertainty, 77, 80, 94 Self‐aYrmation theory, 85 Self‐assessments, 53–54 Self‐categorization‐generated depersonalization, 99 Self‐categorization theory, 71–72, 178, 217, 365, 395–396 Self‐esteem, 6, 71, 85–87, 112 in categorization, 203–204 hypothesis self‐uncertainty manipulations aVect on, 85 social identity research on, 85 uncertainty relation, 85–87, 112 Self‐evaluation maintenance theory, 173 Self‐objectification phenomena, 8, 52 Self‐verification theory, uncertainty‐identity theory and, 110, 113 Shame, 14–15, 41, 57–58 Shifting identity model, 183 Short‐term uncertainty, 74 Shoulder‐to‐shoulder alliance, 48 Side‐by‐side relationships, 48–49, 51 Sign‐learning theory, 259 Situational uncertainty, 83 Skills development, in decategorization, 230–233 Social categorization, 165. See also Categorization changing representation of, 167, 169 multiple bases for, 165–166, 169 processes involved in, 167 psychological process, 171 Social extremism, uncertainty‐identity theory and, 91–97, 113 Social identity, 365 analyses of deviance, 99–100 basic hypothesis tests, 81–84 complexity in categorization, 217–218 definition of, 70–71 group identification for, 79–81, 86 minimal group paradigm, 81, 85–86 processes, 71–72 role of self‐esteem in, 71, 85–87 self‐inclusive social category, 81 theory (See Social identity theory) in uncertainty‐identity theory, 79–87 group identity, 79–81, 86
INDEX hypothesis tests, 81–84 self‐enhancement and self‐esteem relation, 85–87 uncertainty relevance and group relevance, 87 vicarious dissonance and, 365–368 Social identity theory, 69–70, 204, 365–367 development of, 72, 81 of intergroup relations, 71 Social judgment. See also Human judgment; Judgment formation impact of information, perceived relevance and, 275–280 parametric model, 274–291 relative informational impact, 289–291 subjective relevance of information role, 275–283 framing eVects, 280–281 impression formation, trait centrality in, 281 lay theories, 281–283 teaching relevance, 280 task demands and processing resources role, 283–289 base‐rate neglect, 288–289 dispositional attributions, 287–288 persuasion research, 283–287 Social mobilization attitude‐behavior relation, 104–105 uncertainty‐identity theory and, 104–105 Social projection concept of, 98 of uncertainty‐identity theory, 98–99 Social structures and psychological process, 167–168 Social world balance and complementarity, measuring perceptions of, 319 people’s perception, 2 Societal uncertainty, 92 Sociofunctional account of perception and culture, 3–13 characterizing world, 11 generalized other, 8–9 impartial spectator, 8–9 insider’s perspective, 10–11 looking glass self, 8–9 outsider’s perspective, 9–11 participants in, 11–13 phenomenological gestalts, 11
415
tight and loose cultures, 5–7 Spreading myths by debunking, 146–149 Stereotype threat model, for category selection, 172 Subgrouping variability, in categorization, 215–216 Subjective accessibility experiences, 133, 137, 140 role in memory, 153 role in production and reduction of judgmental biases, 143 Subjective certainties, 75 Subjective consciousness, 43 Subjective relevance of information role in social judgment, 275–283 framing eVects, 280–281 impression formation, trait centrality in, 281 lay theories, 281–283 teaching relevance, 280 Subordinate categorization, 170 Success ratings by thought‐listing and time, 136 Superordinate categorization, 170 Sympathy, 14–15, 42, 51 System justification complementary stereotypes eVect on, 335–346 perception of causal link between trait and outcome, 336–341 political orientation, 341–344 Protestant Work Ethic (PWE), 345–346 Panglossian ideology in, 305–353 System justification theory (SJT), 305, 308–312 anticipatory rationalization of status quo, 309–311 justification of arbitrary inequalities among groups, 309 motivation to rationalize status quo, 308–309 out‐group favoritism, 309 and stereotyping as rationalization, 350–352 uncertainty‐identity theory and, 108, 110–111, 113 System‐justifying function of complementary gender stereotypes, 313–323 diVuse system justification, exposure eVects to, 318–321
416
INDEX
System‐justifying function (cont.) gender inequality, altering presumptive context of, 321–323 gender‐specific system justification, exposure eVects to, 315–318 of complementary stereotypes, 312–325 T Task demands in human judgment, 268–269, 274 and processing resources role, in social judgment, 283–289 base‐rate neglect, 288–289 dispositional attributions, 287–288 persuasion research, 283–287 in social judgment, 283–289 Task uncertainty, 83 Taste aversion, 259 Temporal changes, in metacognitive experiences, 155–156 Terrorism, 92 Terror management theory uncertainty‐identity theory and, 108–111, 113 uncertainty‐related motives in, 75 Thematic Apperception Test, 82 Third‐person experience, form of, 3 Thought experiment, 4, 9 Tight cultures, 5–7, 12 Totalistic groups, in uncertainty‐identity theory, 91–97, 113 Transparency illusion, 31–36 Trust, and uncertainty‐identity theory, 102 Truth, implications for public information campaigns, 143–153 Turning warnings into recommendations, fluency‐familiarity‐truth link implications for, 149–151 Two moderators, one route in categorization, 199–203 U Ultranationalism, 92 Uncertainty aVective dimensions, 74 in bad way, 73
caveats of, 73–74 communication perspective, 78 as context‐induced state, 77 feelings of, 69, 73–74, 77, 79, 111 in good way, 73 group identification relation, 72, 79–81, 86 as human motivation, 74–75 identification relation, 70, 72, 79–81, 84 identity and ideology, 103 social mobilization, 104–105 trust, 102 induced identification, 84, 107–108 depersonalization and projection processes in, 99 entitativity role in, 89–91, 108 judgment under, 270 meaningfulness and, 76 motivational role for, 72 in organizational socialization theories, 73 orientation toward, 77–78 politics of, 101 reduction, 72–74, 111–112 hypothesis, 72 pursuit of meaning and, 75–76 theory, 72 through group identification, 72, 79–81, 84, 112 relevance and group relevance, 87 resolution, to satisfy self‐assessment motive, 78 self‐enhancement relation, 85–87 self‐esteem relation, 85–87 self‐related, 77, 80 situational uncertainty, 83 studies of, 94–97 task uncertainty, 83 tolerance to, 77 types of, 77–79 Uncertainty‐identity theory applications of, 97–105 central member, 99–102 compensatory conviction and, 108–111, 113 context approach, 105–107 culture approach, 107–108 depersonalization and projection, 98–99, 112 deviants, 99–102 entitativity, 88–91 extensions of, 97–105
417
INDEX extreme‐totalistic groups, 91–97 features of, 111–113 historical perspective, 70–72 ideology and, 103 implications of, 97–105 leaders, 99–102 marginal member, 99–102 personality approach, 105–107 group‐centrism syndrome, 106–107, 113 need for cognitive closure, 106, 113 uncertainty orientation research, 107 self‐verification theory, 110, 113 social extremism and totalistic groups, 91–97, 113 social identity perspective, 79–87 group identity, 79–81, 86 hypothesis tests, 81–84 self‐enhancement and self‐esteem relation, 85–87, 112 uncertainty relevance and group relevance, 87 social mobilization and, 104–105 system justification and, 108, 110–111, 113 terror management and, 108–111, 113 trust and, 102 uncertainty, (See Uncertainty) Unconditioned stimulus (US), learning by repeated pairing, 258–259 Unconscious inferences, 262 V Values, 2–4, 42–43 uncertainty about, 73 Vicarious cognitive dissonance in interdependent cultures, 386–388 model of, 382 phenomenon, 359 Vicarious discomfort conception, 375–377
vicarious attitude change at service of, 377–378 Vicarious dissonance aVective experience of, 375–378 change at service of vicarious discomfort, 377–378 conception of vicarious discomfort, 375–377 collective and, 384–390 in east and west, 384–386 in interdependent cultures, 386–388 priming cultural identities, 388–390 evidence for, 368–381 group prototypicality and, 378–381 persuasion form, 371–372 process of, 381–384 social identity and, 365–368 translating into actions and attitudes to improve health, 390–395 vicarious attitude change, 372–374 Vicarious hypocrisy, 390, 392–395 Victim‐derogation and victim‐enhancement eVects of system threat on, 340–342 exposure eVects on diVuse system justification, 337–340
W Willful individualism, 3 Women, bivalent stereotypes of, 314 World characterization of, 11, 38–40, 42 perceptions of, 4
Z Zealotry, 72, 113