Multiple Social Categorization

Processes, Models and Applications

Richard J. Crisp, Miles Hewstone

• Publisher: Psychology Press
• Year: 2007
• Print ISBN: 9780415655675
• eBook ISBN: 9781135422950
• Edition: 1
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

'Ethnic cleansing', 'institutional racism', and 'social exclusion' are just some of the terms used to describe one of the most pressing social issues facing today's societies: prejudice and intergroup discrimination. Invariably, these pervasive social problems can be traced back to differences in religion, ethnicity, or countless other bases of group membership: the social categories to which people belong.

Social categorization, how we classify ourselves and others, exerts a profound influence on our thoughts, beliefs, feelings, and behaviors. In this volume, Richard Crisp and Miles Hewstone bring together a selection of leading figures in the social sciences to focus on a rapidly emerging, but critically important, new question: how, when, and why do people classify others along multiple dimensions of social categorization? The volume also explores what this means for social behavior, and what implications multiple and complex perceptions of category membership might have for reducing prejudice, discrimination, and social exclusion.

Topics covered include:

• 
  
  

• the cognitive, motivational, and affective implications of multiple categorization

• 
  
  

• the crossed categorization and common ingroup methods of reducing prejudice and intergroup discrimination

• 
  
  

• the nature of social categorization among multicultural, multiethnic, and multilingual individuals.

Multiple Social Categorization: Process, Models and Applications addresses issues that are central to social psychology and will be of particular interest to those studying or researching in the fields of Group Processes and Intergroup Relations.

Part I Introduction

1 Multiple social categorization

Context, process, and social consequences

Richard J. Crisp and Miles Hewstone

Think about how important your differences are to you. Think about how we all organise our lives in little boxes: man, woman, British, American, Muslim, Christian, Jew, Tory, Labour, New Labour... how could you navigate life if you didn’t know the difference between a child and an adult, an African and an Indian, a scientist and a lawyer?

Bill Clinton, The Dimbleby Lecture 2001, 14 December 2001

At any one time we have access to many identities, including race, sexuality, gender, nationality, class and religion. Far from being neutral, these identities are rooted in material conditions that confer power and privilege in relation to one another. These power relations, however, are not fixed. They are fluid in character, dynamic by nature and, therefore, complex in practice. The decisions as to which identities we assert, when we want to assert them and what we want to do with them are ours. But those decisions do not take place in a vacuum. They are shaped by circumstance and sharpened by crisis. We have a choice about which identities to give the floor to; but at specific moments they may also choose us.

Gary Younge, The Guardian, 21 January 2005

Categorization is integral and essential to human social interaction (Allport, 1954; Chaiken & Trope, 1999; Fiske & Neuberg, 1990). We classify and categorize without, for the most part, giving it a second thought—but it is crucial for our successful navigation of social life. As journalist Gary Younge notes above, our multiple identities are not simple, neutral, passive definitions of our existence; they are laden with implied status, power and value, and therefore of profound consequence for our lives. Categorization is dynamic and fluid—and we can both choose to be categorized, or categories can be chosen for us. Whether we are talking about being 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 various factors, such as context (e.g., Oakes, Turner, & Haslam, 1991) and motivation (e.g., Sinclair & Kunda, 1999) and increasingly it is apparent that in many contexts multiple bases for social categorization can be salient, combined and used simultaneously (e.g., see Brewer, Ho, Lee, & Miller, 1987; Stangor, Lynch, Duan, & Glass, 1992). This volume discusses how our classification of ourselves, and others, along these multiple criteria, can impact on psychological and social processes.

In the following chapters the authors discuss how social categorization— the classification of self and others into broad social groupings—has an impact on how we perceive, think, and behave. Social categories can be thought about with no reference to self-inclusion—Asian, British, young—or they can be classified by membership. That is, we can define social contexts by the groups to which we belong (“ingroups”) and the groups to which we do not belong (“outgroups”). In such intergroup contexts, ingroup and outgroup membership can have significant implications for social judgment and behavior (Hewstone, Rubin, & Willis, 2002) and the chapters in this volume discuss much about how categorization affects intergroup bias—the extent to which someone evaluates their own group more positively than other groups. The term “multiple categorization” refers to any intergroup context that involves perceiving more than a single basis for social classification. Examples abound from everyday life. In the 1983 British General Election a Conservative Party poster depicted a young black man, with the caption “Labour says he’s black. We say he’s British.” Writing about the debate within Britain on what it means to be “British”, Home Secretary Jack Straw returned to this poster and wrote, “Their suggestion that one had to make a choice—that one could not be both—was absurd...” (Straw, 2000). More recently, Abdolkarim Sorush, a leading Iranian intellectual, offered the view that every Iranian has three identities: Shia, Persian, and Western.¹

Research into multiple categorization is appealing because it qualifies and refines existing approaches to the study of social identification and social categorization. Typically, work in intergroup relations has focused on single ingroups and outgroups. In increasingly multicultural, multiethnic, and multireligious societies, there are now, however, multiple ways in which we can be different from others. This complexity of intergroup affiliation presents an intriguing challenge to theoretical accounts of group processes and intergroup relations. In this introductory chapter, we provide a broad overview of the current state of research across these varied perspectives.

FOUNDATIONS

That people can be affiliated to multiple social categories is not a new idea, nor exclusively an idea explored by social psychologists. Anthropologists, sociologists, and political scientists were the first to observe some unique consequences associated with the apparent use of multiple categorization in societal contexts. Evans-Pritchard (1940) and Murphy (1957) both document cases of reduced conflict in cultures with cross-cutting social structures (see also Coser, 1956; Deutsch, 1973; Gluckman, 1955; Lipset, 1960; Rae &Taylor, 1970). LeVine and Campbell (1972) suggested that group members actively make use of cross-cutting affiliations because they ensure security and stability (it is more difficult, for instance, to have conflictual relations with a group based on territory that is simultaneously an ally according to common ancestry). 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 different tribes. Such cross-cutting contexts provide divergent bases for group membership—tribe does not correspond to gente (group memberships neither converge nor correlate).

Research has confirmed that cross-cutting ties can be important on broad societal levels (Cairns & Mercer, 1984; Ross, 1985; see Phinney & Alipuria, this volume) and such category structures have emerged as an important concept for political scientists (see Horowitz, 1985; Wallace, 1973; see Carter, this volume). Based on these observations, models have been developed that focus on the perception and use of multiple categorization. Examples of work in this area focus on varied phenomena such as selection of multiple alternative categories in person perception (Macrae, Bodenhausen, & Milne, 1995; van Rijswijk & Ellemers, 2002), the emergence of novel stereotypes following category combinations (Smith & DeCoster, 1999), or the causal reasoning processes accompanying the emergence of those stereotypes (Kunda, Miller, and Claire, 1990). Other examples include the use of hierarchically ordered subtypes in memory (Stangor et al., 1992) or how multiple categories are represented and used in social judgment (Crisp, Ensari, Hewstone, & Miller, 2002; Klauer, Ehrenberg, & Wegener, 2003) as well as broader sociological, anthropological, and political science perspectives (e.g., Evans-Pritchard, 1940; Horowitz, 1985; LeVine & Campbell, 1972; Murphy, 1957; Wallace, 1973). The chapters in this volume address these very issues.

There are now an increasing number of psychological perspectives on multiple categorization which range from accounts of categorical structuring of knowledge about different social groups (see McGarty, this volume), and the representation of stereotypic attributes and how these attributes are combined (see Smith, this volume), through to the implications of multiple categorization for social judgment (see chapters in this volume by Crisp; Dovidio et al.; Hogg & Hornsey; Miller et al.; Singh; and Vescio et al.). We briefly review these different models of category representation and evaluation in terms of two specific criteria: (a) in terms of the conjunctions formed by considering multiple criteria (i.e., “combined categories”) and (b) in terms of the separable components of targets formed by considering multiple criteria (i.e., “constituent categories”).

A TAXONOMY OF MULTIPLE CATEGORIZATION

There are many ways in which multiple social categories can be perceived, represented, and used. One of the key distinctions in the literature is between models of multiple categorization at varying levels of inclusiveness (where inclusiveness is defined as the property of one category, e.g., European, subsuming another, e.g., Britain). For instance, a considerable body of work has examined the effects on evaluations of making a superordinate social classification salient instead of, or in addition to, a subordinate intergroup dichotomy (Gaertner & Dovidio, 2000; Gaertner, Dovidio, Anastasio, Bachman, &Rust, 1993; Gaertner, Mann, Dovidio, Murrell, & Pomare, 1990; Gaertner, Mann, Murrell, & Dovidio, 1989; see also Crisp; Hogg & Hornsey; and Dovidio et al., this volume). This experimental research has shown, beyond doubt, that the introduction of a superordinate categorization can reduce bias. Whether, or for how long, it does so, however, in extra-laboratory conflicts is another matter. National sporting success is sometimes seen as a means for superordinate national categorizations to dominate more conflictual subordinate ethnic groupings. Thus, when the Rwandan national football team beat favorite Uganda to qualify for the African Nations Cup in February 2003, the victory was seen as having significance way beyond the world of sport, because the Hutu and Tutsi communities (caught up in genocide in 1994) celebrated together. Yet the optimism surrounding South Africa’s Rugby World Cup success in 1995 and France’s Football World Cup victory in 1998 was only brief, despite initial celebrations. South Africa’s sole black player, winger Chester Williams, opined that the “one team, one country” image was “a lie”,² and one of France’s star black players, Thierry Henry, expressed the view, just four year later, that “when we won the World Cup for three months everyone was French. No one was saying someone was black, white, Algerian. It was nice but of course it did not last.”³

Studies of “crossed categorization” explore the effects on intergroup evaluations when two dichotomous dimensions of group membership are simultaneously salient (Brewer et al., 1987; Crisp & Hewstone, 2000a; 2001; Deschamps & Doise, 1978; Ensari & Miller, 1998; Hewstone, Islam, & Judd, 1993; for reviews see Crisp & Hewstone, 1999a; 2000b; Migdal, Hewstone, & Mullen, 1998; Mullen, Migdal, & Hewstone, 2001; Urban & Miller, 1998). Although similar conceptually to the common ingroup identity model, work in crossed categorization makes no specification of the inclusiveness of the cross-cutting categorizations, and examines the effects of both convergent and divergent categorization. Convergent categorization refers to targets that are designated outgroup on multiple dimensions simultaneously. For example, when gender and age dimensions are crossed, one of the composite groups created is convergent (young females are most different categorically from old males—from whom they differ according to two criteria). Divergent categorization refers to the comparison of two targets who share membership on one dimension, but who belong to different groups on another. At the time of Indian independence the philosopher Bertrand Russell said to Indian political leader Pandit Nehru that they had one thing in common, the fact that they were both atheists. “Yes,” replied Nehru, “but never forget, Russell, that you are a Christian atheist, and I am a Hindu one.” Like work on the common ingroup identity model, cross-cutting categorization has also been found to moderate evaluations (e.g., Crisp, Hewstone, & Rubin., 2001; Vanbeselaere, 1987; 1991; see Singh; Miller et al. and Vescio et al., this volume).

Other research has focused on the effects of “subgrouping” (considering multiple categories that are less inclusive than initial target groups; Brewer, Dull, & Lui, 1981; Maurer, Park, & Rothbart, 1995; Park, Ryan, & Judd, 1992; for a review see Richards & Hewstone, 2001). For example, within universities there are typically subgroups that define different areas of study, such as natural sciences, humanities, etc. Within these broad groupings there are further subgroups, so within natural sciences one can study physics, chemistry, etc. Interestingly, while the common ingroup identity and crossed categorization models have focused on the shared and overlapping nature of multiple memberships, subgrouping research has been more concerned with intragroup effects, and has not examined the intercategory implications of shared and overlapping subgroup categorization. Whereas the common ingroup identity and crossed categorization models have been almost exclusively concerned with intergroup evaluations, work into subgrouping has been almost exclusively concerned with stereotyping (typically specific nonevaluative beliefs about groups) and perceived group variability (typically, the extent to which all members of a group are perceived as similar or different from one another; Park & Judd, 1990; Park, Judd, & Ryan, 1991). Subgrouping not only has a moderating influence on stereotyping, but also seems the natural way people structure their intracategory representations (Park et al., 1992). Ingroup information can be organized into subgroups, which can be encoded, recalled, and applied easily and accurately, while there generally appears insufficient information to organize outgroup representations in the same way (see Richards & Hewstone, 2001). While subgrouping has beneficial effects on stereotyping and perceived group variability, its effects on broader positive or negative intergroup evaluations are relatively unexplored.

Although the differential inclusiveness of social categories has undoubtedly revealed distinct effects in the subgrouping compared to the common ingroup identity and crossed categorization literature, this appears to some extent to be due to the different intra- and intergroup foci of the different research traditions (stereotyping versus evaluations, respectively). We have recently examined the predictive power of inclusiveness in differentiating models of multiple categorization. We found common representational and evaluative consequences irrespective of whether categories combined were of different levels of inclusiveness (Crisp, Beck, & Hewstone, 2004). This suggests that, at least with respect to the phenomena considered in this volume, differential inclusiveness does not have a critical moderating role. More crucial is a focus on how categories are represented and used when perceived simultaneously, and it is this focus that we adopt.

REPRESENTING MULTIPLE CATEGORIES

Category conjunctions

When multiple dimensions of social categorization are simu...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Figures
• Tables
• Contributors
• Preface
• Part I Introduction
• Part II Multiple Category Representation
• Part III Multiple Categorization and Social Judgment
• Part IV Cross-Cutting Categorization and Evaluation
• Part V Societal and Political Perspectives
• Part VI Conclusion