This is an open access article distributed under the terms of the Creative Commons
Attribution License (Inbar and Lammers’s (2012) recent survey of the political values of social psychologists makes for revealing, if on reflection unsurprising, reading. It would seem that we are a discipline rooted firmly in the ideology of the left. Fewer than 4% of us, for example, categorize ourselves as ‘conservative’ in our political attitudes towards ‘social issues’. The vast majority of the remainder endorse liberal values, albeit with varying degrees of enthusiasm. Although our attitudes in other arenas (e.g. the economy, foreign policy) are more diverse, they remain predominantly left-leaning. Conservative values and behaviours are seemingly a minority pursuit in our field.
What is true of social psychologists in general, we would conjecture, is doubly true of those working in the field of intergroup relations, where the ‘social issues’ addressed include racism, xenophobia, sexism, homophobia, and anti-Semitism. Indeed, we have seldom met colleagues who espouse conservative beliefs about the nature, origins or solutions of such issues. Not everyone professes to be a card-carrying ‘liberal’ of course. Yet the overwhelming majority appear to view such issues as deriving, in the main, from wider structures of social inequality and institutionalised discrimination. They believe, too, that the point of psychological work is not simply to analyse the processes that sustain inequality and discrimination. We should also be prepared to get our proverbial hands dirty by challenging the status quo and working on behalf of the underdog.
Yet how might we accomplish this goal? What kinds of tools, concepts or insights does the discipline offer that might facilitate interventions to promote change, reducing the injustices that affect the lives of ordinary people in many societies? In addressing such questions, we might draw inspiration from a range of sources. We might capitalize on the contributions of work on political protest, for instance, or look to the insights of liberation, indigenous, feminist or Marxist psychology, to mention but a few examples. Over the course of the discipline’s history, however, one tradition of work above all others has informed how psychologists approach both the problem of intergroup discrimination and the challenge of eradicating it – namely work on intergroup prejudice. As Reynolds, Haslam, and Turner (2012) point out, research on prejudice has consumed literally millions of dollars of research funding and countless hours of research time, and it has dominated the careers of some of our most influential thinkers.
It is disquieting, then, that recent debates in the field have questioned the value of the concept of prejudice in general and the model of social change it underpins in particular (Dixon, Levine, Reicher, & Durrheim, 2012a, 2012b; Reicher, 2007; Reynolds et al., 2012; Wright & Lubensky, 2009). In this paper, we want to distil and evaluate some key elements of these emerging debates and explore their implications for how psychologists could, or should, understand the problem of change. We begin by outlining some underlying assumptions of the so-called ‘prejudice reduction model of social change’. We then describe how our own research on political attitudes in post-apartheid South Africa has led us to query, qualify and sometimes reject those very assumptions. Finally, looking to the future, we discuss the relationship between prejudice reduction and alternative models of social change. Our overall aim is not to advocate a wholesale abandonment of the project of getting us to have nicer thoughts and feelings about others. Who could possibly want that? Rather, we wish to encourage greater reflexivity about the potential limits of prejudice reduction as the standard model of change for social and political psychologists, particularly in societies characterised by historically entrenched inequalities.
The Concept of Prejudice and the Promise of Prejudice Reduction
The concept of prejudice originated in the enlightenment liberalism of the 18th century, an intellectual, cultural and political movement that instituted a sharp distinction between thought as reasoned reflection and thought as blind faith, dogma, and other varieties of irrationalism (Billig, 1988). In the early years of the 20th century, the concept’s referent narrowed. Prejudice came to signify the unreasoning and unreasonable dislike of members of other social groups, often purely on the basis that they belonged to such groups. This narrower definition of prejudice permeated the social sciences, becoming central to its modern usage. As Quillian (2006) notes at the outset of his recent review, “despite the changing nature of prejudice in modern society, most contemporary social science use of the term is highly consistent with Allport’s (1954) early definition of prejudice as ‘antipathy based on a faulty or inflexible generalization’” (p. 300).
It is difficult to overstate the historical significance of this conception of prejudice as ‘irrational antipathy’ within psychological work on intergroup relations. Although some commentators (e.g. Brown, 1995) have questioned the assumption that prejudice necessarily involves an inaccurate perception of others, the idea that it involves negative evaluations that somehow ‘bias’ how we treat others remains fundamental. For one thing, this idea has both directly and indirectly informed the development of many of the field’s foundational concepts (e.g. stereotyping, the Ultimate Attribution Error) and theories (e.g. authoritarianism, symbolic racism). For another, it has spawned a dazzling array of measurement techniques and research paradigms designed, for example, to probe the unconscious and implicit dimensions of our attitudes to others, to identify the biases that arise as unfortunate by-products of ‘information processing’ or, conversely, those that express the most intense intergroup emotions, such as hatred and disgust (see, e.g., Dovidio, 2001; Greenwald & Banaji, 2013; Harris & Fiske, 2006). Many of these innovations have reflected the desire to understand how individuals – sometimes in isolation, sometimes in dishearteningly large numbers – come to hold negative attitudes about others.
Yet why is this important? Why should our discipline have placed such emphasis on understanding and resolving this problem of negative evaluation? The answers to such questions seem self-evident to us now, but this was not always the case. There was a time when what we now call ‘prejudice’ was primarily seen as a natural, if not inevitable, reaction to contact between groups of differing cultural, political and, perhaps especially, intellectual status. If blame was to be attributed, then it was surely the fault of those belonging to lower status groups – the foreigners, immigrants, primitives and exotics with whom ‘we’ now had to deal? The history of the shift from this ideology of ‘difference’ to an ideology of ‘prejudice’ cannot detain us here (see, e.g., Samelson, 1978). However, two positive elements are worth noting. First, the concept of prejudice switched the direction of blame for intergroup ‘problems’, from the victims of discrimination (minorities, the disadvantaged) to its perpetrators (majorities, the advantaged). Second, it inaugurated a promising solution to such problems, based around the project of changing the hearts and minds of the dominant group in order to foster a state of reasonable acceptance of others.
Most research on prejudice has been governed by this thoroughly admirable objective. It has presupposed that prejudice reduction alters not only the private thoughts and feelings of individuals but also the wider patterns of intergroup conflict, discrimination and inequality that they sustain. Certainly, if prejudice reduction had little or no effect on such patterns or if, worse, it inhibited their transformation, then its significance would be immeasurably diminished.
In this respect, the specific pathways through which prejudice reduction is believed to produce social change are worth analysing a little further, even if they have come to appear a matter of common-sense. First, as mentioned already, the targets of change are generally believed to be members of historically advantaged groups – the bigots, rednecks, racists, homophobes and other bogeymen who have come to personify the problem of prejudice but also, more banally, those ordinary members whose reactions to others may be less extreme yet nevertheless help to shore up the status quo. Second, the psychological processes through which change occurs are generally believed to include the reduction of negative beliefs (stereotypes) about and negative feelings (attitudes) towards the disadvantaged. Third, and perhaps most important, the prejudice reduction model of social change holds that the effects of such social psychological shifts ramify beyond the individual. They gradually change the socio-political order of status and power relations and, a fiortori, broader structures of inequality and discrimination. How, when and why the latter process unfolds is, to put it kindly, under-specified in the prejudice literature. It is generally presumed or asserted rather than explained or demonstrated. Nevertheless, societal change lingers as a residual promise that underwrites the value of interventions to reduce prejudice, several of which have generated substantive traditions of research in their own right (see, e.g., Paluck & Green, 2009). There are signs, however, that this state of affairs is beginning to change.
Intergroup Contact and Attitudes Towards Political Change in Post-Apartheid South Africa
Hendrik Verwoerd, often styled as ‘the architect of apartheid’, was a psychologist by training. Before becoming leader of the Nationalist Party in 1950, he held a Chair in Applied Psychology at the University of Stellenbosch in the Western Cape region of South Africa, and prior to that he studied in some of the finest universities in Europe. Although his apologists have often distanced him from the ideologies of fascism and racism, it is a safe bet to assume that he would have fallen within Inbar and Lammers’ (2012) elusive 4% of psychologists who hold a conservative outlook on ‘social issues’. Verwoerd believed passionately in the need to maintain a political order based around racial hierarchy and division. Under the guise of producing good neighbourly relations and efficient labour practices, the government he led consolidated a system of ‘apartness’ that ultimately became an international symbol of racial injustice.
The first and second authors of this article began their research careers as Verwoerd’s system entered its twilight years, buckling under the pressure of mass resistance, armed struggle and international isolation. Our early work tried to map the complex forms of desegregation (and re-segregation) that were accompanying the collapse of apartheid policies such as the Group Areas Act, the Reservation of Separate Amenities Act, and the Prevention of Illegal Squatting Act. We investigated changing relations in contexts of residence (Dixon, Foster, Durrheim, & Wilbraham, 1994; Dixon, Reicher & Foster, 1997), education (Schrieff, Tredoux, Dixon, & Finchilescu, 2005; Tredoux, Dixon, Underwood, Nunez, & Finchilescu, 2005) and leisure (Dixon & Durrheim, 2003; Tredoux & Dixon, 2009). We probed how South Africans responded to such changes, and sought to develop a social psychological perspective on emerging patterns of intergroup contact and avoidance (Durrheim & Dixon, 2005).
This early work drew heavily on observational and interview data. The work that we conducted subsequently was based mainly around a series of large scale surveys, which employed nationally representative samples totalling over 4500 respondents (e.g. Dixon, Durrheim, & Tredoux, 2007; Dixon, Tropp, Durrheim, & Tredoux, 2010; Durrheim et al., 2011). The main goal of these surveys was to explore South Africans’ attitudes to the various forms of social and political change that were sweeping the country, including government policies introduced to redress the legacy of apartheid, such as educational desegregation, affirmative action, land restitution, and black economic empowerment. Understanding how, when and why such policies were supported (or resisted) seemed important to us because, despite the fall of apartheid, South Africa remained one of the most unequal countries on earth. Then, as now, its Gini Index placed it near the top of the list of countries in terms of the relative disparity in the distribution of wealth. Then, as now, the core problems of absolute (and not simply relative) poverty were starkly evident: food insecurity, malnutrition, shortage of housing, illiteracy, lack of access to basic medical care, mass unemployment, escalating crime, and disease (particularly AIDS and tuberculosis).
The results of our three surveys ultimately led us to question the limits of a prejudice reduction model of social change – not in a general or universal sense, but more specifically as a model for promoting meaningful change in South Africa. We had begun our research careers as passionate advocates of this model, schooled in its core theories, methods and assumptions. In particular, we were interested in using Allport’s (1954) contact hypothesis – a classic prejudice reduction theory developed originally in the US – to understand attitudes towards political transformation in the ‘new’ South Africa (for a recent review of the contact literature, see Pettigrew & Tropp, 2011).
Our initial data, hearteningly, suggested that favourable contact with black Africans had a modest but positive relationship both with whites’ racial attitudes and with their acceptance of government interventions to promote racial equality (e.g. Dixon et al., 2007), even if this effect was noticeably stronger for so-called race-compensatory than for race-preferential interventions (Dixon, Durrheim, et al., 2010). Perhaps more intriguing, however, were our data concerning the relationship between contact and the political attitudes of black South Africans, the ‘race’ group who suffered most under the apartheid regime and continue to fare worst on many socioeconomic indicators in the post-apartheid era. Here we again found that positive intergroup contact improved blacks’ emotional responses towards whites, as the contact hypothesis would predict, nurturing warmer feelings and decreasing social distance. Yet it also seemed to shift their political attitudes in a direction that, for us, was worryingly reactionary. That is, it was associated with diminished perceptions of own group relative deprivation and discrimination and decreased support for policies designed to redress the legacy of apartheid, particularly race-targeted policies such as affirmative action or land redistribution.
Meanwhile, beyond the South African context, several other research groups were reporting similar patterns of results, suggesting that our findings were by no means specific to the post-apartheid context (e.g. Glasford & Calcagno, 2012; Saguy & Chernyak-Hai, 2012; Saguy, Tausch, Dovidio, & Pratto, 2009; Sengupta & Sibley, 2013; Tausch, Saguy, & Singh, 2009; Tropp, Hawi, van Laar, & Levin, 2012; Wright & Lubensky, 2009). Their work spanned a wide spectrum of intergroup relations and contexts, exploring, for example, the consequences of contact between whites and blacks in the US, Jews and Arabs in Israel, Hindus and Muslims in India, and Maoris and whites in New Zealand. It employed a wider range of outcome measures than those used in our research, including measures of distributive justice, perceived legitimacy of the status quo, social mobility beliefs, systems justification beliefs and orientation towards collective action. It featured longitudinal and experimental methods as well as cross-sectional survey research methods. The overall message of this emerging work was consistent across research programs: positive intergroup contact encourages us to like one another more, it reduces negative intergroup stereotypes, it produces other positive outcomes such as forgiveness and empathy, but all these benefits come at the cost of decreasing the political awareness and resistance of the historically disadvantaged. Although the evidence supporting this idea is not yet conclusive, early signs indicate that we are dealing with a robust effect.
Yet why would positive contact with the historically advantaged produce this outcome for disadvantaged group members? Why, to echo Cakal, Hewstone, Schwar, and Heath (2011), might it exercise a ‘sedative effect’ on their political attitudes and behaviour? Our survey research provided evidence supporting two potential explanations (see, particularly, Dixon, Durrheim, et al., 2010), and our follow-up qualitative research suggested a third.
The first explanation focuses directly on what prejudice reduction interventions such as the contact hypothesis are designed to accomplish, namely to get us to like one another more. This is an invaluable end in itself in many contexts. In societies marked by relatively stable patterns of inequality, however, positive intergroup feelings may sometimes have a counterproductive effect, reproducing paternalistic relations, structures and ideological beliefs (see also Jackman, 1994, 2005). As a member of a disadvantaged group, for example, it is difficult to view those whom you like or esteem as the beneficiaries of discrimination. It is even more difficult to view them as its perpetrators. Similarly, if others value your display of group stereotypic behaviour (e.g. a woman behaving ‘femininely’), and reward this display with genuine admiration during the course of intergroup contact, then rejecting these ‘positive’ reactions as, say, merely part of a broader ideology of ‘benevolent sexism’ (Glick & Fiske, 2001) becomes a trickier proposition.
The second explanation of the paradoxical effects of contact focuses on its role in shaping attributions about the nature of patterns of intergroup discrimination. We found that positive experiences of everyday interaction with advantaged group members tended to reduce disadvantaged group members’ sense of being targeted personally for ill-treatment and that this individual level perception was readily generalised – arguably over-generalised – to shape their attributions about the nature of collective relations in the South African context. That is, positive interracial contact decreased black Africans’ estimates of the extent to which their group as a whole suffered discrimination in post-apartheid society via its effects on their perceptions of personal discrimination. Interestingly, this pattern of results qualified research on the so-called ‘person-group discrepancy’ effect, which posits the disadvantaged generally rate the discrimination suffered by their group as more severe than the discrimination they suffer personally (see, e. g., Taylor, Wright, Moghaddam, & Lalonde, 1990). That may be so. However, our work showed that perceptions of the two forms of discrimination also remain intimately linked, possibly because personal experiences of discrimination create a ‘false consensus’ effect that informs individuals’ assumptions about the experiences of other group members.
Our most recent work has explored a third explanation for the ironic effects of contact on relations of inequality and discrimination, an explanation that will broaden into a general discussion of the limits of a prejudice reduction model of change. This work has drawn on a qualitative case study of relations between workers and employers in South Africa’s domestic service industry (Durrheim, Jacobs, & Dixon, in press), an arena where intimate communication across racial lines occurs regularly. Although the treatment of domestic labourers has improved in the post-apartheid era, there remain few starker examples of the institutionalisation of racial inequality than those evidenced by working conditions of the million or so ‘maids’ who daily clean, cook and provide childcare to the more advantaged members of the society. Indeed, those working in this industry have recently been cast as “amongst the most exploited, isolated, and marginalised sectors of the working class” (Ally, 2008, p. 8).
At the same time, we found that interview accounts gathered from participants involved in domestic labour relationships were overwhelmingly positive. Notwithstanding the gulf in power and status between parties to the exchange, our analysis suggested that both ‘maids’ and ‘madams’ generally construed their day-to-day relations as fair and supportive. Their accounts were strewn with references to mutual obligation, heartfelt concern, and examples of helping that went well beyond the formal requirements of a labour contract (e.g. employers funding the education of workers’ children). Whilst not challenging the authenticity of such accounts, our analysis demonstrated how this way of constructing the meaning of everyday contact between members of advantaged and disadvantaged groups served as a vehicle for the transmission of systems justifying beliefs (see also Sengupta & Sibley, 2013). Whatever their other functions, such accounts were designed to resolve the political tensions associated with current participation in the historically racist institution of domestic labour by reframing contact between ‘maids’ and ‘madams’ in paternalistic terms.
‘What’s So Funny ‘Bout Peace, Love and Understanding’? The Limits of a Prejudice Reduction Model of Social Change
The latter line of analysis resonates with several strands of recent work that are united by a concern with how apparently positive intergroup evaluations may bolster the status quo. A first strand, for example, has shown how the endorsement of favourable or ambivalent beliefs about women may carry insidious consequences in terms of outcomes related to gender discrimination, both at a societal scale and within specific contexts such as employment (e.g. Glick & Fiske, 2001). A second strand has shown how processes of common identification, which involve the psychological inclusion of the disadvantaged and the advantaged within a superordinate category, may diminish members of both groups’ recognition of their differences and inequalities (e.g. Dovidio, Gaertner, & Saguy, 2009). A third strand has shown how positive intergroup behaviour such as helping may, paradoxically, bolster rather than challenge existing forms of hierarchy and dependency (e.g. Nadler, 2002). A final strand has shown, more broadly, how political structures of inequality are sometimes built around relations of (downwards) benevolence that serve to salve the conscience of the advantaged and to enrol the disadvantaged into bonds of loyal, even affectionate, subordination (e.g. Jackman, 2005).
The upshot of these and related strands of work (e.g. see Greenaway, Quinn, & Louis, 2011; Maoz, 2011) has been a growing critical awareness of the limits of the classic concept of prejudice as ‘irrational antipathy’. Glick and Fiske (2001), for example, have argued for a definition of prejudice that encompasses both positive and negative aspects, as captured in the notion of ‘ambivalent sexism’. Eagly (2004) has similarly argued for a more ‘inclusive’ concept that can accommodate the fact that positive general stereotypes (e.g. women are sensitive) can furnish negative situational outcomes (e.g. women shouldn’t be hired as ‘corporate bankers’). Jackman (2005) has argued that we must ‘dethrone’ the orthodox concept of prejudice as hostility, precisely because it has hindered our capacity to recognise the role of positive and ambivalent feelings in the reproduction of unequal relations within paternalistic social systems.
If we accept the validity of these recommendations, however, where does that leave the traditional project of prejudice reduction, which has been based on the ideal of getting us all to like one another a bit more (or at least dislike one another a bit less)? If positive emotions and beliefs about others can sometimes legitimate wider forms of discrimination, why should we prioritize prejudice reduction interventions as an effective mechanism for transforming such systems? We believe that these are fundamental questions that our field must address in the coming years. Finding adequate answers will not be easy.
One reason why concerns the relationship between prejudice reduction and the other model of social change most commonly advocated by social psychologists, namely the collective action model. The collective action model aims not so much to create more positive intergroup attitudes amongst the advantaged as to motivate the disadvantaged to challenge inequality (even if proponents also recognise that the advantaged may also be motivated to collectively challenge inequality under certain conditions, e.g. Mallett, Huntsinger, Sinclair, & Swim, 2008). Most social psychologists, ourselves included, have long assumed that these models simply reflect different catalysts of a common process of change that leads ultimately to a more equal world. That may still turn out to be the case in some circumstances. However, evidence of their potential incompatibility has steadily mounted in recent years, some of which we have discussed above.
Wright and colleagues have written most trenchantly on this topic (e.g. Wright, 2001; Wright & Baray, 2012; Wright & Lubensky, 2009), clarifying how these two forms of change, though seemingly congruent in their overall objectives, involve social psychological processes that take opposing trajectories. Prejudice reduction typically involves the creation of ‘positive’ intergroup emotions such as liking and trust; by contrast, collective action typically involves the creation of negative emotions such as anger and outrage at injustice. Prejudice reduction involves a reduction of the salience of intergroup boundaries; by contrast, collective action involves a strengthening of such boundaries, often accompanied by stronger political identification as a ‘disadvantaged’ group member. Perhaps most fundamentally, prejudice reduction strives to create harmonious relations that defuse intergroup tensions; by contrast, collection action strives to rupture the status quo, a process that typically increases rather than reduces intergroup conflict.
Given these fundamental social and psychological tensions – powerfully refracted within our own evidence on the effects of intergroup contact in South Africa – Wright and Baray (2012) have cautioned that:
Although it is clear efforts to reduce rampant antipathy, overt expressions of hostility, and active denigration of other groups would need to be part of a scheme to improve many intergroup relations, it also appears reasonable to consider the limitations of a focus on prejudice reduction, and recognize that it may actually directly conflict with another important means by which positive social change occurs – collective action. Failure to recognize these limitations will very likely lead us into the trap that many members of the advantage group seem to fall into – assuming that because interpersonal interactions across groups are convivial and warm that intergroup inequalities are either gone or are acceptable (p. 242).
Future Directions
In the wake of this kind of argument, social psychologists have been exploring several possible routes forward. Different styles of work have emerged that might be crudely categorized as: (a) incommensurablist, (b) compatibilist, or (c) contextualist in orientation. In offering examples of each of below, we want to emphasize that we are not pigeon-holing the overall perspective held by the researchers whose work is outlined. Rather, we are illustrating heuristic alternatives that are instructive at this juncture but will doubtlessly give way to more nuanced resolutions.
An Incommensurablist Approach
This style of work is continuing to map the tensions between prejudice reduction and alternative models of social change. Recent studies conducted by Glasford and Calcagno (2012) and Sengupta and Sibley (2013) provide instructive examples. Glasford and Calcagno examined the moderating effects of positive contact with white Americans (an historically advantaged group) on political solidarity between Latino and African Americans (two historically disadvantaged groups). They found that such contact diminished the effectiveness of an intervention designed to promote common political identification. Theoretically, it also reduced the likelihood that members of these disadvantaged groups might form political coalitions to act together to promote social change that might benefit both communities.
Sengupta and Sibley (2013) explored the relationship between intergroup contact, system justifying beliefs, and policy attitudes amongst Maori New Zealanders. The policy attitudes concerned the infamous Treaty of Waitangi, which, among other provisions, accorded Maori people historical rights of possession of the foreshore and seabed areas of the country’s coastline. As in our research on attitudes towards race-targeted policies in South Africa, Sengupta and Sibley found that positive contact with white New Zealanders was negatively related to Maoris’ support for the Treaty’s provisions. They also found that this effect was mediated by their perceptions that New Zealand is a fair and meritocratic society. In other words, contact affected policy attitudes through its capacity to transmit systems-justifying beliefs, encouraging the disadvantaged to oppose forms of reparation that favoured their own group.
A Compatibilist Approach
This style of work is exploring the conditions under which prejudice reduction interventions fail to produce a ‘sedative’ effect on the political beliefs and actions of the disadvantaged. In one of the first experimental studies on this topic, for instance, Becker, Wright, Lubensky, and Zhou (2013) qualified earlier work by showing that positive contact only reduces the collective action orientation of disadvantaged group members under certain boundary conditions. Notably, this effect occurs when dominant group members either defend the legitimacy of the status quo or keep their political opinions private. When dominant group members acknowledge publically the illegitimacy of the status quo during intergroup contact, however, the sedative effect of contact does not materialise. It appears, then, that the nature and content of intergroup communication is critical to understanding its relationship to collective action.
In a similar vein, Banfield and Dovidio (2013) have just revisited the question of how common categorization – a well-known prejudice reduction technique –shapes white Americans’ willingness to acknowledge racial discrimination. From a more complex set of findings, we wish to extract two themes. First, whites induced by an experimental manipulation to think of African Americans as part of a common ingroup (i.e. as ‘Americans’) were less ready to see them as targets of discrimination and, by extension, less willing to engage in forms of political protest to redress such discrimination. Second and conversely, whites who were induced to think of African Americans as having a ‘dual identity’ – i.e. as both black and American – were more willing to engage in this kind of protest. Thus, although common categorization may create forms of intergroup harmony that diminish recognition of racial injustice, category representations based on more complex dual identities may have the opposite effect.
Why might this be the case? The inclusion of subordinate and dominant group members in a simple common category may lead us to treat them as similar, perhaps even interchangeable, members of an extended ingroup. In so doing, it may quietly occlude the hierarchical nature of their relationship. This is what Dovidio et al. (2009) have evocatively termed ‘the darker side of we’. When common (re)categorization is complicated by a continuing recognition of sub-group identities, however, this kind of psychological – and indeed ideological – occlusion is less likely to occur. In this case, prejudice reduction is less likely to be accompanied by ‘false expectations’ (Saguy et al., 2009) about intergroup equality.
A Contextualist Approach
This style of work is advocating a move beyond generic assessments in order to evaluate the context specific role of different models of social change across different societies, cultures and institutional arrangements. This approach probably fits best with our own current position. By acknowledging what Cherry (1995) might label the ‘stubborn particulars’ of our own research, we are now beginning to re-evaluate its implications for relations elsewhere.
Most of our work to date has focused on intergroup relations rooted in a society characterized by deeply entrenched patterns of racial inequality. The history of South Africa, we believe, exposes clearly the limits of prejudice reduction as a model of social change. The collapse of apartheid had little to do with the project of getting the country’s diverse communities to like one another more or to abandon racial stereotypes. It had little to do with the many well-meaning interventions that sought to ‘bridge the gap’ between the communities that apartheid had torn asunder. In point of fact, intergroup attitudes became more, not less, polarised during South Africa’s transitional phase. Conflict intensified rather than abated. This is because the collapse of apartheid was driven by collective action. Millions of oppressed citizens recognized the injustice of the system, felt outraged, took to the streets, rendered the townships ‘ungovernable’, and through mass protest, often conducted at great personal cost, ruptured forever the status quo. This was by no means the only driver of change, but it was a vitally important one. For this reason, we believe those seeking to understand the conditions that facilitated change in South Africa should look not to some kind of gradual softening of the hearts and minds of the historically advantaged group. Rather, they should look to the particular historical, material and psychological conditions under which collective action was able to emerge, gather momentum, and become effective.
Having made this point, we also want to introduce some important qualifications. We recognise that not all societies are like South Africa. Their history, culture, social structure and political organization are not the same. Nor is the level of inequality that their citizens face. In some societies, the very question of who is advantaged or disadvantaged is far from clear cut, and the nature and consequences of intergroup ‘problems’ may thus be hotly contested (as, indeed, is increasingly the case in post-apartheid South Africa). In negotiating a path through these complexities, social psychologists must appreciate the concrete particulars of intergroup actions and interactions, the historical conditions under which they emerge, and the specific experiences and understandings of participants themselves, whose conceptions of social change may not fit into the neat categories of academic psychology.
For all of these reasons, we are not proposing a universal critique of how psychologists have framed the solutions to intergroup ‘problems’. We are not, to be clear, pitting the general virtues of a collective action model against the general failings of a prejudice reduction model. For example, in societies where relations of intergroup equality have been achieved, where systemic discrimination is under control and where conflict remains a problem, then we agree that prejudice reduction may be an important goal in its own right. In such societies, investing money and other resources into programmes to foster intergroup harmony may have an important role to play. However, in societies where inequality remains institutionally entrenched, then the route from prejudice reduction to meaningful change remains, for us, deeply uncertain. More worrying, when the ideological foundations of inequality are rooted not in overt antipathy and hostile intergroup behaviour, but in gentler, paternalistic forms of discrimination, then prejudice reduction may become part of the disease it is meant to cure.
Conclusions
We began this paper by reflecting on the dominance of liberal values in social psychology, a trend epitomised by a long tradition of research on intergroup prejudice. Indeed, what could be more progressive than a project explicitly aimed at diffusing intergroup tensions, promoting intergroup harmony, and challenging their associated forms of discrimination? Surely only the worst kind of cynic would find fault with such a noble project or carp about its alleged limitations? To put this question another way, echoing the title of the famous Nick Lowe song, ‘what’s so funny ‘bout peace, love and understanding’?
We respect the counter-argument to the kind of perspective that we have favoured here (e.g. see Haslam, 2012, whose paper’s title we have adapted), which strenuously defends the accomplishments of research on prejudice and rightly highlights the limitations of alternative models of social change. There is indeed much to admire about this tradition of work, and we do not doubt that prejudice reduction works in some contexts. Our contention is not that this project is inherently flawed. Instead, we argue that its hegemony within psychology (and beyond) has blinded the discipline to its fundamental limitations in societies characterised by long-standing, stable structures of inequality. For this reason, we hold that the entire ‘problematic’ of prejudice and prejudice reduction is ripe for a careful reappraisal in our discipline (Dixon & Levine, 2012; Wetherell & Potter, 1992).
As indicated, our own position has gradually evolved towards a contextualist perspective, whose implications are worth reiterating in conclusion. In some contexts, we agree, the project of getting us to like one another may be crucial to producing social change. Under conditions of comparative equality between groups, for example, then the project of promoting ‘peace, love and understanding’ is often intrinsically valuable. In other conditions, however, prejudice may be an epiphenomenon that distracts us from the main causes of, and solutions to, problems such as race, class or gender discrimination. In still others, with an irony that is increasingly evidenced by research, prejudice reduction may actually become part of the problem that it is designed to solve, diminishing the extent to which members of historically disadvantaged groups acknowledge and challenge broader structures of social injustice. The point of our paper has been to engage a broader audience in this emerging debate about the conditions under which prejudice reduction interventions are effective.