Tyranny, freedom and social structure: Escaping our theoretical prisons


  • John C. Turner

    Corresponding author
    1. Australian National University, Australia
      Correspondence should be addressed to John C. Turner, School of Psychology, Australian National University, Canberra, ACT 0200, Australia (e-mail: john.turner@anu.edu.au).
    Search for more papers by this author

Correspondence should be addressed to John C. Turner, School of Psychology, Australian National University, Canberra, ACT 0200, Australia (e-mail: john.turner@anu.edu.au).


Reicher and Haslam's (2006) BBC prison study undermines the idea that people passively accept and enact social roles. In this commentary, I point out that this idea is an example of Moscovici's (1976)conformity bias and a wider stability bias in social psychological theorizing. In many key areas, the science prefers analyses that explain how and why social structures, intergroup and power relations, personalities and beliefs maintain and reproduce themselves, and indeed cannot be changed, rather than how and why society constantly generates forces for social change from within itself. This bias distorts reality and produces ideas of limited theoretical or practical power. Human psychology does not make us prisoners of social structure. It makes us capable of collective action to change social structures and in turn re-fashion our identities, roles, personalities and beliefs. Society is not a psychological prison but a means of expanding human possibilities. A reorientation of theoretical emphasis is overdue.

The original Stanford Prison Experiment (SPE; Haney, Banks, & Zimbardo, 1973) seemed to show that ordinary people assigned to the roles of prisoner and guard would naturally and automatically accept and enact these roles and their associated norms for behaviour. The idea that people as guards would act in a tyrannical and brutal fashion and that patterns of domination and submission seemed to follow directly from role assignment was a dramatic one that captured people's attention and imagination. There was also a larger point behind it that resonated with one of the big three metaphors of social science for understanding society, the notion of society as prison. The study seemed to show with great immediacy that there was, in human nature, an inherent tendency for individuals to act as the passive vehicles, indeed victims, of social structures and forces over which they had no control and which constrained their actions. Our psychology made the individual a prisoner of social determinism. This general idea, which Asch (1952) described as a modern sociological version of the group mind thesis, is not as pervasive as the idea of society as organism or society as theatre, but it is still powerful and exerts a particularly pernicious effect in the areas of prejudice, power and personality.

The great merit of Reicher and Haslam's (2006) BBC study is that it shows the problems with the idea of the individual as a prisoner of roles with the same immediacy and naturalism that was a hallmark of the original study. In the BBC study, people do not automatically identify with the roles to which they have been assigned. One group accepts and one group rejects the provided identity, and, paradoxically (for the personal self-interest view of group formation but not self-categorization theory, see Turner & Bourhis, 1996), it is the group identity with status, ‘power’ and resources that is rejected, while members who do accept their group identity still feel free to dissent from the provided meaning and reinvent it. Instead of a picture in which a universal human nature seems to be at work – ‘provide the role identity and they will conform’ – a much more interesting theoretical question emerges. Under what conditions do people accept or reject role identities, under what conditions and how do they challenge and change the meaning of the identities provided, under what conditions and how do groups with apparently little authority, power, resources or status come to redefine their position as illegitimate and create from nothing the social and political power that allows them to exert their collective will and change the social system?

These are all questions to which modern social psychology does have answers. It has been working on the answers for many years, particularly, but not only, in the social identity tradition. But both the questions and answers have been marginalized in the science's official consciousness because, strangely, the picture of human nature apparently illustrated by the SPE seems more attractive to many of our colleagues. They find it a more reasonable and plausible political picture of what goes on in society. Those of us who have never found the SPE picture at all plausible as a historical or political story can now point to the BBC study and say ‘look, when one gives people a chance to act reasonably naturally, with some choice, as they might in reality, over time, then there are instances when far from conforming to imposed roles they reject or change the roles, they reject and change the social structure’. This is not to say that Zimbardo and colleagues' data were flawed. We need not doubt that what Zimbardo et al. said happened did happen, but if we know that it does not always happen, then that is enough to change the whole theoretical problem and metatheoretical lesson.

The data have big implications for other theories tied to the idea that human beings are constrained by an inherent bias towards social stability and conformity to the predefined social structure. The idea of system justification (SJT) and social dominance theories (SDT) that people are inherently motivated to maintain the status quo even to their own disadvantage by derogating themselves or legitimizing the social system that disadvantages them is not supported by these findings. Not much sign of any such ‘ubiquitous’ motive (as Sidanius, 1993, called it) in this setting. It was the dominant group that felt uncomfortable and guilty with the social hierarchy and the prisoners who asserted their self-interest and identity aggressively to overthrow it (SDT's behavioural asymmetry in the wrong direction!). Nobody said ‘let's put ourselves down to maintain the hierarchy’ and wonder of wonders for a social psychology experiment, when participants had the chance, they changed the system. Anyone reading the mainstream literature on power and prejudice in social psychology could be forgiven for thinking that collective revolt and social change never happen. Theories like SJT and SDT invoke specific motives to explain why social change does not happen. Theories like symbolic racism and the whole family of implicit prejudice analyses tell us that political changes in attitudes and beliefs are only superficial because negative affect and stereotypes learnt through early socialization persist in a cultural unconscious no matter what one believes consciously. Anyone who looks outside the window at daily events around the globe will find that, as in the BBC study, resistance, conflict and change are as normal as the sun rising. How did our theory get to be so one-eyed? A big part of the answer, relevant to influence and power, is what Moscovici (1976) referred to as the structural-functionalist conception of society in his critique of the conformity bias. Another has to do with the development of the prejudice field in the USA, in which the particular history and politics of race relations in that country and its commitment to individualism have dominated social psychology's theoretical approach to social conflict.

Theories that tell us explicitly that nothing can be done about social domination and racism because of biology or evolution or the socialized unconscious are only the tip of the iceberg. The whole prejudice field has come to imply the inevitability of racism in human nature. The mainstream field has arrived at the view that prejudice is universal, automatic and inevitable, that it is ‘banal’ (Fiske, 2000). The Allportian (1954) idea, for example, that prejudice is inherent in the universal process of social categorization, which ‘mistreats’ individuals as group members, is lauded as a great discovery (despite a sustained critique by self-categorization theorists). Why are we prejudiced? Because, the general story goes, of cognitive, motivational and emotional processes in our individual psychology such as needs to belong, for self-esteem, for meaning and so on, social categorization, in-group identification, frustration, anger, anxiety, negative affect and so on, and because we are slaves to our social past in the development of our personalities and in the socialization, conformity processes and cultural learning which dictate what we think. Despite theoretical statements that tell us that the conscious and unconscious are highly interdependent, we nevertheless find plausible the idea that a person who thinks he/she is not prejudiced is actually prejudiced unconsciously, because the unconscious is the reservoir of cultural learning that somehow dominates over the social present. In the SPE, what prisoners and guards do is a function of an immediate situational role socialization. In the Allportian tradition, prejudice is a function, inter alia, of personalities, affect and attitudes that embody long-term socialization. In both cases, socialization works against the personal will and freedom of the individual (as in implicit prejudice theories for example). Never mind that no adequate data have ever supported the hypothesis of unconscious racism despite conscious anti-racism (e.g. see Locke & Walker, 1999), it is cited in every textbook because, like the SPE, it fits the desired metaphor.

It is not enough to say that a person is racist because their society is racist. Why is the society racist and why are others not? Why did this person conform, when others resist, fight and change society? Neither children nor adults passively accept what their society teaches (Milner, 1996; Nesdale, 2001) and nor is conformity passive (Turner, 1991, 2005). This theory (that people are racist because they have no choice but to accept a racist culture, consciously or unconsciously) implies an inevitability to racism which is false and derogates human nature. It expresses a strong tendency in contemporary social psychology to formulate theories that make the present inevitable rather than explain why it is temporary, a stability bias. The prejudice field does this by looking for causes in either abstract human nature or prejudiced societies and cultures, despite the clear alternative developed by Sherif and Tajfel of looking to the role of intergroup relations within the politically, economically and historically organized social environment.

The bias is also found in the standard views of power and personality (Moscovici, 1976; Turner, 1991, 2005; Turner, Reynolds, Haslam, & Veenstra, in press). As Moscovici first pointed out, the thesis that a source's influence is based upon the target's outcome dependence upon the source implies that influence only flows in one direction, from those with power (control of resources) to those without power. But if all influence reflects power and flows from the top down, then how does social change take place in which the ‘have-nots’ reject and change the status quo? Anyone who seriously wants to understand social change and the role of power hierarchies in it has to begin by rejecting the dependence theory of power which the field thinks is common sense. In fact, the data clearly reject the supposedly common sense theory if one reviews them seriously rather than just repeats tautological clichés. In terms of personality, the challenge which Kurt Lewin put before us of developing a dynamic conception of personality in which individual behaviour was seen as the contemporary product of a field of current social and psychological forces has been almost completely forgotten in favour of a social learning/Freudian-influenced template. In this template, individuality reflects relatively stable psychological structures that interact with the current situation but were themselves laid down in the past through socialization, early learning and maturation. Personality, it appears, reflects past social learning (or evolution) and only changes through relatively slow processes of individual development. A whole of generation of psychologists has come to accept as virtually axiomatic the idea that personality is hard to change and persists as a source of inertia in the present. This despite all the evidence of social psychology of the incredible flexibility and variability of human cognition and behaviour in response to contemporary social and situational influences. Beginning with the problem of the prejudiced personality, my colleagues and I have begun to try to show that other ways of understanding personality, consistent in principle with personal and social change and with the classic Lewinian position, are possible.

Why does this matter? Because a bias towards social stability distorts not merely reality, but also the kinds of theory we develop. Three problems which follow from the stability bias are: (1) a preoccupation with description instead of explanation, leading often to the reification of empirical relationships as theory; (2) the construction of theory in terms of an abstract, homogeneous, socially unstructured individual, which is a purely ideological fiction; and (3) a refusal to confront the political, collective, ideological dimension of human social life and consider what kind of social psychology makes this dimension possible. It is this bias that explains why we still have an indefensible commitment to individualism in the science. To understand social change, one must look at the psychological group as well as the social individual and at both their difference and interdependence. It is the group through which people are able to change things collectively and politically, and hence ultimately change personalities and individual attitudes. Lewin and others knew this, but modern psychology still finds it an uncongenial fact. Reicher and Haslam have shown beautifully simply the power of the group to change not only social roles but also individual psychological states. This is not an argument for social constructionism, meaning that there is no human nature or that it is infinitely malleable; it is an argument that our psychology includes social-psychological capacities for collective self-regulation and development. Henri Tajfel used to summarize these points by saying that we must develop a social social psychology. He is still right. The good news is that we have made advances in formulating the concepts and theories to do it and that many researchers are making important contributions. A paradox is that the stability bias to which the science adheres holds back recognition of its many achievements.

Because the SPE had a big metatheoretical lesson, so too does the BBC study raise the big issues, but from a quite different and a much more positive perspective. The idea of social structure as something that limits human freedom, that constrains the individual to the endless repetition of past social forms, is alive and well in contemporary social psychology. It has a dulling and regressive effect on our thinking and limits our contribution to social progress and reform. It is one reason why social psychology is so marginal politically and has little political clout. A theory that says that we are condemned to play out social roles invented in the past (by whom and how they were invented is never very clear), as if we can never resist and reinvent them now, is a theory of social stability, of endless stagnation. To show how society changes there must be a dynamic and fully ‘interactionist’ (Tajfel, 1972; Turner & Oakes, 1986) social psychology. I believe this will produce a much richer and more respectful view of human psychology, one that will liberate study of the individual and personality as well as the psychological group. Instead of studying, for example, how the psychology of personality limits and prevents real social and political change, we should be studying how political and ideological changes create new personalities and individual needs and motives. It is the latter that we can see all around us and the former that has failed to make its case after 50 years. Lewin's project of a dynamic conception of personality arising as an outcome of varying forces in the contemporary life-space needs resurrecting as urgently as Sherif's vision of an intergroup perspective on prejudice did in the 1970s.

Zimbardo's SPE was an important piece of research because it showed the power of social psychology to address problems of human social organization and political life and transform people dramatically in short time spans. These are two lessons more forgotten than remembered, but which are still central to social psychology's future. A great service of Reicher and Haslam's BBC study is to resurrect the lessons in the context of a substantive and very worthwhile theoretical debate about the person-role relationship and more generally the way in which society is conceptualized in social psychological research. Behind the tyranny of the prison guards and the abasement of the prisoners in the SPE, there is a view of human beings as the psychological prisoners of society, in turn a working out of a dysfunctional and inescapable human nature. It is never put so bluntly in our textbooks and journals, but it is there and it exerts a pervasive influence and constraint on our ideas and limits our message to society. Social psychology spends much of its time explaining how society is reproduced, how the present recapitulates the past and very little on the other half of the problem, how and why society changes, how the future is created in the social present. Such a huge distortion of the defining problem cannot but harm the science and indeed it does. By the same token, broadening the theoretical focus to embrace the forces for social change as well as social stability, to find ways of explaining why the present does not persist and history does not end, of how people are driven to react and change by the present, holds promise of a new maturity, richness and practical power.


This article was made possible by a grant (DP0342645) from the Australian Research Council. Thanks to Kate Reynolds for her helpful comments.