turner on power

European Journal of Social Psychology

Eur. J. Soc. Psychol. 35, 1–22 (2005)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ejsp.244

Agenda 2005

Explaining the nature of power: A three-process theory

JOHN C. TURNER*Australian National University, Australia

Abstract

Power is an inescapable feature of human social life and structure. This paper addresses the nature of

power. The standard theory is that power is the capacity for influence and that influence is based on the

control of resources valued or desired by others. However, there have always been problems with this

theory and new ones have appeared. The paper summarizes the standard theory and its problems,

outlines the different meanings of power and presents a new theory emphasizing group identity, social

organization and ideology rather than dependence as the basis of power. It proposes that power is

based on persuasion, authority and coercion. A key point is that the theory changes the way these

processes have been understood by reversing the causal sequence of the standard theory. The latter

argues that control of resources produces power, power is the basis of influence and that mutual

influence leads to the formation of a psychological group. The three-process theory argues that

psychological group formation produces influence, that influence is the basis of power and that power

leads to the control of resources. Implications of the theory for social change, coercion, prejudice and

the extent to which power is a social evil are briefly noted. The challenge is to study how power

emerges from and functions within social relationships with a definite social, ideological and

historical content rather than reifying it as an abstract external force producing generic psychological

effects. Copyright # 2005 John Wiley & Sons, Ltd.

Few would deny that power is central to human affairs or a key concept in the social sciences. Roberts

(2003) argues that it is the basis of society: ‘The ability of one person to make a hundred others do his

[sic] bidding is the basic building block upon which all collective human endeavour is based’ (p. xv).

Keltner, Gruenfeld, and Anderson (2003) state that it is ‘a basic force in social relationships’ (p. 265).

It appears to be a universal and indispensable feature of social organization, at work in all political,

organizational and institutional life and, in some views, in every social relationship. Every group,

organization or society must solve the problems of power to achieve its goals or risk failure,

dysfunction or even extinction. But what exactly is power? How should it be defined, what are the

processes through which it operates, what are their distinctive effects on attitudes and behaviour, and

Received 25 August 2004

Copyright # 2005 John Wiley & Sons, Ltd. Accepted 25 August 2004

*Correspondence to: J. C. Turner, School of Psychology, Australian National University, Canberra, ACT 0200, Australia.E-mail: [email protected]

Contract/grant sponsor: Australian Research Council; contract/grant numbers: DP0342645.

how is power gained and lost? These questions define the problem of the nature of power and it is this

problem1 I shall seek to address in this Agenda article.

Social psychology’s first attempt to solve the problem, which is still the standard theory, emerged

from the group dynamics tradition in the 1950s as it translated Lewinian notions of ‘own’ and

‘induced’ forces in the lifespace into a detailed analysis of social influence processes (Festinger,

1953). This theory defines power as the capacity to influence and argues that it is based upon the

influencing agent’s control of resources desired or valued by the target. Power is the potential to

influence and influence is the exercise of power (Bass, 1990; Fiske & Depret, 1996; French & Raven,

1959; Moscovici, 1976; Pfeffer, 1992; Turner, 1991). One consequence is that power has tended to be

studied as an adjunct to social influence rather than in its own right. In recent decades this has changed.

Critiques of the dependence theory of influence have appeared which reject the equation of power with

influence (Moscovici, 1976, 1980; Turner, 1991) and research on power in the context of intergroup

relations has provided new perspectives (e.g. Fiske, 1993; Jost & Major, 2001; Reicher & Levine,

1994a, 1994b; Sachdev & Bourhis, 1985; Stott & Reicher, 1998; Sidanius & Pratto, 1999; Turner &

Haslam, 2001; Tyler, 2001). It has become possible to redefine the nature of power as having more to

do with the basis of organized, collective action than to do with a dependence relationship.

In this article I try to provide a new way of thinking about power, one more consistent with the

currents of contemporary research. I point to problems with the standard theory and then outline a

different causal perspective on the origins of power. Whilst what is proposed will bear similarities to

standard ideas, the basic causal analysis will be different, with the result that both general implications

and specific empirical predictions will differ.

THE STANDARD THEORY OF POWER AND ITS PROBLEMS

The standard theory is not one specific formal theory but rather a set of general assumptions about the

relationship between power and influence which is shared by all the classic theories of social influence

(Deutsch & Gerard, 1955; Festinger, 1950, 1953, 1954; French & Raven, 1959; Kelman, 1958). The

basic ideas are that power is the capacity to influence other people, that it is conferred by the control of

resources (positive and negative outcomes, rewards and costs, information, etc.) that are desired,

valued or needed by others and which make them dependent upon the influencing agent for the

satisfaction of their needs or reaching their goals, and that different types of resources confer different

types of power leading to different kinds of influence.

A further important idea widely shared and relevant to the meaning of power is that some kinds of

influence are more informational or cognitive in character, reflecting needs for information or to

reduce uncertainty, and lead to private acceptance (private attitude change in line with the influence

attempt), whereas other kinds are more social or goal-oriented in character, embodying group pressure

or compulsion, and lead only to public compliance. If one of power’s meanings is the idea of

compulsion, force or domination in contrast to persuasion, then the standard theory that power is the

capacity to influence also contains a sub-theory that outward compliance without inner conviction is a

function of particular kinds of non-informational resources (social approval, rewards and costs)

controlled by the influencing agent.

The standard theory has been extremely successful. It has been applied to a huge range of

phenomena and is still after 50 years the only theory of power that ever gets a mention in social

psychology (e.g. see Keltner et al., 2003, for a recent restatement). Nevertheless, there have always1I shall address the problem solely from a social psychological viewpoint. It is not possible to consider the contributions of othersocial sciences in this article, which does not diminish their relevance.

2 John C. Turner

Copyright # 2005 John Wiley & Sons, Ltd. Eur. J. Soc. Psychol. 35, 1–22 (2005)

been problems with the standard theory and over the last 30 years additional ones have emerged. The

most important can be summarized as follows.

First, the field has never agreed on one view about how the dependence-influence relationship

works and the differences between the classic theories are not trivial. The seminal theorists disagree

about whether there is one (Festinger, 1950), two (Deutsch & Gerard, 1955), three (Kelman, 1958) or

five, six or more (French & Raven, 1959; Raven, 2001) kinds of influence process. They disagree about

the role of group cohesion in compliance. For Festinger (1950, 1953, 1954) being attracted to the

group leads to private acceptance of its influence by members and public compliance reflects dislike

for the influencing group. For Deutsch and Gerard (1955) attraction to the group leads to ‘normative’

not ‘informational’ influence (i.e. compliance). French and Raven (1959) and Kelman (1958) explain

compliance in terms of rewards and costs which are distinguished from both informational processes

and attraction to others (‘referent power’ or ‘identification’).

Even more fundamentally they disagree about the relationship between informational and social

processes of influence. Festinger (1950, 1954) regards them as inseparable and assumes that people

form groups because of their need for information. Deutsch and Gerard (1955) split informational

from group influences, while French and Raven (1959) and Kelman (1958) identify social processes of

influence which are neither informational nor merely compliance (i.e. a kind of acceptance based on

emotional ties). Also highly confusing is that the latter theorists both specify rewards and costs as the

basis of compliance as if other kinds of power did not also involve rewards and costs. In dependence

theory all influence is shaped by rewards and costs (see Keltner et al., 2003). In informational

influence or social reality testing, for example, one shifts towards people with information or expertise

because they provide information which is rewarding in reducing one’s uncertainty. How the rewards

and costs specified as the basis of compliance in these theories differ from the rewards and costs in

other kinds of influence is never explained.

Second, the basic idea that influence reflects dependence is highly problematic (see Moscovici,

1976, and Turner, 1991, for detailed critiques). If one considers the popular dual-process theory of

Deutsch and Gerard (1955), for example, which contrasts an informational process with normative

influence in which one conforms to the expectations of others to gain their approval, neither process

provides an adequate description of the phenomena they are meant to explain. Private conformity is

not a function of people who lack information shifting to people who have it. In fact, even people

confronting an unambiguous stimulus (as in the Asch paradigm) become uncertain when similar

others disagree with them and what reduces uncertainty is subjectively valid information, not

information in the abstract. The perceived validity of information is always a function of social and

relational factors such as the perceived source of a message, the degree to which it has consensual

support and the degree to which the target defines the source as a positive reference group, that is, the

degree to which it is in line with ingroup norms. So-called informational influence is not purely

cognitive but also social and normative. There is in fact no way of defining persuasive or valid

information independently of the social context within which it is apprehended. The same information

which persuades one group will fail to persuade another. One group’s expert is another’s crank. One

does not accept influence from experts because of the information they provide (if one is not an expert,

how can one judge its quality?), but accepts the information as valid because one defines them as an

expert (Moscovici, 1976).

Similarly, the idea that normative influence in the Asch paradigm is non-informational group

pressure is inconsistent with classic findings such as the importance of unanimity rather than group

size in producing conformity and the fact that conformity is still strong even when group surveillance

is removed. Basic findings and indeed basic phenomena (conformity in the classic paradigms, group

polarization, minority influence) do not fit and remain unexplained by the dual-process theory (Turner,

1991). These and other considerations imply that neither cognitive nor group aspects of influence can

Explaining power 3


be reduced to dependence for information or social approval. Moscovici (1976) argues that social

conflict, not informational dependence, is the crucial process in persuasion. Turner (1991) argues that

it is reality testing in the context of a shared social identity.

A third issue is that the standard theory does not sit easily with the facts of historical and social

change. It seems to rule out social change, innovation and minority influence (Moscovici, 1976). It

implies that influence flows only in one direction, from the top down, from the ‘haves’ to the ‘have-

nots’, from those with power to those without it. But there are countless historical examples of social

movements, from the scientific (Freud) to the political (Hitler) to the religious (Christianity), in which

power and control of resources did not precede but followed from processes of influence. New

movements can gain adherents despite often lacking resources, expertise and prestige. The latter are

often the badges of success rather than preconditions. If the standard theory is true, then how is it that

social change from below ever takes place, how is it that subordinate groups for example, do come,

under certain conditions, to reject the legitimacy of the social order and oppose it?

A fourth problem already touched on relates to the role of psychological group formation. The

standard theory is an important component of the social interdependence analysis of the psychological

group (see Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). It is more fully understood as an

implicit causal sequence that runs from ‘resources’ to ‘power’ to ‘influence’ to ‘psychological group

formation’. It is assumed that control of resources gives people power, defined as the capacity to

influence people’s attitudes, beliefs and behaviour, and that influence between people who depend

upon each other leads to shared social norms and values. It is also assumed that positively

interdependent people tend to become cohesive (attracted to each other) and cooperate. Thus mutual

dependence between people is assumed to produce a psychological group, a collection of individuals

sharing mutual relations of cooperation, cohesion and influence, which have become stabilized over

time into a social structure of roles and norms. In this view, then, power and influence precede causally

and theoretically the formation of a psychological group (Turner, 1991; Turner et al., 1987).

But a close examination of the standard conformity paradigms suggests that the psychological

group is a precondition of influence, not simply an outcome. For example, what is experienced as

uncertainty, producing a need for information, cannot be defined solely with reference to the objective,

physical stimulus situation. Uncertainty arises as a social product of disagreement between people

who expect to agree in the given situation. It is where people who expect to agree find that they

disagree that they become uncertain and seek to resolve matters. Self-categorization theory (Turner,

1985) argues that people expect to agree where they define themselves as members of the same group,

in terms of the same social identity, and confront the same shared stimulus situation. If we are the same

in relevant respects and judging the same reality, then we ought to agree. On the other hand, if we are

different, then there is no reason to experience uncertainty about the validity of one’s judgement

simply because we disagree (e.g. David & Turner, 1996).

Thus group formation is a precondition for the experience of uncertainty which sets in train the

processes of informational influence. It is also a precondition for the experience of subjective validity

(confidence in the validity of one’s views), for it is only when similar others agree that one can feel

confident that one’s views reflect the objective public reality rather than personal biases or systematic

error. One accepts information and one seeks approval from those who share one’s beliefs, values and

objectives, from members of one’s ingroups. What this points to is the importance of group formation

rather than dependence as the basis of power and the hypothesis that group formation reflects identity

not dependence.

Finally, power and influence are not properly distinguished in the standard theory. They are

confounded in the general view that all influence reflects a process of submission to the power of

resources (at the same time as, confusingly, some kinds of dependence are supposed to be more

specifically to do with submission to power than others because they produce compliance). Nowhere in

4 John C. Turner


the standard theory is there a place for the sharp antagonism between influence and power commonly

found in everyday discourse, as persuasion versus naked force or domination. The attempt to deal with

this issue through the idea of compliance based on reward and coercive power (French & Raven, 1959;

Kelman, 1958) fails theoretically since all influence is based on rewards and costs in the dependence

formulation. There is also the problem that the conventional concept of compliance does not capture

the idea of naked force or compulsion as a strongly conflictual process. It is still seen as a form of

positive influence in which the powerless are shaped by the powerful, as a form of ‘going along with

the group’.

It is also worth noting that although coercive power is grouped with reward power as a component

of compliance, it functions very differently and produces different results (Festinger, 1953; French,

Morrison, & Levinger, 1960). Further, research on attitude change in the reinforcement and dissonance

traditions has long shown that both rewards and costs can affect private attitudes, each able to increase

or decrease attitude change under specific conditions, which is problematic for this whole approach to

compliance (as public conformity). Even in the power literature, rewards and costs can sometimes

increase and sometimes decrease private attitude change (e.g. Leet-Pellegrini & Rubin, 1974;

Veenstra, Turner, Reynolds, Haslam & Burgess, 2004). Neither is supposed to happen in the

compliance formulation. Although compliance based on rewards and costs appears as a common-

sense process under the influence of the standard theory, there is some doubt as to whether the process

as defined actually exists.

There is also evidence for the idea that what compels outward compliance may ensure private

rejection (i.e. attitude change in the opposite direction to that intended; e.g. Brehm & Brehm, 1981;

David & Turner, 1996; Moscovici, 1980; Raven & French, 1958; Veenstra et al., 2004). The standard

theory barely acknowledges the notion of private rejection since it is inconsistent with its basic thrust

that control of rewards and costs provides power and power is ‘the production of intended effects’. A

theory of the nature of power must explain how and when the demands of the powerful are rejected as

well as when they are effective.

In sum, in the standard theory all forms of influence, from the private to the public, reflect power

based on control of resources and social interdependence for positive outcomes is seen as the basis of

the psychological group. But, in fact, there is no unanimity about how different types of resources

produce different types of influence and the general analysis is problematic. Persuasion is assumed to

flow from informational dependence on others to satisfy some purely individual need when the very

need for and definition of information arises from group and social relationships. Persuasion is not a

process of submitting to people with superior resources, but one of negotiating and validating reality

collectively within one’s reference groups. The standard view also paints a picture of compliance

which fails to distinguish between going along with people one likes and being coerced against one’s

will. All influence is conceptualized as if it were a kind of submission, moving in the direction dictated

by those with resources, because one has no choice, because one is dependent, at the same time as real

coercion (force, domination, compulsion) is lost sight of. The standard theory fails to deal adequately

with the role of psychological group membership and the facts of social and historical change as well

as with the nature of persuasion and coercion. To go beyond the standard theory we need to begin by

considering what a theory of power is meant to explain. What do we mean by power?

THE MEANINGS OF POWER

Defining power has long been a murky business. As Fiske and Depret (1996) comment, we use the

term in ordinary parlance confident that we know what it means until we are asked to define it. It has

Explaining power 5


proved difficult to differentiate consistently from related constructs of influence, compliance, control,

dominance, authority, status and rank. Part of the problem is that there are several legitimate meanings.

An adequate theory must be clear about what it is trying to explain and this in turn requires that the

various meanings of power be spelled out in relation to each other. This is the task to which we now

turn.

The most general meaning of power found in the literature and everyday thinking is that it is the

capacity to cause effects, to have an impact on or change things, to do ‘work’, either in the physical or

social world. At this level of abstraction power belongs to things as well as people and affects things as

well as people. A sub-category is power as the capacity in some way to affect people or society, to

cause them to do things that they would not otherwise have done. A person, institution, physical event

or idea may be described as powerful because it is having a big impact on society; it affects what

people do and think and how they live. This category can be additionally restricted to effects produced

by people, intended or unintended. The idea of a person or group having an impact on others, causing

them to think, feel or act in ways they would not have otherwise, I shall term ‘power as impact’. A

further sub-category is power as the capacity of people to have an intended influence on others and this

is how most theorists have defined ‘social power’. It is the definition that permeates the standard

theory. It rests on an analogy between the processes by which people influence others, to produce

desired changes in their psychological and behavioural states, and the way in which people can exert

some physical power over objects.

But there is also a kind of power which only emerges from human social relationships, from the

capacity of people to organize themselves into groups, institutions and societies. People acting in

concert, cooperating, coordinating and unifying their actions, developing emergent capacities as

members of social systems, are able to have an impact on both the physical and social worlds through

their interrelatedness that would be impossible if they were purely individual beings. Power from this

perspective is the capacity to affect the world, including others, through influencing and controlling

people to carry out one’s will, to act on one’s behalf, as an extension of oneself. It is social power in the

sense that it is a peculiarly social means of having an impact, not because the impact is on people

rather than things. The emphasis here is on getting people to act in line with one’s desires, beliefs,

orders, instructions, commands, suggestions, etc. in order to have an impact on the world through them

(see Figure 1). Power to affect the world and power through people are obviously related. The more

people, groups, organizations one can get to act as collective embodiments of one’s will the bigger the

impact one can have on the world and so reference to ‘power through’ often connotes ‘impact on’ and

vice versa. Part of the general argument to be made is that power through people provides the resources

for impact on the world. Whilst accepting this implication, we shall nevertheless for the moment define

power ‘through people’ as the core social psychological problem of power. How do individuals or

groups influence and control other people to carry out their wishes, suggestions or commands

(Roberts, 2003)?

The standard theory tends to obscure the difference between interpersonal influence and both the

power to affect the world through collective organization and action and the power to create and direct

the collective will of people. It fails to distinguish power through people and impact on them. The

former can be seen as a subcategory of the latter logically but psychologically the meaning of the

process is very different. The problem of social power is not simply how one affects others, I suggest,

but how one gets them to act on one’s behalf, to carry out one’s will, to become part of a

psychologically cohesive social organization.

‘Power through’ can be subdivided further. There is power as influence as opposed to control. This

is influence in the sense of persuasion (cognitive change, conversion), the power to get people to act in

line with one’s desires by persuading them that the desired judgement, decision, belief or action is

correct, right, moral, appropriate. If one can persuade others of the correctness of some belief or the

6 John C. Turner


Fig

ure

1.

Th

en

atu

reo

fp

ow

erth

rou

gh

peo

ple

Explaining power 7


rightness of some action, then they are likely to act on it as a matter of their own volition, as free,

intrinsically motivated and willing agents. The capacity to influence and persuade, to change people’s

private attitudes, values and beliefs, is one basis of ‘power through’ and although often regarded as

‘soft’ is in fact, I shall argue, the decisive process. Influence in contrast to control is not the same as the

exercise of social power in general but is the much more specific idea of changing people’s private

beliefs and attitudes through persuasion.

Power as control is here defined as the capacity to get people to do what one wants where they are

not persuaded of or are uninterested in the validity of the specific belief or act. I suggest it takes two

forms. Legitimate authority is control based on the acceptance by the target of one’s right to prescribe

their beliefs, attitudes or actions. One may not be persuaded that some action is right but one is

persuaded that some person, group or role has the right to prescribe it, by virtue of their position within

the group structure, by virtue of the group norms and values which legitimate such control. Such

control is based on voluntary deference to and private acceptance of ingroup authority. It is not

experienced as an oppression of self or loss of power, but as voluntary submission to a collective will

with which one identifies and can result in a sense of empowerment and pride (Depret, 1995; Tyler &

Degoey, 1995).

Finally there is control as coercion, attempting to control others against their will. In this case some

individual or group, being unable or unwilling to exert influence or authority over a target, resorts to

deploying the human and material resources it controls to constrain, block, compel and manipulate the

target’s behaviour without any effort at persuasion or appeals to legitimate authority. Coercion is an

unusual or one might say a limiting case. It is power ‘through’ in the particular sense that it is directed

to making the target an extension of one’s will on an involuntary basis. It attempts to get the target to

act in line with one’s own will by forcing them to act against their own, by restricting or opposing the

freedom of the target to pursue its own goals. It shades into power as impact, in affecting but not

converting the target, and betokens a more or less conflictual stance which can rapidly degenerate into

open conflict. Traditionally the distinctions between persuasion and control and between voluntary

and involuntary action have been equated in the contrast between private acceptance and compliance,

but in fact these are not the same. Acting without knowing what is the correct response can be

voluntary. Being controlled is not the same as acting against one’s will. Thus neither control nor

coercion as here defined can be equated with the traditional notion of compliance. The latter obscures

the crucial process underlying social organization that people are able to act in terms of a collective

rather than personal will.

Others may prefer different terms to those above but it is the substance of the analysis which

matters. All the meanings of power in current use are specified in relation to each other. As we move

from power in the most general sense down to coercion, we have illustrated the particular

characteristics of each meaning and the level at which it has diagnostic value. Having suggested

that the core problem to be addressed is ‘power through’, we can now outline and explain the processes

of persuasion, authority and coercion by means of a unifying causal scheme.

A THREE-PROCESS THEORY OF PERSUASION, AUTHORITYAND COERCION

The proposed causal scheme turns the standard theory completely upside down (see Figure 2). It is

power and resource control which are assumed to follow from influence rather than vice versa.

Influence in turn is seen as a reflection of social identity not dependence. The unifying framework here

is self-categorization theory, which provides an explanation of psychological group formation in terms

of self-categorization rather than dependence (Turner, 1985). The proposed scheme advances

8 John C. Turner


Fig

ure

2.

Th

est

and

ard

and

thre

e-p

roce

ssth

eori

eso

fp

ow

er

Explaining power 9


hypotheses which are not only counter-intuitive but paradoxical from the standard perspective. What is

supposed to be ‘strong’ power turns out to be weak and the recipe for success turns out to be a

prescription for failure.

The theory has three aspects. First, it argues that causality runs in the exactly opposite direction

from the standard view, that psychological group formation, understood as the development of a

shared social identity, gives rise to influence between group members, that mutual influence is the

basis of power (the capacity to persuade and/or control others to carry out one’s will) and that it is

power which enables people to gain and control resources. In the standard theory influence depends

upon power. In this theory power and control of resources depend upon group influence. Second, it

explains that the link between influence and power needs to be understood in terms of three basic and

interrelated processes, influence itself (here meaning persuasion), authority and coercion. The idea

that influence is the basis of power is both a general proposition and a more specific one with varying

meanings depending on the specific process of power being discussed. Third, the theory distinguishes

between the different meanings of power in terms of these three processes.

Only the general scheme can be put forward in the space available. It is worth noting at the outset,

however, that the idea that psychological group formation reflects the development of social identity

and not dependence for resources is a key proposition of self-categorization theory which has much

evidence in its favour. It is now widely accepted that shared social identity plays a crucial role in

psychological group formation and in group processes of influence, cooperation and cohesion (Turner,

1999). This does not mean, however, that group formation is purely subjective and unconstrained by

social reality. People become a psychological group as they shift to defining themselves as members of

a social category rather than in terms of personal identities as a function of their goals, expectations,

background theories, knowledge and ideologies and the social comparisons they make with others

within the specific social situation (see Turner & Bourhis, 1996; Turner & Onorato, 1999; Turner,

Oakes, Haslam, & McGarty, 1994, for discussion of how group identity forms). Also, the theory

proposes that self-categories exist at many different levels of inclusiveness. Thus a point which applies

to the self at one level can also be applied to higher and lower levels. The same theoretical points can

be made about a leader and a group, a subgroup and an organization, and an organization and society.

People influence and control others through persuasion, authority and coercion. These processes are

distinct but can all be defined in relation to group identity and the influence process which flows from

it. Persuasion is the process of influence outlined in self-categorization theory (Turner, 1987). It is

explained as arising from the collective attempt by a group to develop a consensual response to some

stimulus situation. Neither group formation nor influence is based upon resource dependence. It is

assumed that where group members agree about some stimulus their judgement is validated by the

group consensus and becomes normative. Thus a judgement is assumed to be informational, to provide

evidence about reality, precisely to the degree that it has ingroup consensual support. Where members

disagree, they become subjectively uncertain, since similar perceivers confronting the same stimulus

ought and expect to agree. To reduce uncertainty, which is cognitively inconsistent and impedes

confident action, they must either differentiate the stimulus situation (‘we are talking about different

things’), categorize themselves as different (‘our goals are different’) or engage in mutual persuasion

to reach agreement (‘if we approach it in this way we can see that our views are actually compatible’).

Since the perceived validity of any judgement is a function of the degree to which it embodies an

ingroup consensus or norm, it also follows that where group members seek to influence each other they

will tend to be more persuasive to the degree that they are perceived as relatively prototypical of the

emerging consensus (Turner, 1987). Moreover, the relative prototypicality of a member will vary with

group identity which in turn varies with the social comparative context within which the group is

defined (Turner et al., 1994). Both the persuasive power of the group and of individuals within it are a

function of group identity and consensus.

10 John C. Turner


This idea that relative prototypicality is the basis of individual differences in relative influence or

leadership (Turner, Reynolds, Haslam, & Veenstra, in press) has proved extremely important. It was

developed to explain group polarization (Turner, 1987), showing how extremists gained in influence

over moderates as the group defined itself in contrast to an outgroup (e.g. McGarty, Turner, Hogg,

David, & Wetherell, 1992), has been used to explain how leadership emerges in crowds where there is

no organization (Reicher, 1987), how minorities can influence majorities as they gain in relative

prototypicality in certain social contexts (David & Turner, 1999) and how political leaders become

persuasive by using rhetoric to shape the degree to which they are perceived as prototypical by an

audience of a common identity (Reicher & Hopkins, 1996a, 1996b). More recently, the relative

prototypicality hypothesis has generated much research in the area of organizational leadership (see

Haslam, 2004, for an overview).

This is a very brief summary of the whole analysis, which is of much importance to power. Also

highly relevant is the self-categorization work on group consensualization, schisms and changes in the

social identity of crowds as a function of the actions of the police (e.g. Haslam, Turner, Oakes,

McGarty, & Reynolds, 1998; Sani & Reicher, 1998; Stott & Reicher, 1998). In essence, persuasion and

leadership flow from a shared social identity and changes in the relative prototypicality of individuals,

groups and roles explain how power can be won or lost independently of the resources that they

control.

Where persuasion is the power to get people to believe that certain things are correct, authority is

the power to control ingroup members because they are persuaded that it is right for a certain person to

control them in certain matters. Thus legitimate authority is a product of influence and the formation of

norms within the group. Authority is based on ingroup norms that a person, role or group has the right

to prescribe appropriate beliefs, attitudes or behaviour in certain areas. The group agrees (by custom,

experience, or formal decision) that they ought to follow a specific person or position (that it has the

role of leadership or authority) to express their collective will and short-circuit what might be futile,

divisive and time-consuming arguments about what is the right course of action. It is assumed that an

individual can gain authority from their consistent and successful leadership where he or she persuades

the group of the validity of their point of view over a series of tasks and acceptance of that leadership

becomes normative, or that authority can be conferred directly by formal agreement, custom or the

norms inherent in group identity.

A person’s authority varies flexibly with the identity, norms and goals of the group, its beliefs and the

situation in which it finds itself, in the same way that conformity to any ingroup norm varies. Given an

established authority, all the factors which enhance conformity to the group enhance acceptance of

authority. Given influence within the group, any factor which makes an individual more influential than

another will tend over time to confer authority (including their perceived relative prototypicality, Sunshine

& Tyler, 2003; Weber, Mummendey, & Waldzus, 2002). Authority is not direct persuasion but groups

confer authority in order to get things right. Thus acceptance of leader authority carries a presumption that

the leader is likely to be right about the matter in hand and can lead to validation and internalization of the

leader’s view under certain conditions. Authorities can have the direct right to prescribe private beliefs but

this will depend on the nature of the group and the specific authority conferred. Because the power of

authority flows from its designation as expressing the collective will of the group, obedience can produce

pride and a sense of empowerment when one’s collective identity is salient.

This analysis is in line with French and Raven’s (1959) concept of legitimate power and Tyler’s

(2001) relational model of authority. It agrees with the former that authority is based on social

norms and values and with the latter that acceptance of authority is not a social exchange or

dependence process (although it does assume that authority serves collective self-interest, that it

must get things ‘right’ to be able to lead effectively). Tyler’s model stresses the importance of

procedural fairness for encouraging identification (respect and pride) with authorities, enhancing

Explaining power 11


perceived legitimacy and increasing compliance. This seems fully compatible with the present

analysis. An advantage of the present framework is that it points to a whole range of other factors

relevant to identification with authorities and acceptance of their control as an ingroup norm (e.g.

the ideology and goals of group members, the social comparative context, their history of success or

failure for the group, the degree to which the authorities are perceived as more or less prototypical of

the relevant identity).

Coercion is authority in a dark mirror. It is defined here as the attempt to control a target against

their will and self-interest through the deployment of human and material resources to constrain and

manipulate their behaviour. This process is as close as one gets to the standard theory but is not the

standard theory. Coercion here is not French and Raven’s (1959) reward and coercive power. It is not

influence based on resource control, but a conflictual attempt at control given that one cannot

influence. It is defined more by the conflictual relationship than the deployment of resources.

Authorities commonly deploy resources to influence behaviour, which is accepted as legitimate by

group members because of the nature of the source. The nature of the power process defines the

meaning and value of the outcomes, not vice versa. Being rewarded by an institution one admires is a

positive experience just as being punished by a legitimate authority can engender shame and guilt.

Being coerced by rewards and costs, however, can easily lead to their subjective re-evaluation as

‘bribes’ and ‘insults’ or ‘the glory of a martyr’s death’ by the target so as to frustrate completely the

intention of the source. What is positive and what is negative cannot be defined outside of the social

and psychological nature of the power relationship.

Further, in coercion, deploying resources is at the end of the influence-power sequence. Coercion

depends on influence and authority in that the source must have influence and authority over those

willing to be its coercive agents (Goldhagen’s, 1997, ‘willing executioners’). A government which

directs the police or armed forces to exert force against a target must have influence and authority over

the police or armed forces. History is full of instances where coercion ceases because of loss of

influence and authority over coercive agents by the directing source (e.g. in 1943 in occupied

Denmark, the Nazi authorities’ attempt to extend their ‘final solution’ to Danish Jews was a resounding

failure as countless Danish citizens refused to comply with the round-up but instead helped the great

majority to flee to neutral Sweden in what Kershaw, 2000, describes as ‘the most remarkable rescue

action of the war’, p. 604).

The resources being deployed are themselves nearly always people, organized, with specific

expertise, supported by infrastructure and ‘materiel’, trained to use specific social procedures and

cultural technologies (from the police to mass communications to universities). The actual deployment

of resources is only misleadingly described as a process of dependence, as if the target willingly tries to

please those that can satisfy his or her needs. Coercion being against the will of the target is an inherently

conflictual process, to a mild or more extreme degree. Deploying resources here is much more like

stopping, physically preventing, attacking, compelling the target, as in a social conflict, than like offering

a reasonable bargain. Coercion tends to be confused with legitimate resource control (Bass, 1990). The

control of resources in coercion depends on power ‘through’ (the coercive agents) but in relation to the

target is actually close to power in the sense of ‘impact on’. It affects the target, sometimes in the way

intended, but it is not getting the target to act as a voluntary extension of one’s will.

In fact, it does the very opposite. Coercion, being an attempt to restrict the freedom of the target,

makes the target aware of their difference from and disagreement with the source, it leads to

disidentification, increased social distance and rejection of the source and private rejection of the

position being advocated and hence reduces any residual influence and authority of the source.2 It

2The syndrome of effects hypothesized to follow from coercion in this section are an amalgamation of theory and data from arange of sources.

12 John C. Turner


requires surveillance, which in turn enhances the target’s perceived loss of freedom and fosters

mistrust (Kramer, 1999). It is a divisive, destabilizing and counter-productive means of control. The

more coercion is used the more it must be used, since it undermines influence and authority and leads

to attitude change away from the source at the same time as it provokes resistance and reactance to the

loss of freedom (Taylor, 2000). The more it is used the more it brings into being a countervailing

source of power as the targets develop a collective identity defined by their rejection of coercion and

the goal of defeating the coercive agents who threaten their freedom. The rewards and costs which

encouraged obedience at the outset quickly lose their sting as coercion is persisted in and increased.

Freedom from coercion becomes more important and whatever advances one’s freedom is redefined

positively, even when before it was a cost. The ineffectiveness of such a strategy should be self-

evident. What chance compliance when the target is being motivated to go the other way, evade

surveillance and undermine the power of the coercive source at every opportunity? At the extreme,

coercion threatens the power of the source itself, since it brings into being an enemy dedicated to its

downfall.

Coercion is the power one uses when one does not have power. One resorts to power as ‘impact’

when one does not have power ‘through’ (influence and authority in relation to the target), but even

here one must have influence and authority of some kind (over coercive agents). But the paradox is apt.

Coercion is the weakest and least effective process of power defined as getting the target to act as an

extension of one’s will. That the opposite is frequently thought is because coercive ‘impact’ can be

extreme and dramatic and because it always implies power ‘through’ those that carry it out. Destroying

an enemy is a sign of great power in both senses, but it is also a confession of complete failure in terms

of influencing and controlling the actions of the enemy.

Returning to the general causal scheme of from group to influence to power to resource control,

it can be elaborated in several ways. Group identity unifies and empowers people by giving them a

common self-interest and vantage point. It produces influence and influence enables a group to act

as a unified, coordinated, organized body. Group identity and influence give people the power of

collective action and cooperative endeavour, a power to affect the world and pursue shared goals

much greater than any member wields in isolation. Influence also makes possible authority

whereby a group can function effectively and efficiently through centralized leadership and

control, accepted by the members without the necessity or impossibility of seeking everyone’s

agreement on every relevant matter. Then there is the fact that influence and authority create the

possibility of coercion since only through influence and authority can the human and material

resources be deployed to constrain others against their will. Even where one person confronts just

one other, an attempt at coercion reflects a judgement of relative power, which is always influenced

by people’s judgements about the likely reactions of others, whether they will support or oppose the

actor and target, which in turn is a matter of their identities, norms and values. Finally, there is the

critical point that it is influence, its presence or absence, its effectiveness or not, its intended or

experienced use, that determines whether control will take the route of authority or coercion. The

same commands can be experienced as either legitimate or coercive. Authority can be and

frequently is transformed into coercion and coercion can be and frequently is cloaked in legitimacy

and transformed into authority. What matters is the subjective experience of the target. Does he or

she feel that the source is seeking to persuade on the basis of a common identity, values and

interests, with no ulterior motives, or not? All these are routes whereby influence is the basis of

power.

That such power leads to the creation and control of human and material resources has

been argued in passing. What are called ‘resources’ often on closer inspection turn out to

be people, organized for particular purposes and subject to some social authority. Even material

resources almost never exist independently of the organized bodies of people who use them,

Explaining power 13


the civil, legal, political and economic institutions which give them meaning and specify the

conventions of their use.

One dramatic (but not in fact rare) example of the general point must suffice (Cohen, 2003). In

1947 Ben-Gurion, Chairman of the Jewish Agency in Palestine (the state of Israel in waiting),

expecting attack from Arab states upon declaration of Israeli independence, undertook a major

review of all Jewish military forces and concluded that they were completely inadequate to the

waging of full-scale regular warfare against professional armies. The problems were political,

organizational, attitudinal, technical and material. Against the opposition of many, he set about using

his authority and influence (and a degree of extreme coercion against selected Jewish targets he

defined as undermining his authority and impeding his strategy) to re-organize, train, recruit,

manufacture weapons, equip and re-fashion the Israeli armed forces. He led and he mobilized others

to follow, setting up programmes to transform the situation. His policies expanded Israeli military

forces dramatically. In February 1948 on the eve of Israel’s independence its forces numbered

roughly 16,000; by December 1948 they numbered 92,000. From April 1947 to March 1949 rifles

went from 10,000 to 60,000, submachine guns from 2,000 to 31,000, artillery pieces from 25 (at the

end of 1947) to 500, etc. Israel won its war. Nothing about this expansion was spontaneous or easy. It

came from deliberate policy backed up by the power of the group and the authority it gave Ben-

Gurion. There is no question that the hard military resources came from and were created by the

power of the group and not vice versa.

Established power may appear to depend upon resources, but once social conflict and instability

enter the picture such resources can quickly evaporate or change sides and rising groups create,

attract and develop resources as a function of the influence they exert and the organized activity it

makes possible (Pfeffer, 1977, 1992). As relations of identity, ideology, influence and collective

self-interest change, what were defined and created as, for example, the military resources, the tools,

of one group can change into their new masters, replacing those who supposedly controlled them

(e.g. the New Model Army in the English Civil War, Fraser, 1997, or the Prussian Army in the Great

War, Craig, 1968; see Hanson, 2001, for a discussion of the social and cultural nature of military

power).

The three-process theory turns the standard theory on its head and solves its problems. It is

compatible with the facts of and provides many new insights into social change; it unifies the

informational and group aspects of the influence process in an analysis supported by much research; it

properly acknowledges and specifies the causal role of the psychological group in influence and

power; and it rejects the confusion inherent in the traditional notion of compliance, sharply

differentiating coercive force from the workings of authority whilst showing how they are theoretically

linked as divergent reactions to social identity and influence.

A caveat is that it is not being argued that coercion does not have its place in the armoury of

leadership. Any group facing complex tasks and changing realities is likely to require all three

processes of power to organize its activity and preserve its integrity. The art of leadership is often

to coordinate and balance their use. No leader is likely to persuade every single member of

the group of the rightness of some action; no authority fails to provoke its critics and dissidents.

For the group to function there must be a means of dealing with those who seek to reject or

redefine core values and pursue interests different from the group as whole. Coercion as a highly

selective adjunct to influence and authority can be vital to preserving the very existence of the

group, but as a means of neutralizing and excluding the determinedly recalcitrant, not as an

effective means of power through the group (but see below) and never as a substitute for influence

and authority over the group as a whole. The mistake is to follow the standard theory, to confuse

the resources provided by power with its basis and wield all leadership as a form of unilateral

control.

14 John C. Turner


SOME IMPLICATIONS OF THE THREE-PROCESS THEORY

Social Change and the Basis of Power

The idea that resource control is the basis of power tends to imply that differences in power between

individuals and groups are relatively static and enduring. So long as one controls sufficient resources, it

seems, one has power and those without resources have little option but to submit. It is difficult to see

how power ever changes hands in this view. In practice there are many examples from real life of

relatively rapid gains and losses in power where individuals and groups without initial resources

become more powerful and those with overwhelming resources suddenly lose power.

The three-process theory makes perfect sense in this context since it assumes that power reflects

group identity and that all self-categorizing is relational and dynamic, varying with social comparisons

within and between groups, the specific social context, and the collective goals, values and beliefs of

group members. Thus for example, as intergroup relations change, so that cooperation between groups

is replaced by conflict, then group identity is likely to polarize to emphasize differences from the

conflicting outgroup and more extreme, conflictual members will tend to become more prototypical of

each group than will more moderate members. In consequence, the more extreme members will gain

in influence and authority over moderates. History records endless examples of groups which rapidly

change their leaders in line with changes in group identity as they enter into conflict or conclude a

peace with an enemy and where seemingly secure control of resources suddenly evaporates as the

collective goals, beliefs and even mood of the group change.

The general implication is that power differences in society are constantly shifting and that power

change is as normal as stability. The power of leaders and groups depends on identity, organization and

ideology and these foundations are always being built up or torn down, being developed creatively or

deteriorating in indifference, as a function of partisan interests, collective experience, new tasks and

problems and the endless battle between belief and reality. The power of leaders will rise or fall with any

factor which makes them more or less representative of ingroup identity and authority. Changes in the

collective goals, beliefs, attitudes, circumstances and even mood of the group, in fact any factor which

leads the ingroup to define itself and its collective interests differently, can have implications for which

members will have influence and power irrespective of the resources they control. What matters for

social change is that some subordinate group can develop a distinct identity through which to develop its

own goals, values and beliefs and contest the power of the hitherto dominant group, that it should come

to reject as illegitimate the authority of the social order which subordinates it (Branscombe, Schmitt, &

Harvey, 1999; Simon & Klandermans, 2001; Spears, Jetten, & Doosje, 2001; Tajfel & Turner, 1979).

The three-process theory helps to understand how this happens (where a dominant group, for example,

acts selfishly rather than for the collective good, violates core societal values and beliefs, shows

disrespect, arrogance and coercive brutality). One aspect of this knot of issues is the process of coercion.

The Nature and Effects of Coercion

Coercion is the paradigmatic power process in the standard theory. It evokes an image of control which

cannot be resisted because it is based on overwhelming resources, an image of dependence in extremis.

However, the standard theory gets coercion wrong in reducing it to compliance, which confuses two

different forms of control. Once coercion is seen as a distinct process, as proposed here, we find that it

has quite unique effects. Far from being impossible to resist, a useful if unpleasant form of influence, it

is coercion which is the least effective and most counter-productive form of power (understood as

getting people to do what one wants). It is, I have suggested, the mirror image of authority. It is

Explaining power 15


triggered by a lack of influence and authority and undermines residual influence and authority even

further. The coercion of a target tends to increase social distance from and disidentification with

the source, undermine trust and feelings of control, self-respect and pride, induce private rejection of

the influence attempt and engender resistance and reactance in which the goal of rejecting control

and restoring freedom of action gradually breeds social conflict with the intention of frustrating

control.

Coercion persisted in, without the cloak of legitimacy, tends to produce private attitude change

away from the coercer, reactance, conflict and the emergence of a countervailing force. Ultimately,

coercion creates its own enemy. Surveillance to maintain control like all other strategies aimed at

manipulating and constraining the will of the target likewise frustrates their own purpose (Veenstra

et al., 2004). They create an enemy where there could have been an ally and ensure a waste of

resources to try to get an enemy to do what one is persuading them (through negative influence) that

they should not. How much more efficient to change private attitudes, so that others work willingly,

actively and freely, requiring no burdensome apparatus of external control, to achieve the ends that one

desires? The weakness of coercion is in trying to use it as a form of influence. Selective coercion can

certainly be useful if the aim is to destroy, impede or constrain an enemy, but this is more like power as

impact than power through. The problem arises when people confuse the two on the basis of the

standard theory and produce the very opposite of what they intended. But coercion in any form is a

threat to the power of the coercer as we shall discuss presently.

This analysis of the effects of coercion raises the questions of when coercion is likely to be effective

as control and when it is likely to produce at the same time private rejection and reactance. Coercion is

hypothesized to be more effective (albeit in a limited, short-term fashion) where the response required

is minimally self-defining (ego-involving), the target individualized, the resources deployed are

detached as far as possible from any perceived coercive intent (e.g. presented as ‘objective

consequences’ rather than deliberate human pressure), are positive rather than negative, and cloaked

in whatever legitimacy is plausible. The general rule is where possible to disguise coercion as

legitimate authority and minimize the threat to the target’s perceived freedom.

Conversely, the more that any coercive act is perceived as against one’s own will and interests,

emanating from an opponent rather than an ally or partner, and embodies an intended threat to one’s

freedom of action, the more surely will it produce private rejection and reactance and over time foster

the contrasting collective identity and countervailing beliefs that produce sustained oppositional

power and collective conflict. Since this theoretical conception of coercion has never been defined in

any previous theory of power these hypotheses await systematic testing (but see Kramer, 1999;

Reicher & Levine, 1994a, 1994b; Veenstra et al., 2004). It will be important to test them in a temporal

framework, showing that repeated coercion experienced as such does gradually produce an opposi-

tional stance and identity in the target which erodes the power of the leader and that paradoxically the

authority of the leader could disappear more rapidly the more resources and hence the more perceived

coercive threat he or she displays. An analogy to these ideas is provided by historical studies

suggesting that the deployment of great military power freed from all civil and political restraint

(Cohen, 2003; Craig, 1968) or not based on the civic spirit and loyalty of the soldiers (Hanson, 2001)

has often been the speediest route to military disaster as political support has disappeared, cohesion

and discipline evaporated and enemies multiplied.

Prejudice

The present analysis proposes a novel explanation of the link between power and prejudice. It suggests

that prejudice can be seen as a strategy for maintaining and enhancing the power of a dominant group

16 John C. Turner


in society whilst it undertakes coercion of some other group. Where one coerces a target, I have

suggested, the result is rejection and resistance by the target, one’s influence and authority over the

target is directly undermined. If the experience of a specific target becomes a collective experience of

one’s group or society, then coercion will directly undermine one’s power over the group or society.

But it follows from self-categorization theory that any human being can identify with another and

experience their emotions collectively on any number of bases and any feelings of empathy, sympathy,

of shared social identity, with the target by the group will therefore undermine the authority of the

coercive leadership over the group which is the source of its power.

It is obvious that where coercion is exercised against any dissident subgroup by coercive agents

who are members of the same society or subscribe to belief in a common humanity, there is a real

danger that the agents will identify with their victims unless this outcome is prevented. Coercion can

therefore only be exercised by a dominant group without undermining its own power by negatively

categorizing and stereotyping the target as different from the group as a whole, to break down and

prevent any identification with the former by the latter. This is especially necessary where the target is

a dissident member or subgroup of the very group which is the source of one’s power. It is most

important that those carrying out the coercion feel no identification or they may be transformed from

agents to opponents of one’s power.

This strategy can be taken a step further. If one can categorize the target as a threat, danger, enemy to

the group as a whole, creating an implicit or explicit intergroup conflict, and involve other group

members psychologically in the coercion of the target, then, far from undermining one’s authority,

coercion may work to enhance one’s legitimacy and power, appearing as if for the collective good. At the

extreme, the strategy of demonizing the target to legitimize its coercion can become a strategy of creating

and achieving power over the group by inventing or finding a common enemy that redefines the group

identity to place one at its core and one’s opponents beyond the pale (thus Hitler’s invocation of ‘race’

and the Jewish ‘threat’ to replace the class ideology of his social democratic and Marxist opponents).

Far from being merely an expression of power, therefore, prejudice can provide and maintain power.

This is related to but different from the idea of prejudice as a rationalizing justification of power

differences or as scapegoating to avoid blame or that creating intergroup conflict necessarily enhances

loyalty to leaders. This is prejudice to avoid the destabilizing effects of coercion or to gain power

through redefining group identity. This analysis is in the intergroup conflict of interest tradition in that it

takes for granted that a dominant group which has power to pursue its interests in society as a legitimate

authority will nevertheless to some degree always face opponents from subordinate groups who reject

its policies and which it must deal with through coercion. The prejudice does not cause the coercion,

which is undertaken for reasons of political and economic self-interest (slavery, for example, was a

highly profitable institution for the slave-owners). The prejudice is used by the dominant group to deal

with the political dangers to its authority which would arise in society from those who would be

opposed to its coercion but for the prejudice. Thus it is the coercion which makes the prejudice

necessary and politically the prejudice is directed not just against its victims but also against any of the

dominant group’s own supporters whose loyalty might be less than rock solid. A power analysis of

prejudice requires that it be put in the context of a multi-group social system in which there is a constant

struggle between groups to define society in such a way as to strengthen their own power and position.

The Evils of Power

Power has had a bad press. It is believed to corrupt, lead to abuse, stereotyping and prejudice (Keltner

et al., 2003; Lee-Chai & Bargh, 2001). These notions flow from the confounding of power with the

dominance/submission relationship implicit in dependence and the failure to distinguish coercion

Explaining power 17


from other processes in the standard theory. I suggest that power as persuasion or authority is neither

oppressive nor maladaptive nor inherently detrimental to social cohesion. The very opposite in fact

since the coordination and centralization of group action through legitimate authority empowers group

members to achieve their goals. For a group to pursue its collective will there must be a power structure

through which group identity and goals are realized. The issue is what kind of power process is at

work? Legitimate authority should not reduce members’ feelings of power since it expresses a

collective identity, not a perceived hierarchical difference in which members are subordinated.

Authority can degenerate into coercion, just as coercion can be transformed into authority, and it is

important to study how these changes occur. But it is also important to recognize that the absence or

weaknesses of power structures can be highly frustrating and detrimental to social cohesion and

effective social organization (Pfeffer, 1992; Reicher & Haslam, 2004).

The three-process theory provides a useful analysis for understanding how authority can degenerate

and lead to abuse, since authority provides the opportunity and temptation for reliance on coercion and

at the same time may under certain conditions encourage those in authority to differentiate themselves

from followers and develop partisan, separate interests. It is easy to see how the nature of coercion is

likely to brutalize the authority that wields it and produce a cycle of conflict and even violence in

which mutual hatred, delegitimation and self-justification thrive (Bar-Tal, 1990). It is also easy to see

how a dominant group which has developed interests which conflict with those of subordinate groups

and society as a whole will engage in deception, manipulation, coercion and terror to maintain its

position since by definition it cannot appeal to the common good other than dishonestly. But if one

considers whether there is anything psychologically inevitable about the corrupting effects of power

on those that wield it, some caution is in order. The theories which imply such generic psychological

effects all adopt, behind the supposed variety of influence, one narrow social vision of the power

relationship, that of power as dominance and submission based on dependence (e.g. Keltner et al.,

2003).

But power cannot be reduced to one such process; it takes a variety of forms. Power is an emergent

property of human social relationships, not something that stands outside of them, and power

relations can take as many concrete and nuanced forms as the social relationships they express. If one

looks at the evidence for the supposed generic effects of holding power we find that these effects are

either restricted to just one kind of power process or disappear as soon as the values, beliefs, goals of

the powerful are varied and/or the nature of the social relationship is varied (Depret & Fiske, 1999;

Kipnis, 1972; Lee-Chai, Chen, & Chartrand, 2001; Overbeck & Park, 2001; Reynolds, Oakes,

Haslam, Nolan, & Dolnik, 2000; Rind & Kipnis, 1999; Stott & Drury, 2004; cf. Ellemers, van

Rijswik, Bruins, & de Gilder, 1998). Kipnis (1972), for example, defines ‘power’ as something like

coercion in our terms in contrast to a supposed ‘no-power’ condition of persuasion and authority. He

actually shows therefore that the effects of holding power vary with the type of power process (as do

Rind & Kipnis, 1999).

How people react to holding power is not just a function of some abstract relation of dominance, but

of the identity, beliefs, values and theories of the people involved and the political and social content of

their relationship. Leadership can be about dedication, responsibility, service and self-sacrifice as well

as privilege and rewards, working for the collective good, not greed (e.g. Lee-Chai et al., 2001;

Overbeck & Park, 2001). The beliefs, theories and values of the group define what is legitimate and

what unjust, what is appropriate organization and procedure, what is proper accountability, how

leader-follower relations should be organized and conducted, what should be tolerated and what

should be defied. Thus the issue of corruption, of when it is perceived, of what makes it acceptable or

unacceptable, whether it can or even should be prevented, has a social, political, historical and

ideological dimension, which is in fact always at work in the experimental demonstration of

supposedly generic effects.

18 John C. Turner


A strength of the three-process theory in this respect is its focus on group identity, organization and

ideology as bases of power, since the emergence of immorality and abuse in its exercise cannot be

divorced from these things. From a social psychological perspective, no power is absolute, all power is

constrained, both what can be done and what should not be done reflect the substantive beliefs and

values of people with specific identities in a definite place and time. These brief remarks hardly do

justice to the issue of the evils of power, but they do illustrate the rich possibilities of a fresh approach

to power in considering a question of such import.

CONCLUSION

The standard theory of power tends to conflate all the varieties of influence into one relationship of

dependence for resources. Power is provided by control of resources and those without such resources

are in the power of those who have them. In this view power is like some abstract commodity which

exists in principle outside of social relationships and which can be used by the lucky few to impose

their will on such relationships. I have tried to suggest that this is a reification of what is actually going

on. Power is an emergent property of specific social and psychological relations between people and

these relations shape the form it takes. It emerges from group formation, social organization and the

shared beliefs, theories and values (the culture, ideologies, etc.) which shape social and personal

identity and perceived self-interest. These are the foundations of persuasion, authority and coercion,

the processes by which people are able to get others to carry out their will and have a collective impact

on the world. An implication is that power relations always have a definite social, relational and

ideological content which directly affects how power is gained, lost and used. Leaders gain power not

by possessing resources in the abstract but by standing for, representing, believing, working for

something, or being perceived to do so, with which at least some others concur. Power is never

absolute but always socially constrained and conferred. The social psychological study of power

therefore must put it into the context of the group and societal relationships from which it emerges and

the theories which deal with such relations.

ACKNOWLEDGEMENTS

I would very much like to thank Kate Reynolds, Kris Veenstra, Alex Haslam, Steve Reicher and

Natalie Taylor for their help in writing this article and the Australian Research Council for funding

support (DP0342645).

REFERENCES

Bar-Tal, D. (1990). Causes and consequences of delegitimization: Models of conflict and ethnocentrism, Journalof Social Issues, 46, 65–81.

Bass, B. M. (1990). Bass & Stogdill’s Handbook of leadership: Theory, research, and managerial applications(3rd ed., pp. 225–273). NY: The Free Press.

Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving pervasive discrimination among AfricanAmericans: Implications for group identification and well-being. Journal of Personality & Social Psychology,77, 135–149.

Explaining power 19


Brehm, S. S., & Brehm, J. W. (1981). Psychological reactance: A theory of freedom and control. NY: AcademicPress.

Cohen, E. A. (2003). Supreme command: Soldiers, statesmen and leadership in wartime. London: Simon &Schuster.

Craig, G. A. (1968). The politics of the Prussian Army 1640–1945. London: Oxford University Press.David, B., & Turner, J. C. (1996). Studies in self-categorization and minority conversion: Is being a member of the

outgroup an advantage? British Journal of Social Psychology, 35, 179–199.David, B., & Turner, J. C. (1999). Studies in self-categorization and minority conversion: The ingroup minority in

intragroup and intergroup contexts. British Journal of Social Psychology, 38, 115–134.Depret, E. (1995). Vicarious feelings of personal control and the social categorization of powerful others.

Unpublished paper. University of Grenoble.Depret, E., & Fiske, S. T. (1999). Perceiving the powerful: Intriguing individuals versus threatening groups.Journal of Experimental Social Psychology, 35, 461–480.

Deutsch, M., & Gerard, H. B. (1955). A study of normative and informational social influences upon individualjudgment. Journal of Abnormal & Social Psychology, 51, 629–636.

Ellemers, N., van Rijswik, W., Bruins, J., & de Gilder, D. (1998). Group commitment as a moderator ofattributional and behavioural responses to power use. European Journal of Social Psychology, 28, 555–573.

Festinger, L. (1950). Informal social communication. Psychological Review, 57, 271–282.Festinger, L. (1953). An analysis of compliant behaviour. In M. Sherif, & M. O. Wilson (Eds.), Group relations atthe crossroads (pp. 232–256). NY: Harper & Bros.

Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140.Fiske, S. T. (1993). Controlling other people: The impact of power on stereotyping. American Psychologist, 48,

621–628.Fiske, S. T., & Depret, E. (1996). Control, interdependence and power: Understanding social cognition in its social

context. European Review of Social Psychology, 7, 31–61.Fraser, A. (1997). Cromwell: Our chief of men. London: Arrow Books.French, J. R. P., & Raven, B. H. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in social power

(pp. 150–167). Ann Arbor, MI: Institute of Social Research.French, J. R. P., Morrison, H. W., & Levinger, G. (1960). Coercive power and forces affecting conformity. Journalof Abnormal & Social Psychology, 61, 93–101.

Goldhagen, D. J. (1997). Hitler’s willing executioners: Ordinary Germans and the Holocaust. London: Abacus.Hanson, V. D. (2001). Why the West has won: Carnage and culture from Salamis to Vietnam. London: Faber &

Faber.Haslam, S. A. (2004). Psychology in organizations: The social identity approach (2nd ed.). London: Sage.Haslam, S. A., Turner, J. C., Oakes, P. J., McGarty, C., & Reynolds, K. J. (1998). The group as a basis for emergent

stereotype consensus. European Review of Social Psychology, 8, 203–239.Jost, J. T., & Major, B. (2001). The psychology of legitimacy. NY: Cambridge University Press.Kelman, H. C. (1958). Compliance, identification, and internalization: Three processes of attitude change. Journalof Conflict Resolution, 2, 51–60.

Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and inhibition. Psychological Review,110, 265–284.

Kershaw, I. (2000). Hitler 1936–45: Nemesis. London: Penguin.Kipnis, D. (1972). Does power corrupt? Journal of Personality & Social Psychology, 24, 33–41.Kramer, R. M. (1999). Trust and distrust in organizations. Annual Review of Psychology, 50, 569–598.Lee-Chai, A. Y., & Bargh, J. A. (2001). The use and abuse of power. Philadelphia, PA: Psychology Press.Lee-Chai, A. Y., Chen, S., & Chartrand, T. L. (2001). From Moses to Marcos: Individual differences in the use and

abuse of power. In A. Y. Lee-Chai, & J. A. Bargh (Eds.), The use and abuse of power (pp. 57–74). Philadelphia,PA: Psychology Press.

Leet-Pellegrini, H., & Rubin, J. Z. (1974). The effects of six bases of power upon compliance, identification, andinternalization. Bulletin of the Psychonomic Society, 3, 68–70.

McGarty, C., Turner, J. C., Hogg, M. A., David, B., & Wetherell, M. S. (1990). Group polarization as conformityto the prototypical group member. British Journal of Social Psychology, 31, 1–20.

Moscovici, S. (1976). Social influence and social change. London: Academic Press.Moscovici, S. (1980). Towards a theory of conversion behaviour. In L. Berkowitz (Ed.), Advances in experimentalsocial psychology (Vol. 13, pp. 209–239). NY: Academic Press.

Overbeck, J. R., & Park, B. (2001). When power does not corrupt: Superior individuation processes amongpowerful perceivers. Journal of Personality & Social Psychology, 81, 549–565.

20 John C. Turner


Pfeffer, J. (1977). Power and resource allocation in organizations. In B. M. Staw, & G. R. Salancik (Eds.), Newdirections in organizational behaviour (pp. 235–265). Chicago, Ill: St. Clair Press.

Pfeffer, J. (1992). Managing with power. Boston, Mass.: Harvard Business School Press.Raven, B. H. (2001). Power/interaction and interpersonal influence: Experimental investigations and case studies.

In A. Y. Lee-Chai, & J. A. Bargh (Eds.), The use and abuse of power (pp. 217–240). Philadelphia, PA:Psychology Press.

Raven, B. H., & French, J. R. P. (1958). Legitimate power, coercive power, and observability in social influence.Sociometry, 21, 83–97.

Reicher, S. (1987). Crowd behaviour as social action. In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher,& M. S. Wetherell, Rediscovering the Social Group: A self-categorization theory (pp. 171–202). Oxford:Blackwell.

Reicher, S., & Haslam, S. A. (2004). Rethinking the psychology of tyranny. University of St. Andrews.In submission.

Reicher, S., & Hopkins, N. (1996a). Constructing categories and mobilising masses: An Analysis of Thatcher’sand Kinnock’s speeches on the British miner’s strike 1984–5. European Journal of Social Psychology, 26, 353–371.

Reicher, S., & Hopkins, N. (1996b). Seeking influence through characterising self-categories: An analysis of anti-abortionist rhetoric. British Journal of Social Psychology, 35, 297–312.

Reicher, S., & Levine, M. (1994a). Deindividuation, power relations between groups and the expression ofsocial identity: The effects of visibility to the out-group. British Journal of Social Psychology, 33, 145–163.

Reicher, S., & Levine, M. (1994b). On the consequences of deindividuation manipulations for the strategiccommunication of self: Identifiability and the presentation of social identity. European Journal of SocialPsychology, 24, 511–524.

Reynolds, K. J., Oakes, P. J., Haslam, S. A., Nolan, M., & Dolnik, L. (2000). Responses to powerlessness:Stereotyping as an instrument of social conflict. Group Dynamics: Theory, Research and Practice, 4, 275–290.

Rind, B., & Kipnis, D. (1999). Changes in self-perceptions as a result of successfully persuading others. Journal ofSocial Issues, 55, 141–156.

Roberts, A. (2003). Hitler and Churchill: Secrets of leadership. London: Weidenfeld & Nicolson.Sachdev, I., & Bourhis, R. Y. (1985). Social categorization and power differentials in group relations. EuropeanJournal of Social Psychology, 15, 415–434.

Sani, F., & Reicher, S. (1998). When consensus fails: An analysis of the schism within the Italian CommunistParty. European Journal of Social Psychology, 28, 623–645.

Sidanius, J., & Pratto, F. (1999). Social dominance. Cambridge: Cambridge University Press.Simon, B., & Klandermans, B. (2001). Politicized collective identity. American Psychologist, 56, 319–331.Spears, R., Jetten, J., & Doosje, B. (2001). The (il)legitimacy of ingroup bias: From social reality to social

resistance. In J. T. Jost, & B. Major (Eds.), The psychology of legitimacy (pp. 332–362). NY: CambridgeUniversity Press.

Stott, C., & Drury, J. (2004). The importance of social structure and social interaction in stereotype consensus andcontent: Is the whole greater than the sum of its parts? European Journal of Social Psychology, 34, 11–23.

Stott, C., & Reicher, S. D. (1998). Crowd action as intergroup process: Introducing the police perspective.European Journal of Social Psychology, 28, 509–530.

Sunshine, J., & Tyler, T. R. (2003). Moral solidarity, identification with the community, and the importance ofprocedural justice: The police as prototypical representatives of a group’s moral values. Social PsychologyQuarterly, 66, 153–165.

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, California: Brooks/Cole.

Taylor, N. (2000). When disagreement strikes: Incorporating compliance, conversion, resistance and reactanceinto a model of social response. PhD thesis. Australian National University.

Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive theory of group behaviour.Advances in Group Processes, 2, 77–122.

Turner, J. C. (1987). The analysis of social influence. In J. C. Turner, M. A. Hogg, P. J. Oakes, S. D. Reicher, & M. S.Wetherell (Eds.), Rediscovering the social group: A self-categorization theory (pp. 68–88). Oxford: Blackwell.

Turner, J. C. (1991). Social influence. Milton Keynes, UK: 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: Context, commitment, content (pp. 6–34). Oxford, UK& Cambridge, USA: Blackwell.

Explaining power 21


Turner, J. C., & Bourhis, R. Y. (1996). Social identity, interdependence and the social group. In W. P. Robinson(Ed.), Social groups and identities: Developing the legacy of Henri Tajfel (pp. 25–63). Oxford: ButterworthHeinemann.

Turner, J. C., & Haslam, S. A. (2001). Social identity, organizations and leadership. In M. E. Turner (Ed.), Groupsat work: Advances in theory and research (pp. 25–65). Hillsdale, NJ: Erlbaum.

Turner, J. C., & Onorato, R. (1999). Social identity, personality and the self-concept: A self-categorizationperspective. In T. R. Tyler, R. M. Kramer, & O. P. John (Eds.), The psychology of the social self (pp. 11–46).Mahwah, NJ: Lawrence Erlbaum Associates.

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. A. (1994). Self and collective: Cognition and socialcontext. Personality & Social Psychology Bulletin, 20, 454–463.

Turner, J. C., Reynolds, K. J., Haslam, S. A., & Veenstra, K. (in press). Reconceptualizing personality: Producingindividuality by defining the personal self. In T. Postmes, & J. Jetten (Eds.), Individuality and the group:Advances in social identity. London: Sage.

Tyler, T. R. (2001). Social justice. In R. Brown, & S. Gaertner (Eds.), Handbook of social psychology: Intergroupprocesses (Vol. 4, pp. 344–364). Oxford, UK: Blackwell.

Tyler, T. R., & Degoey, P. (1995). Collective restraint in social dilemmas: Procedural justice and socialidentification effects on support for authorities. Journal of Personality & Social Psychology, 69, 482–497.

Veenstra, K., Turner, J. C., Reynolds, K. J., Haslam, S. A., & Burgess, N. M. (2004). Power and social identity:The differential effects of ingroup and outgroup leadership on power and compliance. Australian NationalUniversity. In submission.

Weber, U., Mummendey, A., & Waldzus, S. (2002). Perceived legitimacy of intergroup status differences: Itsprediction by relative ingroup prototypicality. European Journal of Social Psychology, 32, 449–470.

22 John C. Turner


turner on power

Documents

australian national university

australian research council

cambridge university press

shared social identity

social comparative context

experimental social psychology

psychological group formation

human social relationships