Welcome to ccp web site

Gene Sharp, the world's leading writer on non-violent action, uses a theory of power based on a division between rulers and subjects and on the withdrawing of consent as the main avenue for effecting political change. From the point of view of structural approaches to the analysis of society, Sharp's picture leaves out much of the complexity of political life, such as the structures of capitalism, patriarchy and bureaucracy which do not fit well with the ruler-subject picture. As a set of conceptual tool for social activists, however, Sharp's theory of power is far superior to structural approaches.

* The theory of power discussed in this Review Essay underlies most of Gene Sharp's writings (1970, 1973, 1979, 1980, 1985). The most important explicit statements for studying it are Sharp (1973. pp. 7-62 and 1980, pp. 21-67, 309-378) . I thank Ariel Salleh and Val Plumwood for valuable discussions and Ann Aungles, Stewart Russell and two anonymous referees for useful comments on an earlier draft.

1. Introduction

Gene Sharp is the foremost writer in the world today on the subject of non-violent action. His book The Politics of Nonviolent Action (1973) is widely regarded as a classic. Other important works are two collections of essays, Social Power and Political Freedom (1980) and Gandhi as a Political Strategist (1979), and the more recent Making Europe Unconquerable (1985). More works are forthcoming.

Other writers and activists have made important contributions to the theory and practice of non-violent action, especially Gandhi [1]. Sharp's key role has been to systematize the field in two ways. First, he has classified methods of non-violent action and catalogued hundreds of different techniques along with an extensive array ofhistorical examples. This classification has produced conceptual order amongst the cluttered and scattered experiences of and literature on non-violent actions. Second, Sharp has elaborated a theory of power which offers a framework for understanding how non-violent action works.

Sharp's ideas are especially worthy of critical attention because they have been widely adopted by social activists as providing a theoretical underpinning for their own nonviolent actions. Training sessions on nonviolence often include segments on 'power theory', which typically is a simplified version of Sharp's ideas, based either directly on his writings or on secondary accounts of them. Yet compared to the intensive use of his ideas by activists, scholars have devoted little attention to Sharp.

My aim in this paper is to analyse Sharp's theory of power, especially by comparing it to structural approaches to social analysis. It is not my concern here to confront the standard objections to non-violent action, especially that it doesn't or won't work, nor address the argument that the usual reliance on violence by governments or liberation movements is essential. Criticisms of Sharp from defenders of an ultimate reliance on violence often misconceive his theory [2].

Sharp has written that he welcomes critiques. My analysis is done in the spirit of sympathetic criticism.

2. Sharp's Approach

The essence of Sharp's theory of power is quite simple: people in society may be divided into rulers and subjects; the power of rulers derives from consent by the subjects; non-violent action is a process of withdrawing consent and thus is a way to challenge the key modern problems of dictatorship, genocide, war and systems of oppression.

The two key concepts in Sharp's theory of power are, first, the ruler-subject classification and, second, consent. The ruler-subject classification is one that Sharp uses without detailed justification. The 'ruler' includes 'not only chief executives but also ruling groups and all bodies in command of the State structure' (1980, p. 22). Sharp focuses on the state in his analysis. He spells out the various structures involved in the state, especially the state bureaucracy, police and military, all of which 'are under the command of the person or group which occupies the position of "ruler" at the head of the State' (1980, p. 316). All others besides the rulers are the subjects.

Sharp defines political power, which is one type of social power, as 'the totality of means, influences, and pressures - including authority, rewards, and sanctions - available for use to achieve the objectives of the power-holder, especially the institutions of government, the State, and groups opposing either of them' (1980, p. 27). Sharp counterposes his analysis to the common idea that power is a monolithic entity residing in the person or position of a ruler or ruling body. Sharp argues instead that power is pluralistic, residing with a variety of groups and in a diversity of locations, which he calls 'loci of power'. The loci of power provide a countervailing force against the power of the ruler, especially when the loci are numerous and widely distributed throughout society.

Accepting the argument that power is not intrinsic to rulers, then it must come from somewhere else. Sharp gives the following key sources of power: authority, human resources, skills and knowledge, intangible factors, material resources and sanctions (1973, pp. 1112). What is the basis for these sources of power? This is where the second key concept of Sharp's enters in. He says that these sources of the ruler's power 'depend intimately upon the obedience and cooperation of the subjects' (1973, p. 12). This can be called the consent theory of power. Without the consent of the subjects - either their active support or their passive acquiescence the ruler would have little power and little basis for rule.

Power for Sharp is always contingent and precarious, requiring cultivation of cooperation and manipulation of potentially antagonistic loci. His consideration of the sources of power thus leads him to obedience as the key: 'the most important single quality of any government, without which it would not exist, must be the obedience and submission of its subjects. Obedience is at the heart of political power' (1973, p. 16).

The focus on obedience then leads Sharp to ask 'Why do men obey?' He suggests that there is no single answer, but that important are habit, fear of sanctions, moral obligation, self-interest, psychological identification with the ruler, zones of indifference and absence of self-confidence among subjects (1973, pp. 16-24).

Non-violent action constitutes a refusal by subjects to obey. The power of the ruler will collapse if consent is withdrawn in an active way. The 'active' here is vital. The ruler will not be threatened by grumbling, alienation or critical analyses alone. Passivity and submissiveness are of no concern to Sharp; he is interested in activity, challenge and struggle (1973, p. 65), in particular with nonviolent methods of action.

The account here has abbreviated and simplified Sharp's full exposition, but nevertheless highlights key assumptions made by him. His theory of power is only the beginning of his work on non-violent action, which leads him through methods of nonviolent action to the 'dynamics of nonviolent action', which includes laying the groundwork for action, making challenges, building discipline, building support and redistributing power. The theory of power is important because it is the theoretical foundation for Sharp's other work.

3. Structural Approaches

Sharp's approach can be examined and challenged from many different angles. Here I contrast it with a very general approach to social analysis which focuses on social structures or institutions. Structures frequently selected out as significant include capitalism, the state, patriarchy and bureaucracy.

While social structures are inevitably composed out of numerous social interactions, to focus on the structure is to imply that certain types of social interaction are so regular and entrenched that they take on a dynamic of their own. The Marxist analysis of capitalism is probably the best example of this (e.g. Althusser, 1977; Baran & Sweezy, 1968; Mandel, 1976; Marcuse, 1964; Poulantzas, 1978). Founded on private property, the ownership of the means of production by a small minority of people, and a market on which labour power is purchased and exploited, capitalism appears to behave like a self-regulating system. Whatever the intentions of individual capitalists, if they do not extract surplus labour power from their workers and thus compete successfully in the market, they will be driven to bankruptcy.

Much Marxist scholarship has shown that there is much more to capitalism than blind economic mechanism. Vital to the establishment and maintenance of capitalist relationships are struggles between owners and employees, gender and ethnic divisions within the working class, economic intervention from the state to stabilize and protect markets, social intervention from the state to provide services (education, health and welfare) for reproducing the labour power needed by capital, and police and military intervention from the state to control labour revolts.

The resulting complex of economic and political relationships is still usually called capitalism, and again treated as a system with its own dynamic. Thus we can read about the current restructuring of capitalism, the penetration of capitalism into second and third world societies and into more and more facets of everyday life, and the ubiquitous 'crisis' of capitalism.

Structural approaches hold great power in analysing social systems, if the structures which are conceptualized happen to capture key ways of organizing human interactions which tend to reproduce themselves. This is an elementary but important point. In principle, there is nothing to stop the employees at a factory from simply leaving their jobs and setting up production on their own in a different location. In practice, if the 'different location' were someone else's private property, police would be called in to evict the workers and there would be little support from anyone else in the community. Furthermore, the original company typically would find little difficulty in recruiting new workers. Thus, the system of private property and the market in labour would continue as before. As a shorthand, it could be said that capitalist social relations continued to assert themselves .

It took many decades before the strike, a carefully circumscribed withdrawal of labour power, was accepted as legitimate, and it continues to be attacked by employers. Direct challenges to private property, such as squatting and workers' control, are even more difficult to achieve.

The existence of numerous struggles at the borders of what is conceived of as capitalism makes it hard to argue that capitalism is an automatically self-sustaining type of mechanism. Fundamentally involved is the commitment of individuals to the current order. This is where the concept of hegemony enters (Gramsci, 1971). Hegemony refers to the processes by which a given way of organizing social life, in which one class dominates another, becomes accepted as inevitable and desirable by most people. These processes include the mass media, formal education, the family, popular culture, and routines of daily life at work and leisure.

Other concepts of social structure, such as patriarchy and the state, have been elaborated (and disputed) in a fashion similar to capitalism. The power of such concepts is shown when analysing large-scale developments (the law of uneven development of international capitalism; international politics as struggles between states) and also when understanding social struggles (such as conflicts and accommodation between capitalism and patriarchy in labour history).

One of the major dangers in using such concepts is the reification of categories. Capitalism, for example, is frequently presented as if it operates and evolves independently of the people whose interactions make it up.

The use of a structural analysis does not commit one to a particular method of political action. Historically, Marxist analysis has been linked to vanguard parties whose aim is to capture state power in the name of the working class, and for whom tools such as violence and the state are neutral. But others using a Marxist analysis favour more populist methods, involving themselves in mass struggles or working with the 'new social movements' such as the environmental, feminist and peace movements.

4. Limitations of Sharp's Approach

Structural approaches provide a useful contrast for examining Sharp's theory of power [3]. Sharp's focus on consent is individualistic and voluntaristic in orientation, as shown by his attention to psychological reasons for obedience. An analysis of social structure provides another way to understand consent (Moore, 1978).

An understanding of the power relationships associated with capitalism would seem essential to developing effective non-violent methods of struggle. While Sharp gives numerous examples of non-violent action by workers - he devotes an entire chapter of The Politics of Nonviolent Action to 23 types of strikes (l973, pp. 257-284) - he gives no examination of capitalism as a system of power, and misses out on insights provided by Marxist analysts.

While in principle an oppressive ruler can be opposed by workers walking off the job, in practice there are many factors to be taken into account in mobilizing them to do so. The workers are likely to be divided along lines of status, skill, wages, gender and ethnicity; the mass media may provide little support or active disinformation; certain workers may have been tied to the regime by dispensation of special favours. being involved in corruption, or compromised by participating in repression of minorities; education in nationalism may make it easy for the ruler to raise the spectre of foreign enemies, external agitators and hurting the national interest.

Furthermore, the 'system', whatever its oppressiveness, may still serve to benefit large groups of people in certain ways. Many members of the working class, while exploited by capitalists, at the same time receive wages sufficient to offer a life seen as better than those of their parents. Capitalism as a social system simultaneously oppresses and benefits those who live in it.

Sharp also gives no analysis of the social system of bureaucracy and how its hierarchy, division of labour and regular procedures serve to mesh everyone - including top bureaucrats - into patterns of behaviour which are hard to escape. Contrary to the usual picture, political struggles do take place within and around bureaucracies (Weinstein, 1979), and since these struggles are almost always non-violent, Sharp's approach may offer some insights. But the ruler-subject dichotomy is of limited value here, since in a typical bureaucracy, nearly everyone has both superiors and subordinates. To be of use, the dynamics of non-violent action would have to be elaborated in light of studies of the dynamics of bureaucracy.

Patriarchy is another system of power which Sharp has not analysed in detail. The social practices by which males dominate over females can hardly be seen as ones simply of ruler and subject. Complex processes are involved, including upbringing, expectations of characteristic behaviour, the gender division of labour, direct discrimination, harassment, rape and other violence, all of which are linked to other systems of power, including economic structures, the state, trade unions, churches and the military. In particular, patriarchy is intertwined with the power structures of the state and the military which are the focus of Sharp's analysis. Non-violent action and the giving or withdrawing of consent by women undoubtedly are important in the maintenance of male domination. But without any analysis of patriarchy as a structured set of social relations which can hardly be 'turned off' by the simple withdrawal of consent, Sharp does not provide the basis for studying this power dynamic.

Another key factor in systems of power is technology. Rather than being neutral tools, technologies can be said to embody social relations (Dickson, 1974). In other words, particular artefacts are easier to use for some purposes and by some social groups than others. For example, nuclear weapons can serve the ends of state elites and perhaps some terrorists, but not the ends of environmentalists or even the police. Small-scale solar energy embodies values of self-reliance and decentralization, whereas fossil fuels are more easily linked to dependence on centralized suppliers.

The practical possibilities for 'withdrawing support' depend in part on the technological infrastructure. If a community can feed and shelter itself without massive outside assistance, it is better able to oppose aggressors using non-violent methods. Technologies for person-to-person communication, such as the telephone and short-wave radio, provide a stronger basis for non-violent resistance than one-directional technologies such as television. An analysis of the social relations of technology, and the social struggles around technology, therefore is vital to a full understanding of how present-day society is maintained and how non-violent challenges to oppression can best proceed. Sharp does not bring in such an analysis.

Another important factor is the knowledge and experience of individuals and groups. People with a tradition of independence and social struggle, and with practical experience of opposing authority, are likely to be in a much better situation to make use of non-violent action. Knowledge and experience of this sort depend on a number of factors, including styles of upbringing, formal education, the prevalence and mode of activity of community groups, and the organization of work. Other power structures enter in here, such as bureaucracy and patriarchy, since they shape the understandings and experiences of those who 'live in them'. Knowledge and experience are taken into account by Sharp in his discussions of the psychology of obedience and loci of power, but he provides no structural analysis of how people come to have the knowledge and experience that they do. Knowledge is not something that resides in books and experts but can be interpreted as a social relationship, as a feature of processes of negotiation and the exercise of power. Governments provide funds to research and develop certain kinds of knowledge; the media select and construct knowledge in certain ways schooling promotes and validates certain things as knowledge. These and other factors affect the potential for non-violent action, but Sharp's categories do not provide a convenient entry point for examining them.

Sharp would quite correctly reply that areas such as bureaucracy, patriarchy and technology are perfectly compatible with his picture. In his listing of 'sources of power', the category 'skills and knowledge' would cover the factor of knowledge and experience noted above, while the category 'material resources' would cover the factor of technology. With a bit of stretching, the factors of bureaucracy and patriarchy might be included in his category of 'human resources'. Sharp's picture can be made to include things raised by structural approaches, but only with some difficulty. The point is that Sharp's picture focuses first and foremost on the ruler-subject dichotomy and on consent and its withdrawal, whereas a detailed analysis of the structures of power can only enter as an afterthought or as a general context for the consent picture.

Like any moderately adaptable political theory, Sharp's theory of power can be extended or adapted to cover facets that initially seem to be left out. Indeed, a careful reading of Sharp's work reveals an awareness of many of the points raised here. Touching on issues in a general way, however, is quite different from integrating them into the core concepts. The adaptability of the theory does not remove its central focus, and it is this focus which shapes how the theory is used and who is likely to use it.

The consent picture works best, as theory, when there is an obvious oppressor. It is not by chance that Sharp regularly refers to Stalinism and Nazism. His examples of challenges to authority largely concern situations which are widely perceived as oppressive by contemporary Western political judgment. In retrospect, and as a result of incessant reinforcement, most people in the West today assume without question that opposition was the only moral stance to be taken against Stalinism and Nazism. Sharp challenges one aspect of this received wisdom in his emphasis on the support for and lack of resistance to the Stalinist and Nazi regimes.

Yet on another level Sharp does not provide a sufficient challenge to the picture of ruler and oppressed (if consenting) populace. To fully understand the phenomenon of Stalinism, it is essential to analyse the mobilization of support and suppression of dissent through the Communist Party, the process of industrialization, the reconstitution of the hierarchical army in the 1918-21 war against the Western attack on the revolution, the social inheritance of Tsarism, and the international political scene. Similar comments apply to Nazism. The point is that Stalinism and Nazism were much more than simply systems of ruler and oppressed, and that a full understanding of 'consent' requires a deep social analysis (e.g. Gouldner, 1977-78).

This point is clearer in the context of present-day struggles, where the judgment of history has not yet become conventional wisdom in school history classes and bipartisan political rhetoric. The meaning of nonviolent action is the result of social struggle rather than following immediately from a simple examination of rulers and subjects. Those such as the Berrigans who have taken non-violent direct action against facilities linked to the capacity for nuclear warfare can be interpreted as acting for humanity against evil rulers who are willing to risk mass killing to defend systems of power. But only a minority of people accept this interpretation; in practice, the civil disobedients to the nuclear war machine are engaged in political practice to convince people that their concerns should be the concerns of others. These activists have found that the dynamic of non-violent action does not automatically click into place to generate greater support. Sharp could only agree; he continually stresses that non-violent action is not guaranteed to succeed. The trouble is that his theory of power does not provide the conceptual tools needed to determine whether direct action against nuclear facilities is a particularly effective way to challenge the current systems of power and the current ideologies which mobilize much of the population to support organized violence as 'defence' against an 'enemy'.

Sharp comes closest to a structural approach in his discussion of loci of power. For example, he describes how the distributed power of the nobility under feudalism constrained the monarch, who in principle had unlimited authority (1980, pp. 33-35). But Sharp does not introduce any concepts convenient for analysing these structures. The major purpose of his examples is to argue for his thesis that constitutions are not sufficient to control rulers, that replacing rulers does not lead to control over rulers, and that devolution and diffusion of power among many groups is necessary to control the ruler's power (1980, p. 47). In short, his discussion of structures is used to support his basic ruler-subject picture. Once established, the structures tend to be dropped out of the picture. It is perhaps significant that when Sharp does discuss structures of power it is usually using historical examples such as feudalism or Fascism rather than examples also quite relevant today such as capitalism or patriarchy.

Even Sharp's discussion of loci of power gives a very simplified picture. Sharp argues that 'In order for effective control over the ruler's power to be possible in the long run, power must be effectively devolved and diffused among various social groups and institutions throughout the society' (1980, p. 47). This ignores the possible supportive relationships between the loci ('various social groups') and dominant social groups, and conflicts between the loci themselves. For example, trade unions arose out of workers' struggles against oppressive working conditions under capitalism, and were only set up in the face of vigorous opposition by capitalists and governments. Therefore, trade unions seem to be a perfect example of loci of power. Yet, once established, many trade unions have been incorporated into the 'system' and act to control the workers, for example in opposing grassroots worker initiatives and wildcat strikes. The existence of hierarchy and bureaucracy in trade union structures belies the image of a straightforward process of devolution of power.

Trade unions, too, have been key agencies for maintaining the gender division of labour, often in the face of the acceptance or preference of employers for women at a lower wage (Walby, 1986). Women's groups in their struggle against discrimination in employment have gained some leverage from state power, for example in the form of equal employment legislation. This seems to be a process of one locus of power, the women's movement, drawing on state power (the 'ruler') to challenge features of another locus of power, namely patriarchal work practices supported by trade unions. A similar analysis could be made of the dual role of other organizations, such as political parties or environmental lobbies, which act both to gain concessions and coopt radical ferment. The message from such examples is that Sharp's idea of strengthening the loci of power is not nearly as straightforward as it might seem, while the complexities are hard to grasp using Sharp's conceptual framework.

Sharp argues that the use of non-violent action tends to diffuse power: 'Changes achieved by nonviolent action are therefore likely to be more lasting' (1980, p. 62). Sharp's lack of structural analysis makes it difficult to say anything more than this vague claim. The practical results of non-violent action depend on the political context, and a detailed analysis needs to be made to determine the role of nonviolent action (e.g. Zielonka, 1986).

For example, the Iranian Revolution in 1978-1979 was won largely through the mass use of non-violent methods mobilized through the decentralized loci of the bazaars. Furthermore, in the early stages of the revolution there were some important social initiatives, for example towards equality for women (Albert, 1980). Yet the revolution quickly turned into a system of centralized repression. Factors involved in this transformation include the availability of the state bureaucracy and military forces from the Shah's regime, the hierarchical structure and ideology of Shi-ite Islam, and the global political configuration. The point here is that a simple analysis of the 'dynamics of nonviolent action' leaves out much of the social complexity needed to understand the Iranian events. Structural analysis has much to offer in understanding the process of revolution (Skocpol, 1979, 1982).

5. Strengths of Sharp's Approach

From the point of view of structural approaches to social analysis, Sharp's theory of power is much too simple to capture the full dynamics of society, if it is not misconceived entirely. But this critique has been made using a tacit assumption, associated with structural approaches, about what a theory of power is supposed to achieve. To unearth this assumption, it is useful to start with a basic question: what is the point of having a theory of power in the first place?

The usual answer to this question in social science would appeal to some unexamined notion of achieving a better 'understanding' of social reality. But, to pursue the point, what is the purpose of better understanding? Whose ends does this understanding serve?

If the aim to advance the careers of intellectuals who stand by the side observing society but preferring to avoid interaction with it, then a complex, erudite theory serves admirably. On the other hand, if the aim is to provide some insights which can be used by activists, then a simple, straightforward, easy-to-apply theory is far superior, so long as it grasps certain basic insights. By this criterion, Sharp's theory is highly successful.

Sharp explicitly states that he aims to be readable (1980, p. xii). While his jargon-free accessibility is important, however, far more so is the ease with which his approach can be applied by activists. Sharp's picture is essentially voluntarist: people, by deciding to withdraw consent, can topple even the most repressive dictatorship. Sharp provides not only a host of examples of non-violent action, but also describes a simple dynamics which shows how seeming weakness - nonviolence - can lead to increasing support.

Sharp has been taken up as the patron theorist of non-violent action around the world. His ideas about power are regularly presented in nonviolent action training sessions, his examples of non-violent action are endlessly re-used in talks and leaflets, and his authority is routinely invoked in support of non-violence. Arguably, Sharp has a higher profile among grassroots social activists than any other living political theorist. (The major influential figures of Gandhi, Mao and Marx are dead.)

Ironically, while Sharp's analysis is most applicable to authoritarian regimes which more closely approximate the ruler-subject picture, his ideas have gained the greatest following in liberal democracies where the complexity of power structures limits the relevance of his theory.

In being taken up by activists, Sharp's ideas are often communicated and applied in a highly simplified form. Simplification and transformation (distortion) are inevitable in the translation from theory to practice. This also applies to structural analysis; one result has been 'vulgar Marxism'. This is one reason why it is important to examine the core ideas in a theory rather than be distracted by complex elaborations. The strength of Sharp's approach is that his core ideas are ideally suited for fostering non-violent action, whereas the core ideas in structural approaches are better suited for analysis than action.

Compared to his enormous currency 'in the field', Sharp has had relatively little influence in policy or scholarly circles. For most policy makers, who deal in the nitty-gritty of practical politicking and its assumptions of top-down decision-making and the ultimate reliance on violence, Sharp's commitment to non-violence and diffusing power is far too radical and hence is dismissed as impractical or utopian. Furthermore, training populations in methods of non-violent action would make the task of 'governing' society - that is, maintaining the reality and legitimacy of inequalities in power, wealth and status - immensely more difficult, and would jeopardise the positions of the policy makers themselves.

The scholarly neglect of Sharp's work [4] is more complex. A possible (and highly unflattering) explanation is that his power theory is too simplistic to attract the attention of political scientists while his studies of nonviolent action, which can be attacked as based on taking historical examples out of context in order to prove a point, are not convincing historical scholarship. Whatever the strength of such charges in relation to Sharp, shortcomings such as these are rife in political and historical work. For example, the theory of nuclear deterrence is based on several untenable assumptions, such as the rationality and unity of national actors. Yet it is taken extremely seriously by numerous scholars. Furthermore, nuclear deterrence theory cannot be claimed to be more concerned with 'real world politics' than nonviolent action, since the latter is a ubiquitous component of political struggles throughout the world.

I prefer an explanation that rests on the content of Sharp's ideas rather than on a prejudicial evaluation of his 'scholarship'. Arguably, liberal theorists have not taken up Sharp's approach because they are not interested in promoting social change from the bottom up, while the major critical tradition within social science, Marxism, is historically linked to violent liberation struggles, vanguard political parties and structural analysis. The major intellectual traditions which are most in tune with Sharp, anarchism and Gandhian political and economic analysis, have little following among Western social scientists .

It might be objected that although Sharp's approach is superior in terms of mobilizing social action, its lack of structural insight will lead to the failure of campaigns. While persuasive as an abstract argument, this objection overlooks the immediate circumstances associated with at least some organized nonviolent action. Those who plan non-violent action frequently have a deep understanding of the local political situation. This understanding in many cases is exactly what a structural analysis would look like if applied to the local political scene. Thus the most talented and experienced activists, even if untutored in the intricacies of abstract social theory, may incorporate the equivalent of a structural analysis into their practical activities. In this context, the shortcomings of Sharp's theory are far less important than its strengths. (By comparison, most structural approaches offer little immediate direction for campaigning, and can result in the familiar 'paralysis of analysis'.)

This ideal situation will not always apply, and so it is worth asking whether it is useful or desirable to combine the insights of Sharp's theory and structural approaches. Starting with Sharp's picture, it is relatively easy, in principle, to 'add in' social structures. Rather than assuming a stark ruler-subject dichotomy, a more complex picture of an array of partially supporting and partially antagonistic forces can be developed. The activist aim of withdrawing support then poses the difficult question of which particular intervention will best mobilize support, empower oppressed groups and lead to lasting change. From the point of view of activists, it should be possible to combine Sharp's insistence on the importance of withdrawing consent with a structural analysis which could help indicate those avenues where non-violent action would have the most desired effect.

Starting from a structural analysis, it is not so obvious what it means to incorporate insights about consent until it is made clear what is to be done with the result. If the purpose of theory is 'understanding of society' in some general scholarly sense, then there is little immediate purpose in disrupting a coherent structural analysis by adding on some material which is at cross purposes theoretically. Admittedly, structural approaches often suffer from a failure to include human agency, but it is not clear that Sharp's method is the most successful way to do this from a theoretical point of view.

On the other hand, if one of the aims of a structural analysis is to provide insights for social action, it is first necessary to go beyond a static picture of society to examine tensions, frictions, contradictions and struggles. Then it is necessary to conceptualize the role of intellectuals and social theory in social struggles. For example, what should the theorists say if their theoretical suggestions for action are bypassed in favour of some more pressing or popular campaign? In other words, how does structural analysis link its prescriptions to the theoretically untidy day-to-day welter of issues, campaigns, coalitions and counter-strategies by dominant groups?

The difficulty confronting theory in this middle-ground of providing guidance for campaigns 'on the ground' is one reason why the task is often forsaken in favour of producing erudite works which provide deep insight into structures but almost no guidance about involving oneself and others in social activism. High level theory also has the 'advantage' (for the theorist's prestige) that it is unlikely to be 'proven wrong' by the next turn of events.

The task of linking structural analysis to direct action is a vitally important one; the point here is that it is quite difficult partly because structural approaches do not have obvious and immediate implications for social action.

Thus, while it is relatively easy to criticize Sharp's theory of power at an intellectual level, it is immensely more difficult to propose an alternative theory which is more suited for effective practical application. Sharp's approach, through its simplicity and immediate relevance, throws the spotlight on apparently more sophisticated approaches by suggesting the simple question, 'What can you do with them?'

NOTES

1. On Gandhi see Gandhi (1927), numerous articles in the journal Gandhi Marg and interpretations and developments by Gregg (1966) and Shridharani (1939).

2. For example, Koch (1984) claims that Sharp's idea of withdrawing consent 'disregards the fact that. in the process of state-formation, so many sources of power accumulate and concentrate into the machinery of the state that the withdrawal of popular support would have little effect on the existing power of the ruler' (p. 3). This argument misconceives the process of withdrawing consent as applying only to those outside the machinery of the state. Sharp would include state functionaries (including police and the military) among those whose consent is necessary for the ruler to continue in power.

3. Another way of analysing the limitations of Sharp's approach is through the contrast between one-, two- and three-dimensional views of power as described by Lukes (1974). According to this classification, Sharp's approach, with its focus on behaviour and decision-making. is one-dimensional. The two- and three-dimensional views of power, which use the concepts of non-decision-making and control over the political agenda, bring in the same considerations as the structural approach as described here. Within peace research, Geeraerts (1977), among others, has distinguished between the instrumentalists, which include Sharp, and the structuralists, such as Galtung. A number of the points made here about Sharp have been made in general terms of the instrumentalists by the structuralists.

4. Useful treatments that do exist include Lipsitz & Kritzer (1975) and Summy (1983).

REFERENCES

Albert, David H., ed., 1980. Tell the American People: Perspectives on the Iranian Revolution. Philadelphia: Movement for a New Society.

Althusser, Louis, 1977. For Marx. London: New Left Books.

Baran, Paul A. & Paul M. Sweezy, 1968. Monopoly Capital. Harmondsworth: Penguin.

Dickson, David, 1974. Alternative Technology and the Politics of Technical Change. London: Fontana.

Gandhi, Mohandas K.. 1927. An Autobiography or the Story of My Experiments with Truth. Ahmedabad: Navajivan Press.

Geeraerts, Gustaaf, 1977. 'Two Approaches to Civilian Defence', pp. 5-19 in Gustaaf Geeraerts, ed., Possibilities of Civilian Defence in Western Europe. Amsterdam: Swets & Zeitlinger.

Gouldner, Alvin W., 1977-78. 'Stalinism: A Study of Internal Colonialism', Telos, no. 34 (Winter), pp. 5-48.

Gramsci, Antonio, 1971. Selections from the Prison Notebooks. London: Lawrence & Wishart.

Gregg, Richard B., 1966. The Power of Nonviolence. New York: Schocken.

Koch, Koen, 1984. 'Civilian Defence: An Alternative to Military Defence?', Netherlands Journal of Sociology, vol. 20, no. 1, pp. 1-12.

Lipsitz, Lewis & Herbert M. Kritzer, 1975. 'Unconventional Approaches to Conflict Resolution: Erikson and Sharp on Nonviolence', Journal of Conflict Resolution, vol. 19, no. 4 (December), pp. 713-733.

Lukes, Steven, 1974. Power: A Radical View. London: Macmillan.

Mandel, Ernest, 1976. Late Capitalism. London: New Left Books.

Marcuse, Herbert, 1964. One Dimensional Man. London: Routledge & Kegan Paul.

Moore, Barrington, Jr, 1978. Injustice: The Social Bases of Obedience and Revolt. London: Macmillan.

Poulantzas, Nicos, 1978. Political Power and Social Classes. London: Verso.

Sharp, Gene, 1970. Exploring Nonviolent Alternatives. Boston: Porter Sargent.

Sharp. Gene, 1973. The Politics of Nonviolent Action. Boston: Porter Sargent.

Sharp, Gene, 1979. Gandhi as a Political Strategist. Boston: Porter Sargent.

Sharp, Gene, 1980. Social Power and Political Freedom. Boston: Porter Sargent.

Sharp, Gene, 1985. Making Europe Unconquerable: The Potential of Civilian-Based Deterrence and Defence. Cambridge, MA: Ballinger.

Shridharani, Krishnalal, 1939. War Without Violence: A Study of Gandhi's Method and its Accomplishments. London: Victor Gollancz.

Skocpol, Theda, 1979. States and Social Revolutions. Cambridge: Cambridge University Press.

Skocpol, Theda, 1982. 'Rentier State and Shi'a Islam in the Iranian Revolution', Theory and Society, vol. 11, no. 3 (May), pp. 265-283.

Summy, Ralph, 1983. 'One Person's Search for a Functional Alternative to Violence', Gandhi Marg, no. 49 (April), pp. 26-44.

Walby, Sylvia, 1986. Patriarchy at Work: Patriarchal and Capitalist Relations in Employment. Cambridge: Polity Press.

Weinstein, Deena, 1979. Bureaucratic Opposition. New York: Pergamon.

Zielonka, Jan, 1986. 'Strengths and Weaknesses of Nonviolent Action: The Polish Case', Orbis, vol. 30, no. 1 (Spring), pp. 91-110.