The concept of power is at the heart of political enquiry. Indeed, it is probably the central concept of both descriptive and normative analysis.
When we talk about elections, group conflicts and state policies, we seek to explain events and processes in the political world by fixing responsibility upon institutions and agents. We are thus talking about power.
When we ask about the constitution of the good or just society, we are constrasting present conditions with some projected alternative set of arrangements that might better enable people to conduct their lives. Here too we are talking about power. It would seem impossible to engage in political discourse without raising, whether implicitly or explicitly, questions about the distribution of power in society.
It is at least partly for this reason that social and political theorists have spent so much time arguing about the concept of power—what it means, what it denotes, how it might figure in appropriately scientific analysis or how it might be ill-suited to such analysis, and, finally, why scholars and citizens should care about any of the above. Indeed, it is a striking fact that while most political theorists would agree that power is a focal concept, they would probably agree upon little else.
This has led to some awkward situations where theorists speak to each other using the same terms but meaning radically different things. Such problems of translation have never reached a point of incommensurability, and it is probably fair to say that most political theorists operate with some basic core conception of power.
The core is the notion, articulated in different ways, that the concept of power refers to the abilities of social agents to affect the world in some way or other.
The word ‘power’ derives from the Latin potere, meaning ‘to be able’. It is generally used to designate a property, capacity, or wherewithal to effect things.The concept has clear affinities with the concept of domination.
The latter means some sort of mastery or control; derived from the Latin dominium, it was originally used to designate the mastery of the patriarch over his household or domain (Tuck 1979).
While the concept of power has often been interpreted as a synonym for domination, the latter connotes an asymmetry about which the former is agnostic. The concept of power also has close connections with the concept of authority. But the latter has a normative
dimension, suggesting a kind of consent or authorization, about which the former is similarly agnostic.
The grammars of these concepts, and their interrelationships, are interesting and important (Pitkin 1972; Morriss 1980), but we will here concentrate upon the core notion of power as capacity to act, a genus of which the concepts of domination and authority can be seen as species.Such a core, however, is itself quite nebulous, and it certainly admits of many different interpretations. As a consequence, a good deal of substantive inquiry and debate has been muddled by seemingly interminable and often rarified conceptual argument.
A cynical commentator would chalk up much of this disagreement to the endless methodological fixations of political theorists, who sustain subdisciplines, journals and careers by furthering meta-theoretical argument ad infinitum (Shapiro 1989). Such cynicism would not be unwarranted, but we think that there is more to it than this.
If it is true that it is impossible to carry out political analysis without implicating the concept of power, it is also true that it is impossible to talk about power without implicating a broader set of philosophical, indeed metaphysical, questions about the nature of human agency, the character of social life and the appropriate way to study them. These broader questions are, as the history of modern social science attests, deeply contentious, and it should thus be of no surprise that this controversy has extended to the concept of power as well.
In an essay such as this it would be impossible to provide a detailed and nuanced account of such controversy. we will thus present its rough outlines.There are, we would suggest, four main models of power in modern political analysis:
1 a voluntarist model rooted in the traditions of social contract theory and methodological individualism;
2 a hermeneutic or communicative model rooted largely in German phenomenology;
3 a structuralist model rooted in the work of Marx and Durkheim;
4 a post-modernist model, developed in different ways in the writing of Michel Foucault and certain contemporary feminists.Each of these models offers not only a definition and elaboration of the concept of power, but a conception of humans, social institutions, and methods of analysis as well.
Before outlining these models, we should make three things clear.First, we will treat models as no more than rough categories or general ‘ideal types’.
we in no way intend to suggest a kind of substantive consensus among theorists typical of each model who, despite certain similarities, often share many differences on all sorts of matters. Second, while each of these models is sufficiently distinct and autonomous to be discussed separately, it is not the case, the views of methodological ideologists notwithstanding, that these models are in all respects mutually exclusive.
This is, of course, a complicated question, but we will suggest that each model in fact presents some important insights, and that theorists of power should probably think in more synthetic terms than they are accustomed to. Third, what we will discuss below are different models of the concept of power, not different theories of its distribution in particular forms of society. The discussion, in other words, will be largely meta-theoretical.
Many political theorists, including participants in conceptual debates about power, have mistakenly believed that there is a one-to-one correspondence between metatheory and theory, so that, for instance, a subscriber to Robert Dahl’s arguments about the behavioural study of power is necessarily a pluralist, and vice versa. As we have argued elsewhere (Isaac 1987), this is not the case.