Let us be frank and admit that there is no such thing as power – just as there is no such thing as “the elephant” or “the rhinoceros.” It pays to be nominalist in all these cases and avoid a metaphysics of power just as much as a metaphysics of biological kinds. A noun makes us look for a corresponding object and an abstract noun for an abstract entity. Wittgenstein has shown how that misleads us. So, no power, but no harm will be done with the term, if we take it in the right way. Let us say, then, that there exists a field of relations of something affecting (bearing on, controlling, shaping, transforming, destroying, etc.) something in some way or other. Like Foucault, we can call this the field of relations of mobile inequality. It is from this field that we usually pick a subset we call relations of power. But the choice is wide open. Thus, we end up with disputes about the nature of power, disagreements about how power is to be defined. These arise only from an ill-conceived essentialism and should be relegated to the metaphysical dustbin.
There is no single thing called “power.” There is no single subset of relations of mobile inequalities that properly constitute relations of power. We can speak, for instance, of the power of nature, or of the power of individual agents, or that of institutions, or speak of power only when the relations in question are considered to be legitimated, or when groups of agents work in unison. In each case we are carving out a different domain from the field of relations of mobile inequality. There is no disputing about which is the right one.
This being so, it may be best to think first about the totality of mobile relations of inequality. Foucault has proposed that we call them all power relations. His notion of power is, thus, a bare, minimal one; but we can proceed from it, if the need arises, to richer and more restrictive notions. We can talk, for instance, of social and political relations of power, which are, in fact, the ones that interest Foucault. There is much to recommend this method of starting from a bare concept and then to advance through a process of conceptual enrichment. But we need, perhaps, first to make more explicit what is meant by mobile inequality. Instead of calling the relations in question “mobile” we might also speak of them as “active” or “dynamic.” We are, in other words, not considering conceptual, logical, or mathematical relations of inequality. To say that proposition Q derives from proposition P does not mean that P exercises power over Q. We also don’t mean comparative relations of inequality. “A is taller than B” does not imply a power relation. We are concerned rather with mobile, active or interactive, relationships that generate a dependence and thus an inequality of one relatum to the other. “Going for a walk together” is a dynamic relation in which the partners interact with each other; but, as a symmetrical one, it is not a power relation. “Persuading someone to come along for a walk” is, on the other hand, an example of a power relation in the intended sense.
This broad notion of power proves its usefulness when we start thinking about social and political matters. It allows us to specify different mechanisms and functions of power in society and politics. We can distinguish, for instance, between prohibitive and productive relations of power. While judges exercise power mostly in a prohibitive manner, teachers are meant to exercise power productively. Society and culture exemplify both prohibitive and productive power. Power relations are, in fact, ubiquitous in society – though, of course, not universal. There are symmetrical social relations in addition to the asymmetrical ones. Sometimes the symmetrical relations arise from and are, in fact, constituted by (asymmetrical) relations of power. Social equality is often a fragile achievement teetering on a multitude of relations of inequality.
Our minimal notion of power helps us, further, with characterizing the relation between the social and the political. Power operates both in society and politics. We need to ask then: what is specific about political power? Here again we must say: there is no unique and prescribed way of doing so. We can carve out political relations from the totality of relations of mobile inequality in more than one way. Politics, like power, is not a natural kind. We can define politics, political power, and relations of power in more than one way and it is not the case that one of these definitions is the right one. And because politics is not a natural kind, it does not make sense to assert, like the Aristotelians, that we are political by nature. The only thing we can possibly say is that power relations are endemic to human life and in this sense “natural.” The identification of a particular subset of power relations as political is always a pragmatic choice. The reasonable thing is to look for a concept of political power that is diagnostically useful. But our choice will always be contestable. Hence the disputes over what is political and what is not. Is the enforcement of morals a political matter? Is religion a political concern?
In order to clarify the issue, we must revert to the previous strategy and begin with a minimal characterization of political power. This, too, is Foucault’s way of proceeding. We can follow him in saying that political relations of power are relations that exercise power on relations of power. Political relations are, thus, of second- or higher-order. The law giver, for instance, acts politically in passing laws that regulate the social interactions of citizens: these laws forbid, regulate, or nurture certain exercises of power. They forbid child abuse, regulate business, and nurture a political consciousness. Any political exercise of power can, in turn, be subject to an exercise of political power. The legislature’s exercise of power may be reviewed by a court. In a modern state there are characteristically multiple levels of the exercise of power on political relations of power. Political power thus operates in a multi-level fashion.
Foucault’s minimal concept of political power has its uses but it can also mislead us. The exercise of power on power relations is ubiquitous in society – even in those parts we normally consider to lie outside politics. Parents exercise power on relations of power when they encourage, control, or intervene in their children’s play. Given our minimal concept of political power we will have to say that the parent is then acting politically. A large class of social relations involves, in fact, the exercise of power on relations of power. We are forced to conclude that social life is suffused with politics. Some of Foucault’s readers have come to believe that he has made the stupendous discovery that politics is everywhere. But that “discovery” is due only to his choice of a minimal concept of political power. By using it we draw attention to analogies between private, family, and social life, on the one hand, and what we are used to call more narrowly politics, on the other. The danger of Foucault’s way of speaking is that we come to think of these domains as more similar than they actually are. We may thus be misled into thinking that family life is really (against all possible evidence) just as cold, calculating, and self-serving as large-scale politics can be or, alternatively, that large-scale politics is just as personal and petty as family life often is.
Such concerns justify the introduction of a narrower concept of political power. Given the obvious difference between the informality of family and social life as against the formalized exercise of power in the state, it makes sense to isolate the concept of an institutional exercise of power as a distinct notion. Doing so has, however, significant implications. Frans de Waal has argued that we can identify political power relations in the life of primates. Could this not be helpful for understanding the evolution of human politics? If we insist that politics presupposes an institutional order, we may lose hold of this insight. But talking about “chimpanzee politics” may also lead us to overlook the distinctive character of the human variety. We can try and navigate around this difficulty by distinguishing between a “proto-political” exercise of power in animal life and the properly political exercise of power in an institutional order. It may even be useful to distinguish a whole variety of uses of the term “political.” But our language is not helpful in giving expression to that possibility. A solution may be to use the term with numerical subscripts.
To speak of political power in the more specific, institutional sense forces us to be clear about the nature of institutions. Institutions we may say, for short, are, in fact, complexes of power relations or, more typically, multi-level, staggered, and hierarchical complexes of power relations. But this is still not enough. We need to add that such complexes are commonly built on a material base, require material means, and have material effects: they have a location, they occupy buildings, they process documents, they manage machineries and armaments. If we speak of institutions as systems of rules or practices, we will overlook this material aspect. The material base of institutions changes, of course, over time and with it the political relations of power. Political power is thus not a fixed quantity, but something that has a history.
The history of human power has, in fact, a dual character. There are the actual relations of power and there is their interpretation. Power is most effective, Foucault has argued, when it is invisible and thus remains uninterpreted. But what we think and say about power can both enhance and deplete it. When we believe that someone has power, his exercise of power may become more effective. When we say that someone is legitimated to exercise power, we will be more ready to submit to it. Saying that someone “has power” means that he is capable of exercising it or that he is legitimated to exercise it. It is possible to have power in one sense but not in the other. Both the exercise of power and its interpretation change over time.
This history displays what Carl Schmitt has called a dialectic of power. The more centralized and complex power relations become, the less they will be controllable by individual agents. The concentration of power and its dispersion go hand in hand. The rulers of modern states have enormous power, but their exercise of power is dependent on those who supply them with information, on the one hand, and those who execute their decisions, on the other. Donald Trump has all the power of an American president at his disposal but his decisions are determined by what he has just seen on television. So, who is the one who actually exercises power in this situation? And when the president issues one of his intemperate commands, a judge or a bureaucrat or a general may well obstruct its execution. So, who exercises power over whom at that moment? It may make sense to speak of a sovereign holder of power in simple settings, but in the complex institutional arrangements of modern life sovereignty becomes an illusion. There is no one to whom the people could hand over all power and there is no power which could be handed over entirely to the people. Forms of government constructed on the principle of sovereignty exist – but only in the imagination. And it must be admitted that this imaginary sovereignty can redirect the actual flow of relations of power.
The cycle of birth and death leads, in any case, to a constant transfer of power. At every moment someone gains and someone loses power. The transfer of power may go on gradually and unnoticed; it can also be visible and programmed or even sudden and violent, chaotic and unforeseen. Systems of political power (monarchy, dictatorship, democracy, etc.) differ not only in the way power is exercised within them but also – and perhaps more importantly – in how it is transferred. Many factors determine the nature and speed of that transfer: biological, economic, cultural, and ecological. Technological development contributes greatly to the instability in the distribution of power in human hands. As a consequence, human history results in a constant accumulation and concentration of power in some places and its dispersion in others. The balance of power is always in flux. If those in power were actually able to control this process, relations of power would already have settled in a stable pattern a long time ago. But there are instabilities, revolutions, the acquisition and deprivation of power. There is, however, no natural law that this cycle will go on forever. Who can say what the ultimate outcome will be: a complete absorption of power into a single centre, a black hole that attracts and annihilates all power around it, or a dissipation of power into an anarchic cloud of galactic dust?
The process is not entirely in our hands. The human exercise of power depends on what the material substratum will allow or what it requires. In institutional contexts those constraints will be particularly stringent since the functioning of the institution is so dependent on its material base. Our increasingly technologized world may eventually come to circumscribe the possibilities and thus the power of human agency. In the end, the power of nature is bound to overwhelm that of human action.