Foucault’s Critical Ethics

Approaching Power from a New Theoretical Basis

It is opportune, perhaps today even mandatory, that we develop a more relevant psychology and philosophy of power relationships beyond the simple conceptual framework provided by our traditional formal politics. Indeed, it may be imperative that we give some attention to defining a theory of politics which treats of power relationships on grounds less conventional than those to which we are accustomed. (Millett 1970, 24)

To understand a Foucauldian ethics, we must situate that ethics within the frameworks in which it acts. At the heart of Michel Foucault’s philosophical and historical work is an analysis of power relations and their fundamental role in society—an analysis that goes beyond “the simple conceptual framework provided by our traditional formal politics,” as Kate Millett put it, and has fostered and grounded new approaches to political engagement. For Foucault, that means that ethics must emerge within a context of relations of power, which are “everywhere” and thus inescapable. (As will become clear in the course of our exposition, the claim that power is everywhere is not a condemnation to slavery, a Weberian “iron cage,” or Hobbesian brutishness. Rather, it is for Foucault the condition of possibility for freedom.) To begin, then, an imperative first step is to understand Foucault’s conception of power relations and their role in society.

Foucault’s most explicit thinking about power is presented in the work he did in the mid- and late 1970s, particularly in his two published works, Discipline and Punish (1975) and La volonté de savoir (1976, translated as The History of Sexuality, vol. 1: An Introduction), as well as his courses at the Collège de France between 1974 and 1979. The first chapters of this work will present Foucault’s understanding of power in these years. The initial step is to grasp the core framework that Foucault proposed for understanding power and its operation.

In this chapter, then, we’ll look at his most condensed and generalized presentation, in Part 4 of La volonté de savoir. This will allow us to accomplish four tasks. First, we will be able to present the reasons why I will call Foucault’s analyses a “theory” of power; second, we will identify the mistaken theories of power that his analysis is meant to supplant—the theories against which he is arguing. We will next be able to see the basic characteristics of power according to Foucault’s theory: a network of force relations throughout society, relations that are characterized by resistance and that interact by means of local tactics and larger strategies. Fourth and finally, we will discuss the methodological guidelines that Foucault outlined for the use of this theory.

Once we understand power as a network of force relations that infuse all our social interactions, we will be able to trace in more detail the development of Foucault’s account of power. After our initial theoretical exposition, we shall examine three “moments” in the evolution of Foucault’s theory of power and his approach to ethics. In Chapter 2, we will address the first moment—roughly in 1974 through 1976, especially Discipline and Punish. Here, Foucault traces the emergence of a new kind of power—disciplinary power—out of older models of sovereign right and power in the seventeenth and eighteenth centuries. Disciplinary power is an absolutely central element of Foucault’s analysis of modern power—indeed, it is what he is most famous for. But his articulation of it, especially in Discipline and Punish, makes two related mistakes: it takes disciplinary power as the only modality of modern power, and it can give the impression that the omnipresence of disciplinary power leaves us trapped in a bleak, despairing world.

The second moment (and our third chapter) focuses on how Foucault revised and expanded his understanding of power from that point—a process that took place in 1976 through 1978. In his 1976 course at the Collège de France, for example, Foucault uses a study of “war” as the principal or guiding metaphor for power relations as an occasion to reflect upon his own methodological presuppositions as well as to trace further the limits of disciplinary power. In this period, he also came to recognize that relations among macro-level phenomena—in particular, “populations”—are not adequately described by disciplinary power, and so his catalog of the modes of power must be broadened and supplemented.

He thus concludes both the 1976 course and La volonté de savoir with the introduction of a new kind of power, “biopower,” a nondisciplinary power over “man as species” (CdF76ET, 242). Foucault’s discussion of biopower complements that of disciplinary power, allowing for a more complete description of the contemporary regime of power relations. Much of his work at the Collège de France in 1978 and 1979 is devoted to an articulation and development of the concept of biopower and what he terms “pastoral power,” (a premodern ancestor of both biopower and disciplinary power). The various modalities of modern power can then be brought together under an umbrella term, “governmentality.”

Throughout this discussion, Foucault’s analysis of power relations underscores two key facts. First, power relations in and of themselves are inadequate for a full analysis of society—we must speak of other kinds of relations (such as ethical relations) as well. Second, there are a number of resources for an ethics within the very analysis of power relations that Foucault has developed. Thus Foucault is prepared to undertake the third moment, which we will examine in Chapter 4—in which the ethical trajectories emerging from this analysis of power relations come to the foreground. Using the umbrella term “governmentality,” Foucault is thus able once again to reframe his analysis of these power relations in terms of “governing oneself and others.” This construction emerges from Foucault’s studies of ancient Greek and Roman practices. Marked by an emphasis on the self, it yields possibilities for a richer conception of the self and of agency within a framework of power relations and a regime of power. “The notion of governmentality allows one, I believe, to set off the freedom of the subject and the relationship to others, i.e., that which constitutes the very matter of ethics” (DE356.2, 20).

A “THEORY” OF POWER

What we can call a “theory” of power emerges from Foucault’s mid-1970s analyses of psychiatry, the prison, and sexuality. This theory is not restricted to descriptions of one empirical period or “regime” but describes certain general characteristics of power and its operation across historical epochs and periods.

Foucault disliked the term “theory.” He noted in La volonté de savoir that “the aim of the inquiries that will follow is to move less toward a ‘theory’ of power than toward an ‘analytics’ of power” (1976ET, 82). Foucault emphasized analysis over theory in part because he was reluctant to make any claim to a permanent or complete understanding. In his 7 January 1976 lecture at the Collège de France, Foucault indicated at least part of his distrust for theory: “given that the question ‘What is power?’ is obviously a theoretical question that would provide an answer to everything, which is just what I don’t want to do . . .” (CdF76ET, 13). It is only insofar as theories can be used “untheoretically” in this sense—that is, without claiming to answer everything—that they can be valuable: all-encompassing or global theories “have, I think, provided tools that can be used at the local level only when, and this is the real point, the theoretical unity of their discourse is, so to speak, suspended” (CdF76ET, 6). Nevertheless, he did refer to his own project as a theory: his task “is a question of forming a different grid of historical decipherment by starting from a different theory of power” (1976ET, 90–91, my italics).1 He would again describe it as a “theoretical shift” in The Use of Pleasure (1984aET, 6; 1984a, 12: “déplacement theorique”). For Foucault, the term theory must be used with caution; we should embrace theory only in the sense of “a theoretical production that does not need a visa from some common regime to establish its validity” (CdF76ET, 6).

Given Foucault’s reluctance about the term theory, however, we may recall Heidegger’s remark in a 1924–1925 lecture course:

For one who has learned to understand an author it is perhaps not possible to take as a foundation for the interpretation what the author himself designates as the most important. It is precisely where an author keeps silent that one has to begin in order to understand what the author himself designates as the most proper. (Heidegger 1924, 32–33)

Even if Foucault himself is silent about the theoretical significance of his analyses of power, we can take it as one of the foundations for our own reading and for his analyses of ethics.

With this terminological caution in mind, I shall use the term “theory” with respect to Foucault’s work in an experimental sense: a theory is a hypothesis to organize diverse data but also to be tested and revised or abandoned in light of that data. That a theory aims to be more general than a description of a single historical period or epoch is an essential part of its value and usefulness for our understanding of the phenomena it encompasses, and it remains a useful term with respect to Foucault’s analyses of power. This generalizability does not entail, however, that such a theory need “a visa from some common regime” or frame of analysis, such as Marxism or psychoanalysis, for its authority; on the contrary, a theory’s warrant will come from the empirical data that it organizes and that supports it. An essential part of my project, then, is to present Foucault’s theory of power as such. As we shall see in the following, Foucault continued to revise his theory in light of his ongoing empirical studies.

Foucault’s theory of power encompasses a second kind of generalizability as well: power is omnipresent—that is, it can be found in all social interactions. As he put this in 1977, “it seems to me that power is ‘always already there,’ that one is never ‘outside’ it” (DE218.1, 141). This omnipresence has two central characteristics. First, power is coextensive with the social body, with the field of social relations. Second, it does not stand alone but is interwoven with, revealed in, other kinds of social relations.

Note that this conception of omnipresence is a restricted one.2 First, Foucault is not saying that power functions as a trap or cage but rather that it is present in all of our social relations. That is to say that in any social interaction, from our most intimate and egalitarian³ to the most hierarchical, power constitutes an at least implicit aspect of that interaction. Second, Foucault is also not saying that all relations reduce to power relations but simply that power is present in all social relations. This leaves open the possibility that kinds of relations other than power are also operative in particular social relations.

We can look at each of these limitations in turn. First, the claim that there is no position “outside” of power does not mean that freedom is impossible. This is because power relations always entail resistance, risks and, struggles and the possibility to change the situation and the distribution of power within it. This is indicated in the logic of power relations and resistance. Properly understood, the recognition of power relations within a situation includes recognizing the possibility of altering the situation.

We are always in this kind of situation [inflected by power relations]. It means that we always have possibilities, there are always possibilities of changing the situation. We cannot jump outside the situation, and there is no point from which you are free from all power relations. But you can always change it. (DE358.1, 28)

Second, power relations’ omnipresence does not entail that they are the only kind of relation in society.

The omnipresence of power: not because it has the privilege of consolidating everything under its invincible unity, but because it is produced from one moment to the next, at every point, or rather in every relation from one point to another. Power is everywhere; not because it embraces everything, but because it comes from everywhere. (1976ET, 93)

Power does not “consolidate everything” or “embrace everything”; rather, it emerges as immanent in all kinds of social relations. That social relations could not be exterior to power relations does not entail that they consist entirely or exclusively of power relations. And so power alone may not be adequate to explain all, or every aspect of, social relations.

Foucault’s theory of power, then, has two kinds of generalizability: synchronic and diachronic. Power relations can be found in all of our social interactions, and the various kinds of power relations that Foucault describes, while most clearly recognizable in modern social institutions, are also present in earlier epochs.

Judith Butler, a feminist theorist profoundly influenced by Foucault, highlights both the dangers and necessity of a “theory of power”:

On the one hand, any analysis which foregrounds one vector of power over another will doubtless become vulnerable to criticisms that it not only ignores or devalues the others, but that its own constructions depend on the exclusion of the others in order to proceed. On the other hand, any analysis which pretends to be able to encompass every vector of power runs the risk of a certain epistemological imperialism which consists in the presupposition that any given writer might fully stand for and explain the complexities of contemporary power . . .

This demand to think contemporary power in its complexity and interarticulations remains incontrovertibly important even in its impossibility. (Butler 1993, 18–19)

The strength of Foucault’s analysis is that it allows us to recognize multiple vectors. In fact, his theory continued to evolve—for example, with the introduction of the notion of biopower—precisely because certain vectors were initially insufficiently elaborated. I think that his theory can also avoid the charge of “epistemological imperialism” because this imperialism would become an issue only if one attempted to give a “total” explanation of a particular given situation. Foucault’s theory will recognize that an analysis of any particular situation will be partial and incomplete—“the ‘distributions of power’ . . . never represent only instantaneous slices taken from processes” (1976ET, 99)—but what his theory does offer is a toolkit that is as complete as possible for those analyses. As Butler recognizes, the task that the toolkit makes possible is “incontrovertibly important” even if such a complete description remains impossible, an ideal only to be approached asymptotically.

As I’ve just noted, Foucault’s theory of power is not a static conception. His work focused explicitly on power in the mid-1970s, and his theory evolved, or rather was refined and specified, in Foucault’s later work. (Foucault himself remarked [DE326.1, 237] that the themes he identifies in Discipline and Punish were in fact present at the heart of his much earlier History of Madness. Of course, there is a certain retrospective rereading and reinterpretation going on in this claim, but it is not entirely irrelevant. Certain problems come into focus in later work that were not yet made explicit in earlier work, and aspects of Foucault’s theory of power become more sharply delineated over time, as well. Indeed, in his 1973–1974 Collège de France course he introduced this new emphasis upon power as a basis for social analysis through a critique of History of Madness [CdF74ET, 12–16].) We will look at each of these moments in detail in this and subsequent chapters. Our first task is to see how Foucault first articulated his understanding of power, when it became the primary object of his analyses in the mid-1970s.

Foucault’s single most explicit published discussion of power in this period is given in Part 4 of La volonté de savoir, “The Deployment of Sexuality.” The first two of four chapters in this part, “Objective” and “Method,” are the most important for our purposes. He states his task clearly at the beginning of the first chapter—it is to work “toward an ‘analytics’ of power: that is, toward a definition of the specific domain formed by relations of power, and toward a determination of the instruments that will make possible its analysis” (1976ET, 82).

HOW NOT TO UNDERSTAND POWER

Foucault first distinguishes his theory from three mistaken, inadequate, or misleading conceptions (each of which corresponds to a tradition or school of social thought).

The word power is apt to lead to a number of misunderstandings—misunderstandings with respect to its nature, its form, and its unity. By power, I do not mean “Power” as a group of institutions and mechanisms that ensure the subservience of the citizens of a given state [such as characterize many liberal analyses]. By power, I do not mean, either, a mode of subjugation which, in contrast to violence, has the form of the rule [typical of psychoanalytic approaches]. Finally, I do not have in mind a general system of domination exerted by one group over another [i.e., class oppression], a system whose effects, through successive derivations, pervade the entire social body [as in many Marxist views]. (1976ET, 92; my comments in brackets)

Foucault’s worry is not that these analyses are entirely useless but that they often mischaracterize an accidental feature of power in a particular context as an essential characteristic of power in general. “The analysis, made in terms of power, must not assume that the sovereignty of the state [liberal], the form of the law [psychoanalytic], or the over-all unity of a domination [Marxist] are given at the outset; rather, these are only the terminal forms power takes” (1976ET, 92; my comments in brackets). So each of these features may in fact be present in certain contexts as terminal forms, but none is fundamental. We must first develop a new method—based on a richer theory—that begins with the basic molecules of power relations and then builds upward to more complex forms.

The most important misconception is what he terms a “juridico-discursive” understanding of power—the bulk of the first chapter of Part 4 is devoted to a description of this “juridico-discursive” analysis. This conception, he notes, is common to many “political analyses of power, and it is deeply rooted in the history of the West” (1976ET, 83). This conception is characteristic of both psychoanalytic (Freudian and Lacanian) and critical-theoretical (Marcusean) approaches to sexuality; it is especially important therefore that we recall Foucault is not endorsing but criticizing this theory of power. His argument, in fact, is that this conception, so generally accepted, has functioned as a mask by which much of the actual operation of power is obscured, thereby making many of the actual mechanisms of power tolerable (1976ET, 86). (He pursues a similar project of unmasking power’s actual modes of operation in Discipline and Punish.)

According to this “juridico-discursive” theory, power has five principal characteristics. First, power always operates negatively, that is, by means of interdictions. Second, power always takes the form of a rule or law. This entails a binary system of permitted and forbidden, legal and illegal. These two characteristics together constitute the third: power operates through a cycle of prohibition, a law of interdiction. Hence (and fourth), this power manifests in three forms of prohibition—“affirming that such a thing is not permitted, preventing it from being said, denying that it exists” (1976ET, 84)—which reveal a logic of censorship. Fifth and finally, the apparatus of this power is universal and uniform in its mode of operation: “From top to bottom, in its over-all decisions and its capillary interventions alike, whatever the devices or institutions on which it relies, it acts in a uniform and comprehensive manner; it operates according to the simple and endlessly reproduced mechanism of law, taboo, and censorship” (1976ET, 84).

Notice how Foucault has characterized this uniformity, “in its over-all decisions and its capillary interventions alike.” Implicit in this characterization is an important distinction between macrostructures (the “over-all decisions”) and micropractices (“capillary interventions”)—a distinction that will be very important in the development of Foucault’s own understanding of power. His analysis begins at the micro level (in Discipline and Punish, for example) and is modified as it encompasses the macro level (especially in the 1978 and 1979 Collège de France courses). That this distinction is not made in the “juridico-discursive” view is just another indication of how it differs from Foucault’s own analysis and how it is mistaken about, and masks, the actual operation of power.

Why does Foucault term this view a “juridico-discursive” representation of power? First, it is juridical because it is modeled upon the law, upon prohibition: “it is a power [more precisely a representation of power] whose model is essentially juridical, centered on nothing more than the statement of the law and the operation of taboos” (1976ET, 85). But as Foucault makes clear, the actual operation of power cannot be reduced to one model—the law, the state, or domination—but functions in a variety of forms and with varying means or techniques.

Second, according to this view, power is in essence discursive: its prohibitions are tied together with what one can say as much as what one can do; in this way restrictions on language should also function as restrictions upon reality and action—this is the heart of the “logic of censorship.”

It links the inexistent, the illicit, and the inexpressible in such a way that each is at the same time the principle and the effect of the others: one must not talk about what is forbidden until it is annulled in reality; what is inexistent has no right to show itself, even in the order of speech where its inexistence is declared; and that which one must keep silent about is banished from reality as the thing that is tabooed above all else. (1976ET, 84)

While this view emphasizes discourse as the primary arena in which power’s effects manifest, Foucault notes that discourses are related to power in much more complicated ways than this view would suggest:

Discourses are not once and for all subservient to power or raised up against it . . . discourse can be both an instrument and an effect of power, but also a hindrance, a stumbling-block, a point of resistance and a starting point for an opposing strategy. (1976ET, 100–101)

We have here a skeletal overview of the kind of analysis that Foucault rejects. But several of these characteristics merit more detailed comment.

First, the logic of censorship (the fourth feature of power, which we have just seen at the heart of the “discursive” emphasis) is also the logic of repression, the principal form in which power is manifested according to this theory—particularly in its psychoanalytic variants:

These are the characteristic features attributed to repression, which serve to distinguish it from the prohibitions maintained by penal law: repression operated as a sentence to disappear, but also an injunction to silence, an affirmation of nonexistence, and, by implication, an admission that there was nothing to say about such things, nothing to see and nothing to know. (1976ET, 4)

This passage comes from Part 1 of the book, “We Other Victorians.” This part of the book is frequently misunderstood by students because it is written almost entirely in an ironic voice—the views expressed therein are largely a prose portrait of a view (including this conception of power) that Foucault wants to throw into relief in order to criticize. (His critical distance is indicated in this passage by the phrase “attributed to.”) It is interesting that in this description, repression is distinguished from penal law since penal practice in fact provides the empirical basis and paradigmatic example for Foucault’s own analyses of power. This distinction functions as one aspect of the masking that this theory provides for power’s actual operation.

There is also something more to be said about the universal uniformity of power on this view (the fifth principal characteristic). The problem here is that the model is too reductive—it collapses a variety of kinds of power to one model, the law. But it is not just that such a unidimensional analysis fails to describe adequately diverse manifestations of power. It also fails to recognize an important distinction between micro and macro forms of power—one of the key distinctions that will allow Foucault’s own analysis to be much more nuanced. Microrelations occur in local interactions, between individuals and small groups such as families. Macro forms are larger, representing interactions between institutions (such as the state or the judicial system) as well as patterns formed out of many microrelations (such as cultural norms). These diverse forms may in fact operate according to very different principles and models.

Furthermore, the uniformity that Foucault criticizes here is a synchronic uniformity. The “juridico-discursive” view asserts that all kinds of power at a given time operate according to one model. It is important that we distinguish this synchronic uniformity from a diachronic consistency. I have argued above that part of what qualifies Foucault’s analyses as a proper “theory” of power is diachronic consistency or generalizability: it gives us resources to understand the manifestation and operation of power not only in our contemporary epoch but also in earlier periods. (This can be seen in Discipline and Punish, as Foucault traces the transformations of penal power from one paradigm or epoch to another. We will return to this point later.) In short, a theory that recognizes multiple models has more resources with which to describe diverse historical periods.

Finally, we should note that the “juridico-discursive” view of power entails a concomitant view of the subject:

Confronted by a power that is law, the subject who is constituted as subject—who is “subjected”—is he who obeys. To the formal homogeneity of power in these various instances corresponds the general form of submission in the one who is constrained by it—whether the individual in question is the subject opposite the monarch, the citizen opposite the state, the child opposite the parent, or the disciple opposite the master. A legislative power on one side, and an obedient subject on the other. (1976ET, 85)

Note that Foucault here explicitly relates the “obedient subject” who is “subjected” to law as the image of the subject correlated with the juridico-discursive account of power. Many understand Foucault to posit just such an “obedient subject” as the subject of power properly understood; this misunderstanding is at the root of many criticisms of his work. Yet that he defines it as a function of the juridico-discursive view strongly suggests that we should not be so hasty. The implication is that this understanding of the subject is itself mistaken or inadequate, part of the “mask” of this account of power.

It may in fact be the case that Foucault’s analyses in this period are somewhat confused and that this understanding of the subject is still at work in his own analyses—that he occasionally lapses into this view and does not entirely peel away the “mask” that this view provides for power. Indeed, his discussions of disciplinary power (in the first moment) do bear a certain resemblance to this juridico-discursive view. One of the motivations behind Foucault’s continuous revisions of his analysis of power, as I hope to show, as well as his turn to ethics, is the need to account adequately for agency on the part of individuals and institutions. So we need to remember that part of his ongoing project is to overcome this reductive view of the subject as merely an “obedient subject.” As he notes in the introduction to The Use of Pleasure (volume 2 of The History of Sexuality, which was published months before his death in 1984):

It appeared that I now had to undertake a third shift, in order to analyze what is termed “the subject.” It seemed appropriate to look for the forms and modalities of the relation to self by which the individual constitutes and recognizes himself qua subject. (1984aET, 6)

Foucault will come to this new shift on the basis of his new analysis of power. And we can now turn to Foucault’s own analysis, having completed our survey of those views of power from which Foucault wishes to free his own analysis.

It is this image that we must break free of, that is, of the theoretical privilege of law and sovereignty, if we wish to analyze power within the concrete and historical framework of its operation. We must construct an analytics of power that no longer takes law as a model and a code. (1976ET, 90)4

A FOUCAULDIAN VIEW OF POWER

It is time now for us to turn to this constructive task and to begin to articulate Foucault’s own positive understanding of power. Foucault describes his task as:

a question of forming a different grid of historical decipherment by starting from a different theory of power; and, at the same time, of advancing little by little toward a different conception of power through a closer examination of an entire historical material. (1976ET, 90–91)

The first chapter of Part 4 was devoted to characterizing the view of power that must be overcome—one of his principal objectives in this work. The next chapter, “Method,” will begin the positive task by outlining Foucault’s new theory of power, which will guide his empirical investigations by providing the hypotheses to be tested.5 (In this section, I shall offer what is essentially a commentary upon this chapter, quoting it extensively as Foucault’s most condensed presentation of his theory of power in these years.)

He begins:

It seems to me that power must be understood in the first instance as [1] the multiplicity of force relations immanent in the sphere in which they operate and which constitute their own organization; as [2] the process which, through ceaseless struggles and confrontations, transforms, strengthens, or reverses them; as [3] the support which these force relations find in one another, thus forming a chain or a system, or on the contrary, the disjunctions and contradictions which isolate them from one another; and lastly, as [4] the strategies in which they take effect, whose general design or institutional crystallization is embodied in the state apparatus, in the formulation of the law, in the various social hegemonies. (1976ET, 92–93; my numeration in brackets)

There is much to unpack in this sentence. The bracketed numbers indicate four principal aspects of Foucault’s initial definition. We have a set of “force relations,” processes by which these relations are transformed, systems or disjunctions that are constituted by the interplay of these force relations, and larger strategies (or “terminal forms”) with general and institutional characteristics that emerge from these relations, processes, and systems. He begins at the micro level, looking at local relations of force rather than at the macro level of hegemonies and states, which can only be fully understood as functions of the local relations. (As we’ll see later, however, macro forms of power are not entirely built of these micro forms. Or rather, macro forms can create a sort of feedback loop that impacts the micro forms, not to mention that there are a number of micro forms that interact with one another in a variety of ways.)

First of all, then, power must be understood at the micro level as relations of force. Foucault is unambiguous on this point: “It is in this sphere of force relations that we must try to analyze the mechanisms of power” (1976ET, 97). But what are these “force relations” at the basis of power? Force relations seem to be the basic unit, the undefined or given, in this approach to power. Very broadly, force relations consist of whatever in one’s social interactions that pushes, urges, or compels one to do something. With this term, Foucault makes an explicit analogy to physics—he refers on numerous occasions, for example, to the “micro-physics of power” (1975ET, 26; CdF74ET, 16; cf. also CdF78ET, 295–296).

We must be careful not to make too much of the analogy to physics—I do not think, for example, that Foucault means to propose that analysis of social relations should be reduced to equations of force relations.6 As he notes at the Collège de France, “explaining things from below also means explaining them in terms of what is most confused, most obscure, most disorderly, and most subject to chance” (CdF76ET, 54). So to speak of microrelations and force relations is not necessarily to simplify or clarify but to complicate our analysis of events and actions. Furthermore, an important part of my larger argument is that these force relations, while important, do not constitute the whole of social interactions for Foucault. (Modern physicists also tell us that even for physical phenomena, which can be analyzed in terms of equations of force, there are simply too many variables for a complete or accurate analysis [Caraher 2004].) Given this caution, however, we can note certain implications of the analogy.

We can use the analogy to help us understand this notion of force relations as the basic unit of power. In Newtonian physics, force is defined as mass times acceleration. Force is thus the extent to which a body will be put into motion—larger objects (greater mass) will require a greater force to begin moving; a greater force will also be necessary to make an object move more quickly (greater acceleration). The important point here is that “force” is whatever serves to put an object into motion regardless of the origin or source of that force. Force may be introduced by gravity, magnetism, or some other means. Its action thus can be described independently of any particular agent or object as the “creator” of that force. Analogously, Foucault speaks of power relations in terms of force relations without reference to a source or agent. This suggests that Foucault does not mean to imply that individuals cannot act as agents within power relations but rather to draw our attention, especially for methodological reasons, to the force relations as such rather than to agents or actors. Closer examination of the characteristics of these force relations should help make this clearer.

To recall, Foucault began with the claim that “power must be understood in the first instance as the multiplicity of force relations immanent in the sphere in which they operate and which constitute their own organization” (1976ET, 92). Three features of these force relations are thus delineated: there is a multiplicity of force relations, force relations are immanent in social interactions, and these force relations are organized internally.

First, that there is a multiplicity means that we will find many different relations of force intersecting and overlapping in our social interactions. What is more, this suggests that these force relations will not all be of the same quality or kind: there will be multiple sorts of force relations, which may have different particular characteristics or impacts. To draw on the analogy to physics again, we could say that different forces will be present in the same field, as are gravity and magnetism, and that some of these forces will be stronger than others, and some stronger in certain contexts but not in others.

What sort of presence do these relations have, then? The second feature delineated in this description is that force relations are “immanent in the sphere in which they operate.” That these relations are “immanent” means that they exist only within a certain domain or discourse. In other words, they are not concrete, like bodies, but incorporeal, like the laws of physics. They are nevertheless genuinely present—and, like laws, their presence can be felt in very concrete ways. The analogy to physics is again useful here. As physical bodies interact, they exert relations of gravity, magnetism, etc. upon each other. Similarly, social interactions are constantly permeated by these relations of force, power relations. Foucault thus describes force relations as a substrate: “it is the moving substrate of force relations which, by virtue of their inequality, constantly engender states of power, but the latter are always local and unstable” (1976ET, 93). He notes in the 1976 Collège de France course that “power is never anything more than a relationship that can, and must, be studied only by looking at the interplay of the terms of the relationship” (CdF76ET, 168).7 This means that “relations of power are not in a position of exteriority with respect to other types of relationships (economic processes, knowledge relationships, sexual relations), but are immanent in the latter” (1976ET, 94). So “power is not an institution, and not a structure; neither is it a certain strength we are endowed with; it is the name that one attributes to a complex strategical situation in a particular society” (1976ET, 93). This has an important corollary: power is omnipresent (as discussed above) “because it is produced from one moment to the next, at every point, or rather in every relation from one point to another. Power is everywhere; not because it embraces everything, but because it comes from everywhere” (1976ET, 93). (Recall our earlier discussion of power’s omnipresence.)

So there is a multiplicity of force relations, which are immanent in social interactions. The third feature in this initial characterization is that these force relations “constitute their own organization.” On the one hand, these force relations “are the immediate effects of the divisions, inequalities, and disequilibriums which occur in [other types of relationships]” (1976ET, 94). But on the other hand: “If in fact [these force relations] are intelligible, this is not because they are the effect of another instance that ‘explains’ them, but rather because they are imbued, through and through, with calculation: there is no power that is exercised without a series of aims and objectives” (1976ET, 94–95). These calculations, these aims and objectives, which Foucault will refer to as “tactics” and “strategies,” constitute the internal organization of these power relations. We will come back to explain this more carefully later.

Several other propositions also emerge from this core understanding of power. Foucault delineates five. First, since power emerges in relationships and interactions, power is not possessed but exercised. “It is not the ‘privilege,’ acquired or preserved, of the dominant class, but the overall effect of its strategic positions” (1975ET, 26). At stake here are two competing models of power—one based on a contract (possession), the other based on perpetual battle (strategies or war). Foucault traces the struggle between these two models in extensive detail in both Discipline and Punish and “Society Must Be Defended.” As he notes in Discipline and Punish, his study “presupposes that the power exercised on the body is conceived not as a property, but as a strategy . . . ; that one should take as its model a perpetual battle rather than a contract regulating a transaction or the conquest of a territory” (1975ET, 26).

Second, power relations are not exterior to other relations. This reiterates the point above about immanence.

Third, “power comes from below; that is, there is no binary and all-encompassing opposition between rulers and ruled at the root of power relations and serving as a general matrix” (1976ET, 94). This elaborates upon the observation above that there are a multiplicity of power relations, and Foucault makes two important points with this observation. First, power is not reducible to a binary relationship; second, power comes from below. That power is not binary means that we cannot reduce all sorts of power to one model. Furthermore, we cannot understand power as a simple relation between ruler and ruled, oppressor and oppressed, master and slave. Rather, power emerges from a variety of overlapping and intertwined relationships.

In part, Foucault insists that power is not binary in order to extricate his view from certain misunderstandings or criticisms (as we saw earlier)—to underscore that this is not a structuralist or deconstructive account of power, nor will it reduce all power to repression or class oppression, as psychoanalytic or Marxist approaches would.

But Foucault also stresses here that power comes from below. We cannot begin to understand power in the first instance by looking at monarchies or states, by looking at the top of any chain of command. Rather, we must look at the complex webs of interwoven relationships—what Foucault calls the “microphysics” of power (1975ET, 26). Power develops in the first instance in specific, local, individual choices, behaviors, and interactions. These combine in myriad ways to constitute larger social patterns, eventually yielding macroforms, which one typically thinks of when one thinks about “power,” such as societies, states, and kings—just as everyday objects are constituted by atoms and molecules.

In fact, Foucault’s own analysis of power will stress interactions at both the micro level of individuals (which he will describe as disciplinary techniques of the body) and the macro level of populations (which he will term biopolitics). But his work in 1975–1976 is focused almost exclusively on the micro level. He begins to approach an analysis of the macro level only late in this period, at the end of “Society Must Be Defended” and the last part of La volonté de savoir. A more careful articulation of the macro level, the biopolitics of populations will be developed in the Collège de France courses of 1978 and 1979.

Fourth of the five propositions that emerge from Foucault’s conception of power is that “power relations are both intentional and nonsubjective” (1976ET, 94) This juxtaposition is, frankly, puzzling and has led to quite a bit of misunderstanding about Foucault’s analysis. First, power is intentional: power relations are “imbued, through and through, with calculation: there is no power that is exercised without a series of aims and objectives” (1976ET, 95). Foucault refers to these aims and objectives as “tactics” and “strategies” and notes that these are what constitute its “rationality.” But power, he insists here, is also nonsubjective: “But this does not mean that it results from the choice or decision of an individual subject” (1976ET, 95). Nor, he continues, can it be located in groups such as economic decision makers, governing castes, or the state apparatus.

The problem here, the apparent paradox, is that according to Foucault’s description, power has to be exercised by someone or something, but if it is nonsubjective, there can’t be a “someone” exercising that power. Given what he says in Discipline and Punish about how the subject is constructed by power relations (which we’ll discuss in the next chapter), Foucault seems to be erasing agency or locating agency in a noncorporeal “power” rather than individuals and institutions. I think the problem can be resolved with two observations. First, part of his point here is that the effects of the exercise of power reach beyond any individual’s (or group’s) intentions or control. As we’ve already seen, Foucault is arguing against the view that “the state” acts as a monolith, and he is arguing for the importance of microevents, with their ripples and interactions, in order to understand macrophenomena. This means that local actions often have unintended macroconsequences and that one’s control of macroprocesses will always be limited and incomplete. Macrophenomena result from the concatenation of many microevents, but they are not the direct result of any particular individual action or choice. This is, then, an argument for a system-level, rather than individual-level, understanding of power relations. Foucault’s distinction between tactics and strategy parallels this micro/macro distinction. Tactics are local, micro; strategies are macro, systemic. But as I’ve suggested above, Foucault has not yet come to a mature understanding of power at the macro level—that will not emerge clearly until 1978–1979. (This may account for some of the confusion in his language here.)

The second observation to be made here is a logical one. To understand subjectivity as constituted (in part) through power relations is not to deny that subjects can act intentionally. Nevertheless, the status of subjectivity will become a focal point of Foucault’s investigations in later years. We will return to this theme in later chapters.8 It is an important one, because one of the fundamental questions for ethical action has to do with the individual’s ability to make decisions that are not “merely” determined by the relations of power in which they emerge—in other words, a question of freedom.

This question may in fact lie at the heart of readers’ varied reactions to Foucault’s analyses. Those who understand the claim that individuals are constituted by power relations to constitute a denial of freedom find his vision bleak. Those of us who find in his analyses the tools with which to increase our self-awareness, and hence our own freedom, hear wellsprings of hope in his discussion of the continuous transformations of power through history. At a minimum, on the latter view, if power relations are in fact best understood as a necessarily ongoing battle, then the battle is never utterly lost. (We’ll return to this question of hope, too, in later chapters.)

Indeed, Foucault seems to anticipate this objection, this worry about freedom, in the fifth of the five propositions that he discusses here. Power is always accompanied by resistance; resistance is in fact a fundamental structural feature of power: “Where there is power, there is resistance, and yet, or rather consequently, this resistance is never in a position of exteriority in relation to power” (1976ET, 95). Without resistance, without two bodies (or minds) pushing or pulling against each other, there is no power relation.

This fifth point encapsulates each of the preceding four. Power is exercised (first proposition) in the very interplay of force and resistance, this interplay is present in all social interactions (second proposition), force and resistance are manifest even in microinteractions between individuals as well as states (third proposition), and while each person may choose to apply force or resist, the ultimate outcome of the relation cannot be controlled by one party (power is intentional and nonsubjective—fourth proposition).

On the role of resistance in constituting power, Foucault’s position will not change. In a 1984 interview, for example, Foucault reiterates that “in the relations of power, there is necessarily the possibility of resistance, for if there were no possibility of resistance—of violent resistance, of escape, of ruse, of strategies that reverse the situation—there would be no relations of power” (DE356.2, 12). He adds that “there cannot be relations of power unless the subjects are free. . . . If there are relations of power throughout every social field it is because there is freedom everywhere” (DE356.2, 12).

Let’s stop for a moment and recall where we are in the discussion. We began this section with a long quotation, repeated again here, in which Foucault identified four principal aspects of power:

Our discussion so far has focused on explicating only the first of these, that power consists of multiple force relations. (We’ve covered a lot of ground along the way.) We can now quickly address the remaining three aspects that Foucault identified here.

The second point that Foucault makes here is that these force relations are processes, not static, and are constantly being transformed. These transformations take the form of ceaseless struggles and confrontations between the original force and its accompanying resistance, and these sometimes strengthen the power relations but other times weaken or reverse it. These processes also produce a number of interrelationships and systems, as various power relations reinforce or undermine one another (third point). Here Foucault introduces a distinction between tactics and strategy: tactics are the local rationalities of power in particular cases; strategies, on the other hand, are the larger systemic or global patterns of power. And (fourth point) these strategies are built out of combinations and concatenations of those local tactics.

The rationality of power is characterized by tactics that are often quite explicit at the restricted level where they are inscribed (the local cynicism of power), tactics which, becoming connected to one another, attracting and propagating one another, but finding their base of support and their condition elsewhere, end by forming comprehensive systems. (1976ET, 95)

These comprehensive systems, or strategies, constitute “institutional crystallizations” out of the interaction and combination of locally fluid power relations and become recognizable terminal forms like the state and the other types he enumerated.

Foucault elaborates further on strategies and tactics in Discipline and Punish:

If there is a politics-war series that passes through strategy, there is an army-politics series that passes through tactics. It is strategy that makes it possible to understand warfare as a way of conducting politics between states; it is tactics that makes it possible to understand the army as a principle for maintaining the absence of warfare in civil society. (1975ET, 168)

His use of the army as a symbol for tactics here is part of a broader use of military metaphors. Just as Foucault has drawn an extended analogy to physics to explain his understanding of power, he also uses an analogy to war—inverting Clausewitz’s saying that war is politics by other means 9—to underscore the hypothesis that power relations lie at the molecular level of all social interactions. (Indeed, in his 1973–1974 Collège de France course, he even acknowledged that his is a “pseudo-military vocabulary” [CdF74ET, 16].)

In this military analogy we can see the “Hobbesian hypothesis” about power as the organizing principle of society quite clearly articulated in Foucault. Or rather, as I suggested in the Introduction, we can see Foucault’s exploration and testing of the hypothesis—in order to determine its limits, to show that it is not sufficient, even if necessary, for a complete understanding of social relations.

Foucault makes an interesting and important addition at this point in La volonté de savoir:

Power’s condition of possibility, or in any case the viewpoint which permits one to understand its exercise, even in its more “peripheral” effects, and which also makes it possible to use its mechanisms as a grid of intelligibility of the social order, must not be sought in the primary existence of a central point, in a unique source of sovereignty from which secondary and descendent forms would emanate; it is the moving substrate of force relations which, by virtue of their inequality, constantly engender states of power, but the latter are always local and unstable. (1976ET, 93)

A number of points can be seen when we consider this long single sentence (portions of which we’ve already discussed) as a whole. First, Foucault emphasizes power’s peripheral effects—an examination of these peripheral effects will allow us to observe the actual operation of power, which is obscured and masked if we only look at larger, central forms like the state. Following from that, this shift in perspective provides a new viewpoint from which to begin a study of power—and will thus entail a new methodological approach. Third (a point that Foucault has been making throughout this discussion), this allows us to grasp the actual local basis of power’s larger effects. So this approach will make possible a fully realized test of the Hobbesian hypothesis—by using mechanisms of power as our grid of intelligibility for the social order.

Finally, and perhaps most intriguingly, Foucault uses explicitly Kantian language when he speaks of power’s “condition of possibility” here—even if he immediately shies away from its implications (his “or in any case the viewpoint . . .” seems to be a rhetoric of retreat). Kant has loomed in the background of Foucault’s work throughout his career, from his translation and commentary on Kant’s Anthropology (1798) for his secondary thesis (OT-61-02), to his analysis of Kant’s “What Is Enlightenment?” (1784) in his last years. There Kant’s thought will play an important role in the articulation of Foucault’s ethical vision, and we will take it up in our final chapter. It is enough for us to note here that he is already using Kantian language in his thinking about power.10

The movement of Foucault’s analysis here is from the micro level to the macro, from the molecular to the everyday—from (1) specific, individual force relations through (2) their processes of transformation and (3) the networks or systems that their interplay produces, to (4) their larger, strategic manifestations in the state, the law, and other hegemonies, such as ownership of the means of production. In the end-forms that Foucault identifies here, we should recognize the three traditions of analysis that Foucault earlier criticized as partial and inadequate (liberalism, psychoanalysis, Marxism). Even though each of these strategies may be unidimensional, the networks of power taken in sum are multidimensional and cannot be reduced to only one strategic mode, be it juridico-discursive or something else. Foucault’s point is that while they may have adequately described some particular strategy (or terminal form) of power, each approach fails to grasp the fundamental form or operation of power at the molecular level. So if, following Hobbes in order to challenge him, we are to understand social relations in terms of power, Foucault’s approach will be more effective.

In short, it is a question of orienting ourselves to a conception of power which replaces the privilege of the law with the viewpoint of the objective, the privilege of prohibition with the viewpoint of tactical efficiency, the privilege of sovereignty with the analysis of a multiple and mobile field of force relations, wherein far-reaching, but never completely stable, effects of domination are produced. . . . And this, not out of speculative choice or theoretical preference, but because in fact it is one of the essential traits of Western societies that the force relationships which for a long time had found expression in war, in every form of warfare, gradually became invested in the order of political power. (1976ET, 102)

What we have discussed so far provides only a basic framework—a set of theoretical presuppositions that constitutes the heart of Foucault’s theory of power. There are significant elements of this theory that we have not yet discussed, such as the relationship between power and knowledge and between power and “the subject.” (We’ll take these up in the next chapter.) And many important aspects of Foucault’s analysis of power will evolve over the next decade—for instance, as we will see, he begins by characterizing modern power solely as disciplinary power but then supplements this with an analysis of other forms, like biopower. This basic framework, however, is consistent throughout the theory’s subsequent development and elaboration.11

METHODOLOGICAL GUIDELINES FOR STUDYING POWER

To review, we have unpacked a dozen or so dense pages at the heart of La volonté de savoir. What we’ve seen so far of Foucault’s theory of power, however, is general and formal—it is of little value for engaged social analysis. David Halperin even criticizes Foucault’s presentation here as dogmatic:

Volume One [of Foucault’s History of Sexuality], for all its admittedly bright ideas, is dogmatic, tediously repetitious, full of hollow assertions, disdainful of historical documentation, and careless in its generalizations . . . (Halperin 1995, 5, quoting Halperin 1986, 277; his brackets).

Halperin’s criticism, while not entirely inaccurate, is unfair and fails to consider the book in the context of Foucault’s broader investigations.¹² Foucault’s analysis of power as presented here is not a series of “hollow assertions” but a “grid of intelligibility” (1976ET, 93) that entails several methodological principles—methodological precautions—that guide his historical investigations. Second, it is based upon an extensive empirical foundation—and much of the “historical documentation” for Foucault’s claims here is given in other published works; additional support was intended to be published in subsequent, unrealized volumes.

In his 1976 Collège de France course (given in the months before La volonté de savoir was published), Foucault delineates five methodological precautions for his study of power, sovereignty, and domination. First, Foucault notes that “our object is . . . to understand power by looking at its extremities, at its outer limits at the point where it becomes capillary” (CdF76ET, 27), in other words, to study power by observing the infinitesimal, local, micropowers that are its most molecular manifestations rather than by looking at global terminal forms. Second, do not “ask the question (which leads us, I think into a labyrinth from which there is no way out): So who has power?” (CdF76ET, 28). This question is a trap, because power is not possessed but exercised. “The goal was, on the contrary, to study power by looking . . . at the point where it relates directly and immediately to what we might, very provisionally, call its object, its target, its field of application . . .” (CdF76, 28). Again, avoid mistakes based on erroneous theories of power and look to local interactions to see its initial effects.

The third and fourth of Foucault’s rules reflect his theory’s conception of how the infinitesimal nodes combine and interact in a matrix to create larger effects of domination or sovereignty. Third:

Do not regard power as a phenomenon of mass and homogeneous domination. . . . Power must, I think, be analyzed as something that circulates, or rather as something that functions only when it is part of a chain. . . . Power is exercised through networks, and individuals do not simply circulate in those networks; they are in a position to both submit to and exercise this power. (CdF76ET, 29)

The circulation of power, exercised by and on individuals, will serve, at least in part, to constitute these individuals as subjects. But it will constitute larger social patterns, strategies, and institutions. Thus the fourth rule:

We should make an ascending analysis of power, or in other words begin with its infinitesimal mechanisms, which have their own history, their own trajectory, their own techniques and tactics, and then look at how these mechanisms of power, which have their solidity and, in a sense, their own technology, have been and are invested, colonized, used, inflected, transformed, displaced, extended, and so on by increasingly general mechanisms and forms of overall domination. (CdF76ET, 30)

The fifth methodological guideline that Foucault delineates here has to do with the relation between power and knowledge—a relation that we will articulate in more detail shortly: “the delicate mechanisms of power cannot function unless knowledge, or rather knowledge apparatuses, are formed, organized, and put into circulation” (CdF76ET, 33–34).

The common denominator in these five precautions is that one must not misinterpret power by taking one terminal form (sovereignty, prohibition, etc.) as its general form.

To sum up these five methodological precautions, let me say that rather than orienting our research into power toward the juridical edifice of sovereignty, State apparatuses, and the ideologies that accompany them, I think we should orient our analysis of power toward material operations, forms of subjugation, and the connections among and the uses made of the local systems of subjugation on the one hand, and apparatuses of knowledge on the other. (CdF76ET, 34)

These guidelines thus serve to allow one to interpret historical and sociological data with a new unfiltered lens. And these rules can be adapted to particular contexts—Foucault explicitly does so in his studies of prisons and sexuality.

Discipline and Punish follows four such general rules:

1. Do not concentrate the study of the punitive mechanisms on their “repressive” effects alone. . . . As a consequence, regard punishment as a complex social function.

2. Analyze punitive methods . . . as techniques possessing their own specificity in the more general field of other ways of exercising power. Regard punishment as a political tactic. . . .

3. See whether there is not some common matrix [between the history of penal law and the history of the human sciences] or whether they do not both derive from a single process of “epistemologico-juridical” formation; in short, make the technology of power the very principle both of the humanization of the penal system and of the knowledge of man.

4. Try to discover whether . . . the insertion in legal practice of a whole corpus of “scientific” knowledge, is not the effect of a transformation of the way in which the body itself is invested by power relations. (1975ET, 23–24)

The first two of these rules reiterate what we have seen above: power does not have the general form of “repression”—as the “juridico-discursive” theory would have it—but is exercised in a myriad of specific ways. The various techniques of punishment, and imprisonment in particular, are but a few of the specific micropractices of power in its capillary action. The last two of the rules address the mutually productive relationship between power and knowledge. And this method makes certain new insights possible: as Foucault suggests here, the very humanization of punishment (a movement away from violence inflicted upon the body) may itself be an effect of new technologies of power.

Foucault enumerates four similar rules for his study of sexuality in La volonté de savoir. Here again we see that the investigation is grounded in Foucault’s new theory of power. The first is what he terms a “rule of immanence”: Do not assume that a “sphere of sexuality” exists independently of relations of power.

If sexuality was constituted as an area of investigation, this was only because relations of power had established it as a possible object; and conversely, if power was able to take it as a target, this was because techniques of knowledge and procedures of discourse were capable of investing it. . . . We will start, therefore, from what might be called “local centers” of power-knowledge. (1976ET, 98)

Beginning from these “local centers,” the second and third rules jointly articulate how larger and more complex relations develop. The “rules of continual variations” remind us that “the ‘distributions of power’ and the ‘appropriations of knowledge’ never represent only instantaneous slices taken from processes. . . . Relations of power-knowledge are not static forms of distribution, they are ‘matrices of transformations’” (1976ET, 99). And the third rule, the “rule of double conditioning,” emphasizes that these fluid matrices combine and interact to produce larger patterns and strategies.

No “local center,” no “pattern of transformation” could function if, through a series of sequences, it did not eventually enter into an over-all strategy. And inversely, no strategy could achieve comprehensive effects if it did not gain support from precise and tenuous relations serving, not as its point of application or final outcome, but as its prop and anchor point. (1976ET, 99)

So local interactions and global patterns are mutually interwoven. If Foucault initially emphasizes how the larger strategies are grounded in local power relations (here, for example—local centers are anchoring points, not final outcomes, of these strategies—and more generally in his analysis of disciplinary power), his later analysis of “biopower” will bring out how the global strategies can shape and mold local relations.

The fourth rule, which Foucault terms the “rule of the tactical polyvalence of discourses,” underscores the role of resistance as a constituent of relations of power while highlighting the complexities of power’s strategies.

We must make allowance for the complex and unstable process whereby discourse can be both an instrument and an effect of power, but also a hindrance, a stumbling-block, a point of resistance and a starting point for an opposing strategy. Discourse transmits and produces power; it reinforces it, but also undermines and exposes it, renders it fragile and makes it possible to thwart it. In like manner, silence and secrecy are a shelter for power, anchoring its prohibitions; but they also loosen its holds and provide for relatively obscure areas of tolerance. . . . Discourses are tactical elements or blocs operating in the field of force relations; there can exist different and even contradictory discourses within the same strategy; they can, on the contrary, circulate without changing their form from one strategy to another, opposing strategy. (1976ET, 101–102)

Foucault’s treatment of discourses here—as part of the tactics and strategies of power, reversible, not monolithic—contrasts explicitly with the “juridico-discursive” view that he had earlier critiqued.

In La volonté de savoir, as in the 1976 Collège de France course, Foucault notes that “these are not intended as methodological imperatives; at most they are cautionary prescriptions” (1976ET, 98). This caveat is important. It reminds us that these rules are intended, in the first instance, as “prescriptions” to prevent us from falling into familiar but mistaken habits of thinking—interpreting power according to the discounted approaches. On the other hand, neither Foucault’s own approach nor the methodological guidelines that follow from it are “imperatives”; rather, they are tentative proposals, subject to critique and revision in light of the ongoing investigations.

Indeed, these methodological propositions, as well as this theory of power itself, emerged from Foucault’s ongoing self-critique. In the opening lecture of his 1974 Collège de France course, Psychiatric Power, Foucault began to articulate this new conceptualization of power in response to what he then perceived as shortcomings in the analyses of his first work, History of Madness. On the fourth page of that course, he initially observes that

in the asylum, as everywhere else, power is never something that someone possesses, any more than it is something that emanates from some one. Power does not belong to anyone or even to a group; there is only power because there is dispersion, relays, networks, reciprocal supports, differences of potential, discrepancies, etcetera. It is in this system of differences, which have to be analyzed, that power can start to function. (CdF74ET, 4)

Later in the initial lecture, Foucault explicitly identifies this new theoretical approach to power—and to psychiatry in terms of power relations—as a correction to presuppositions in History of Madness. “I accorded a privileged role to what could be called the perception of madness. Here . . . I would like to see if it is possible to make a radically different analysis and if . . . we could start from an apparatus of power” (CdF74ET, 13). Arnold Davidson in his introduction notes that

without having yet developed all of the tools of his own analysis, Psychiatric Power already exhibits Foucault’s awareness of the shortcomings of available conceptions of power. . . .

Psychiatric Power can be read as a kind of experiment in method, one that responds in historical detail to a set of questions that permeated the genealogical period of Foucault’s work. (Davidson 2006, xiv, xviii)

Indeed, we might say that this “experiment in method” led to significant results.

We began this discussion of methodology as the first part of a response to Halperin’s suggestion that Foucault’s analysis of power in La volonté de savoir was dogmatic. The second part of this response is to highlight how Foucault’s condensed and formal presentation in those pages was grounded in and developed through very careful empirical work. This work, throughout his career, included careful analyses of the development of asylums, hospitals, and prisons, as well as of sexuality. La volonté de savoir itself was intended only as the first of six volumes—a schematic overview and a framework for understanding the empirical work of the subsequent volumes. (Of course, that work was not published as such, since Foucault’s project had radically changed by the time the two final volumes actually appeared.) To trace this empirical development in detail, and to elaborate further Foucault’s conception of disciplinary power, we shall turn next to Discipline and Punish.