Knowledge and Power in Foucault

Knowledge and Power in Foucault

David L. Potts

10 mins
February 27, 2011

This commentary is part of The Atlas Society's 1999 online "CyberSeminar" entitled " The Continental Origins of Postmodernism ."

To properly interpret [Michel] Foucault’s statements in The History of Sexuality: An Introduction ("HSI," Vintage Books, 1990; first published 1976), I believe it is necessary to understand his more fundamental philosophical framework, especially his views on epistemology. Accordingly I will begin, in section I, with a brief exposition of his “first philosophy,” as I understand it. Then, in section II, I will attempt to relate Foucault’s epistemology to his doctrine of “power” in HSI. Section III will then interpret Foucault’s views on sexuality in light of the framework developed in sections I and II. Finally, in section IV I raise some additional questions and issues.

To support some of my claims about Foucault I have occasionally made recourse to his writings outside the assigned material. I have tried to keep these to a minimum, but I felt that at least some were necessary to establish a context, especially in epistemology, in which to understand the assigned pages.


A nice summary of Foucault’s basic philosophical framework occurs in his preface to his book The Order of Things ("OT," Vintage Books, 1973; first published 1966).

Any intelligible ordering requires a “system of elements” or grid in terms of which similarities and differences, or any other basis of organization, may be cast (OT xx). For example, when we group objects together or distinguish them from one another on the basis of shared or different properties, it is this system of properties that comprises the grid in question. And, to repeat, there is no organization, no intelligibility, without an antecedent grid.

We make reality intelligible by not just one grid but by a whole complex of grids, arranged in three levels. At the most basic level are “primary codes,” which include grids comprised of language (the words we apply to things), the schemata of sense perception, and assorted cultural practices, techniques, and values (OT xx). These grids are basic in the sense that they determine the “empirical,” which is of course something of a mirage inasmuch as it is determined by a priori grids. The primary codes are transparent, at least at first; we do not experience the spectrum of colors, for example, as a “grid” but as simply there, as an aspect of the way things are.

At the other end of the scale of grids, at the most derivative level, are our schemes of conceptual understanding, our systems of categories, our scientific theories.

In the middle level lies the grid which is the most fundamental and important but also the most difficult to grasp, which Foucault calls the “episteme.” We experience the episteme as the principle of order itself. Is order “continuous and graduated or discontinuous and piecemeal, linked to space or constituted anew at each instant by the driving force of time, related to a series of variables or defined by separate systems of coherences...” (OT xxi)? These are the sorts of questions determined by the episteme. Foucault introduces the idea of the episteme with a tale from the Argentinean writer Borges about a supposed Chinese encyclopedia that classifies animals as: (a) belonging to the Emperor, (b) embalmed, (c) tame, (d) sucking pigs, (e) sirens, (f) fabulous, (g) stray dogs, (h) included in the present classification, (i) frenzied, (j) innumerable, (k) drawn with a very fine camel hair brush, (l) et cetera, (m) having just broken the water pitcher, (n) that from a long way off look like flies. The main thing that strikes us about this taxonomy is that it transcends questions about better or worse, valid or invalid. For it mixes up the very principles from which taxonomies proceed. That is, it is not a possible taxonomy. The Chinese encyclopedia violates our sense of order itself, a sense we aren’t even aware of having until we feel it violated by pathological phenomena such as the Chinese taxonomy.

It is an episteme that supplies us with this sense of order per se. The episteme allows us to critique our grids at both the theoretical and the primary coding levels. The episteme is the “firm foundation” for general theories, that which provides the reference standard on which they are built and by which they are appraised, and which is more true than any theory. In conflicts between theory and empirical evidence, the evidence may have to be revised, but not the episteme. Indeed, it is by reference to the episteme that we can use theory to force revisions to our perceptual judgments. Foucault pictures the episteme as an “epistemological field” or “space of knowledge” (OT xxii) within which competing theories and concepts exist and are evaluated--and without which they could not be. The episteme is the “condition of possibility” of all knowledge.

Foucault’s view is not particularly unique but has clear lines back to Kant.

However, the episteme is not built into our consciousness like the Kantian categories. It is culturally and historically determined. It is said to be “constructed”--and Foucault’s view may be called “constructionism”--although the term is perhaps misleading inasmuch as the construction is neither conscious nor deliberate. Between different cultures, or between different epochs of the same culture, there may be radically different epistemes. Foucault is saying, therefore, that, for example, Borges’s Chinese taxonomy of animals is only impossible within our Western epistemological field and that it is entirely possible that a radically different culture would find the Chinese taxonomy not only possible but reasonable.

As I have already remarked, we are largely unaware of the episteme and to become aware of it is difficult. Yet it is important to try to do so, since it is the episteme that sets the terms for all knowledge and it is the episteme of a culture or epoch that must be grasped to correctly understand the beliefs and practices of that culture or epoch. Foucault calls the project of trying to elicit the episteme of a culture or epoch “archeology.” (The subtitle of OT is “An Archeology of the Human Sciences.”)

In his own work Foucault did not examine foreign cultures but different epochs of Western European civilization, mostly within just the last few hundred years. He believes there have been three distinct epochs during this period. First, the Renaissance, which ended about 1650. Then the “classical” epoch, from 1650 to 1800. Then the “modern” epoch, from 1800 to the present. Further, he thinks the modern episteme has about run its course and is due to be replaced by a new one (OT xxiv), a postmodern epoch.

Thus, Foucault’s view is not particularly unique but has clear lines back to Kant. The contemporary figure to whom I find it most helpful to compare Foucault is Kuhn. For episteme read “paradigm.” For epoch read “period of normal science.” Both authors find it difficult to say there is progress in the history of knowledge--particularly not if progress means that we discover more of the truth. Both deny that there is any “theory neutral” access to the way the world is. Both have difficulty saying exactly what a paradigm/episteme is. And both hold that paradigms/epistemes are largely unconscious and are cultural creations which can suddenly dissolve and be reconstituted. The chief difference is that Foucault’s vision is considerably more grandiose than Kuhn’s. Kuhn restricts himself to the domain of scientific theories, and even that only in fields of science that are relatively well developed. Foucault on the other hand wants to cover all knowledge in any human culture. His concept of “episteme” is accordingly broader than “paradigm.” Whereas a paradigm determines a particular theory, an episteme determines what theories are possible.


Foucault and Kuhn also both hold that the adoption of a given episteme/paradigm is not rational. It could hardly be, since only within an epistemological field can there be standards of rationality. Therefore, an episteme/paradigm being a social construct, the forces governing its change must be social. For Kuhn the change is precipitated by a crisis of confidence in the scientific community and then by the outcome of a sort of popularity contest of competing theories and scientists. Lakatos goes so far as to call Kuhn’s process a matter of “mob psychology” (The Methodology of Scientific Research Programmes, Cambridge UP, 1978, 91). True or false, Kuhn’s is a relatively simple process to describe. Kuhn’s theory after all applies only to the members of a comparatively tiny scientific community, and only to a portion of their lives--their scientific work.

Foucault’s theory, by contrast, is supposed to apply to every member of an entire culture and to every aspect of knowledge and cultural activity. People do not make organized, explicit decisions about social knowledge and publish them in journals. The determinants of an episteme must therefore be pervasive, governing all aspects of social practice and belief from the bottom up. For Foucault, the mechanism that does this is, apparently, power.

I say “apparently” because Foucault never talks about epistemes in HSI, and so I feel slightly uncomfortable speculating about their relation to power, which is the central concept of HSI. Foucault speaks of the mechanisms of power as “a grid of intelligibility of the social order” (HSI 93), which is tantalizing inasmuch as an episteme is also a grid of intelligibility. But doesn’t this conflate two separate issues? For I began by asking what determines a change of episteme, but now I am asking whether the field of power relations might not be the episteme. Of course, the answer to both questions could be the same. The field of power relations could comprise the episteme, and then perforce any reconstitution of that field would comprise a change of episteme.

Whether or not the field of power is the epistemological field, it is evident that “power” is in the driver's seat in the world of HSI and therefore must, almost by default, govern changes of episteme. Power and knowledge are intimately tied. According to the “Rule of Immanence” (HSI 98), for example, knowledge and power are internally related. Sexuality can become an area of investigation only when power establishes it as such, and at the same time power can operate (to control people) through knowledge of sexuality only after sexuality has been constructed by the sciences. Therefore the constructs of knowledge and the strategies of power mutually emerge in and through one another.

For Foucault, the mechanisms of power determine scientific theories, knowledge, and ultimately truth itself.

Again, truth is a “production” “thoroughly imbued with relations of power” (HSI 60). For example, in the nineteenth century, bourgeois society “put into operation an entire machinery for producing true discourses concerning [sex]. Not only did it speak of sex and compel everyone else to do so; it also set out to formulate the uniform truth of sex....Causality in the subject [i.e., human being], the unconscious of the subject, the truth of the subject in the other who knows, the knowledge he holds unbeknown to him, all this found an opportunity to deploy itself in the discourse of sex. Not, however, by reason of some natural property inherent in sex itself, but by virtue of the tactics of power immanent in this discourse” (HSI 69-70). In other words, sex is not “in fact” a particularly significant aspect of human life--facts are ultimately historical constructs. Rather, sex emerged, in the field of power relations at this juncture in history, as an object, discourse around which was encouraged by the then-current tactics of power. Therefore “sexuality”--the knowledge structure designed to embody the truths constructed about sex (HSI 68)--was invented and “deployed” by the bourgeoisie as a political tool (HSI 120-127).

Although I have used intentional language just now, and Foucault uses it constantly, to describe how the mechanisms of power determine scientific theories, knowledge, and ultimately truth itself, it must not be thought that power is wielded by any central, guiding hand. Rather, power is dispersed throughout society in all the multitudinous “tactical” power relations between people as individuals. “And ‘Power,’ insofar as it is permanent, repetitious, inert, and self-reproducing, is simply the over-all effect that emerges from all these mobilities” (HSI 93). In other words, the power of people in the aggregate of classes and institutions emerges as the recurrent patterns and strategies implicit at the individual level. Power grows from the bottom up--from any social relationship in which there is inequality (HSI 93)--and thus suffuses society. Still, it is “intentional” (HSI 94), since strategies of power in the aggregate inherit the aims and objectives of the tactics by which individuals wield power (HSI 95).

Thus large-scale enterprises, such as the sciences of medicine, pedagogy, and economics, and other forms of discourse, are deployed so that knowledge can be wielded as an instrument of power. And this is deliberate although no one is in charge. There are two further points to make about this.

First, the internal relations between power and knowledge do not directly determine what is true or sayable but only make a certain space of argument--er, excuse me, “discourse”--available. For example, although sodomy was recognized, there was no clinical category of homosexuality until the late nineteenth century. The development of homosexuality as an object of medical research and social concern enabled “homosexuals” to be submitted to the legal system, medical and penal institutions, and other apparatuses of power. The same space, however, allowed homosexuals eventually to speak in their own behalf, demand recognition and tolerance, claim normalcy, and so forth (HSI 101-2).

Second, there is a deeper sense in which no one is in charge than the point made earlier about there being no central authority. For Kuhn, the scientific community is small and identifiable, and its problem well defined and of limited scope. Foucault on the other hand is talking about entire civilizations and the entirety of their knowledge and institutions. Therefore, Kuhn’s scientists may lack rational grounds for their decisions, but at least they decide. People in Foucault’s world, by contrast, seem reduced to cogs in a machine, able only to choose from among the culturally available tactics of “power-knowledge.” What adds particularly to this sense of pettiness is the fact that our knowledge choices do not bring us any closer to reality nor represent success in dealing with nature. Foucault repeatedly emphasizes that, in investigating our sexuality we are not learning about ourselves, not if that means probing the whys and wherefores of our inner selves (e.g., HSI 105-6). Sexuality is not something in us to be discovered; it exists only in the discourses we construct as moves in our power struggles. Indeed, the only reality, it would seem, is what we construct in the course of our collective personal struggles in Foucault’s sunshiny world of semi-sadistic relations of power.


Sexuality, then, is a historical construct (HSI 105). What is the history of sexuality? It seems to be rather short, actually, sexuality having only been invented at the end of the eighteenth century. Its roots go back much further, however. Remembering the historical epochs mentioned at the end of section I above, we could say that sex began to be “put into discourse” at the dawn of the classical epoch (HSI 12). Prior to that, people just had sexual relations, of sundry kinds, and although this aspect of life was hardly invisible, it was not considered to hold the keys to human nature, any more than, say, people’s different eating practices.

During the classical epoch, the “great prohibitions” concerning sex were invented (HSI 115): “the exclusive promotion of adult marital sexuality, the imperatives of decency, the obligatory concealment of the body, the reduction to silence and mandatory reticences of language.”

Then, at the beginning of the modern epoch there emerged “a completely new technology of sex” (HSI 116) through the institution of medicine especially, but also through pedagogy and economics. There were “four great strategies” (HSI 103-5). (1) The “hysterization of women’s bodies,” in which women were identified as especially determined by sexuality, and their sexuality was claimed to be critical to the maintenance of children and the family but at the same time prone to pathology. (2) The “pedagogization of children’s sex,” by which Foucault means obsession with child masturbation. (3) The “socialization of procreative behavior,” meaning that population control came to be regarded as a legitimate province of state and societal concern and intervention. (4) The “psychiatrization of perverse pleasure,” in which deviant sexual practices began to be attributed, by the medical community, to underlying pathologies of “sexuality” (e.g., as previously mentioned, the conversion of sodomites into “homosexuals,” “sexual inverts,” etc., with a nature, an etiology, in short a “sexuality” to be investigated and from which society had to be protected).

It is to these modern epoch “technologies” that Foucault seems to be referring most of the time by “sexuality.” (Example: “‘Sexuality’: the correlative of that slowly developed discursive practice which constitutes the scientia sexualis” (HSI 68).) Their utility as instruments by which some people can get others in their power is obvious. It is Foucault’s thesis that that is exactly what they are for. That in fact must be why sexuality is interesting to Foucault to begin with. He is hardly in a position to claim that he is probing human nature, after all. Sexuality “appears rather as an especially dense transfer point for relations of power...Sexuality is not the most intractable element in power relations, but rather one of those endowed with the greatest instrumentality: useful for the greatest number of maneuvers...” (HSI 103). To a person who thinks social/political power is the root cause of everything, sexuality is a field with especially rich soil.

Foucault hints, however, that sexuality may not last forever. What society creates, society can destroy. People of the future may look back in amazement at the importance we attribute to sex (HSI 157-9). Just as the modern epoch may be drawing to a close, so sexuality may be rupturing in the twentieth century as we see a new period of sexual tolerance, the lifting of taboos, and the loosening of other mechanisms of repression (HSI 115; it is worth remembering here that HSI was written in the ’70s).


By way of wrapping up, there are two additional issues I want to briefly raise. First, how accurate is Foucault as history? Is he an accurate historian with a bizarre interpretive overlay or does his philosophy distort his history? I believe the answer is more the former than the latter. I am unfamiliar with most of the historical facts Foucault discusses in HSI, and I have not yet read volume 2, The Use of Pleasure (Vintage Books, 1990; first published 1984), which discusses sex in Greek antiquity. If one may judge from David Halperin’s 100 Years of Homosexuality (Routledge, 1990), however, a Foucaultite tract which has been extremely influential in recent years and which is meant to extend and support Foucault’s claims about sex in ancient Greece, distorting the facts is not the problem.

The problems rather are three. First, the facts are constantly interpreted in terms that must seem bizarre to those who are not true believers in Foucault’s vision. Second, Halperin seems not really interested in any facts not relevant, pro or con, to his philosophical program. Despite cascades of footnotes (literally hundreds per chapter), the book contains no original research on antiquity and very little about Greek sexuality that can’t be found, sans Foucaultite jargon (i.e., relentless use of words like “discourse,” “power,” “inscribed,” “constitute,” “text,” etc.), in K. J. Dover’s excellent Greek Homosexuality (Harvard U.P., 1978). In short this is “activist scholarship” and is therefore less interested in discovery or even in understanding than in making a case. Halperin does not, however, so far as I can see, distort or otherwise play fast and loose with the evidence. Third, there is a tendency to exaggerate discontinuities in history. This is an inevitable consequence of the thesis that history is not continuous, and so something to watch out for in these authors. Kuhn, needless to say, is also criticized for this.

Finally, readers may recall a previous post in which I asked whether Heidegger’s arguments matter. The question bears repeating in the case of Foucault. Does Foucault offer any reasons to suppose, for example, that sexuality is a historical construction? Any reasons, that is, that don’t presuppose his own philosophical framework? He does finally consider, near the end of HSI (152-7), the objection that in studying sexuality people may be trying to understand our underlying sexual nature. His answer is entirely to be expected: “it is precisely this idea of sex in itself that we cannot accept without examination” (HSI 152). That is, the whole idea of there being an underlying sexual nature which it is important to discover is simply false; or rather, it is a “truth” which sexuality constructed! It is after all Foucault’s whole thesis that sexuality is just that knowledge structure according to which sex is a deeply rooted and critically important aspect of “human nature” which must be investigated, understood, and controlled. Sex therefore, as an “underlying reality,” proceeds from sexuality, not vice versa. One might say that this reply is fine if you already accept that sexuality is a social construction, but what if you don’t?

Doesn’t Foucault dodge the issue it seemed he was about to engage, viz., whether sexuality is a social construction, not our conception of our sexual nature as it exists independently of our understanding of it? If so, I have found no other place where Foucault engages this question.

Response by Eyal Mozes

Response by David Ross and others

Does Foucault Have an Argument?

Back to Michel Foucault, The History of Sexuality