Niklas Luhmann once affirmatively observed: “The ownership of knowledge confers authority. This person can teach the others. The individual who claims authority, must therefore base it on knowledge. The knowledge function and political function cannot be separated, in the final analysis” (Luhmann 1990: 149).Footnote 1

In this observation, he is not alone. Just as much as the conclusions of the leading Social Democrat, Wilhelm Liebknecht ([1872] 1891) or the classic investigation of the oligarchical tendencies of the Social Democratic Party by Robert Michels ([1911] 1949], Michel Foucault’s (e.g. [1975] 1977: 32) major work is designed to display the complicity of knowledge in disciplining, governing and repressing people. Knowledge and power, it would appear at least initially, are like Siamese twins: “Power produces knowledge; […] power and knowledge directly imply one another; […] there is no power relation without the correlative constitution of a field of knowledge, nor any knowledge that does not presuppose and constitute at the same time power relations.”

Furthermore, one should not expect, Foucault goes on to explain, that the conflation of power and knowledge can somehow eventually be disentangled: “Knowledge and power are each an integral part of the other, and there is no point in dreaming of a time when knowledge will cease to be dependent on power […]. It is not possible for power to be exercised without knowledge, it is impossible for knowledge not to engender power” (Foucault 1977: 15).

The question Foucault’s statement leaves open is of course: who is exercising power? Are knowledge and political authority inseparable and do they always benefit only a particular social stratum, namely the stratum of the powerful?

Power and Government

Michel Foucault assumes a highly critical stance toward the powerful segments of society in his archeology and genealogy of social problems and topics – such as madness, clinical medicine, the penal system and sexuality, which blend the problems of power and knowledge. Yet he takes for granted that each of the disciplinary sciences associated with these organized activities is “successfully” implicated in modern society’s attempt to control and shape its citizens (“gouvernementalité”).Footnote 2 And although Foucault enlarges more traditional conceptions of power to encompass all social relations, more often than not, the addressees of his critical interventions seem to be the large social institutions representing the loci of power.

What on the surface had been primarily a political and legal matter becomes invested with newly fabricated dimensions of scientific knowledge. Practical or political knowledge, like power, is a context-specific phenomenon. One needs to examine the socio-political practices in which knowledge is embedded. Consequently, Foucault ([1969] 1972: 194; see also Foucault [1975] 1977: 305) formulates his “knowledge-guiding interests” as follows:

Instead of analyzing this knowledge – which is always possible – in the direction of the episteme that it can give rise to, one would analyze it in the direction of behavior, struggles, conflicts, decisions, and tactics. One would thus reveal a body of political knowledge that is not some kind of secondary theorizing about practice, nor the application of theory […]. It is inscribed from the outset, in the field of different practices in which it finds its specificity, its functions, and its networks of dependencies.

Practical knowledge embedded in various discursive activities within different institutional settings is successively described by Foucault as a “political anatomy, a political economy, a discursive formation, a discursive disposition, and a political technology. Repressions and prohibitions, exclusions and rejections, techniques and methods bring individuals under surveillance” (Lemert and Gillan 1982: 60).

However, the power joined to knowledge and exercised by the modern state as “the political form of centralised and centralising power” (Foucault 1981: 227) in labeling, masking, censoring, segregating, prohibiting, normalizing, surveying and oppressing (for example, Foucault [1975] 1977: 304) its subjects by the powerful state is not total, but not quite enabling either. The conditions for the possibility of recalcitrance shown by the subject are not completely displaced; as a result, there are limits to the power exercised by state agencies (cf. Foucault 1980: 119). Although overly efficient, the knowledge/power axis, according to Foucault, is more complex and leaves room for enabling results among those that are oppressed. However, resistance as a reaction to the seemingly well-functioning social controls remains a blind spot in Michel Foucault’s approach.Footnote 3

The primary impression of Foucault’s theorizing is that of the enormous authority of knowledge embedded in the discursive practices of the state. Such a conception of the omnipotence of the state resonates strongly with much of the historical literature on empire that portrays the “overwhelming power of the early modern state in its relationship with subject people, be they members of lower social classes, bureaucrats, and administrators, or indigenous populations” (Edwards et al. 2011: 1399). This impression cannot be set aside. Emphasizing, as Foucault does, the efficacy of knowledge attached to power and the extent to which it works by forcing its imprints on subjects and society, it is difficult, if not impossible, to allow or account for the possibility of societal discontinuities. How does social change come about? Power is enabling, if only because it has unanticipated consequences, but its agency for those who are subject to power appears to pale against the background of its productivity in stabilizing social figurations.

In the face of this conception of the overwhelming power of normalization Alain Touraine ([1992] 1995: 168) asks with good reason: “Why reduce social life to the mechanism of normalization? Why not accept that cultural orientations and social power are always intertwined, and that knowledge, economic activity and ethical conceptions therefore all bear the mark of power, but also the mark of opposition to power?” (see also Megill 1985: 140–252).

Power and Discourse

Let us take a brief look at the genealogy of Foucault’s thinking about the relationship between knowledge and power, domination and agency. In 1966, Michel Foucault published a book that was to become a philosophical bestseller in France, Les mots et les choses (translated as The Order of Things in 1970). In this book he examined biology, economics, and medicine, noting a fundamental transformation of each of them towards the end of the eighteenth century: “A monetary reform, a banking custom, a commercial practice can be rationalized, developed, maintained or dissolved each according to its appropriate form; they are always founded upon a certain knowledge: a dark knowledge that does not appear in itself in a discourse, but the necessities of which are precisely the same for abstract theories and speculations without any seeming connection to reality” (quoted in Paras 2006: 23).

This “dark knowledge” that informed all discourses he was to call épistémè, something that was unique to a historical period. During the eighteenth century the dominating épistémè was a tabular representation of reality. Anything that existed could be represented in tables.Footnote 4 However, there is no trace of the subject who assembles these tables, a practice Immanuel Kant was to ridicule as “tabular reason” (tabellarischer Verstand). Foucault’s conclusion was that there was no place for “man” in such an épistémè. Only with the advent of the modern discourse “man” moved to the center of the discourse. The implication was, at least for Foucault, that with the end of the modern discourse man will again disappear “like a face in the sand at the edge of the sea”, as he famously put it in the last sentence of The Order of Things.

In his next book, The Archaeology of Knowledge ([1969] 1972), Foucault develops a programmatic statement about the analysis of what he calls discursive formations: “Whenever one can describe, between a number of statements, such a system of dispersion, whenever, between objects, types of statement, concepts, or thematic choices, one can define a regularity […] we will say […] that we are dealing with a discursive formation […]” (Foucault [1969] 1972: 41). Here he outlines his understanding of how knowledge and power, and discourses and objects in society relate to each other.

Drawing on the example of madness he asks what it is that makes a discourse unified. He rejects the idea that there are objects out there that could be described more or less accurately through scientific language, as there cannot be “madness itself, [with] its secret content, its silent, self-enclosed truth.” Rather, “mental illness was constituted by all that was said in all the statements that named it” (Foucault [1969] 1972: 35). One cannot, therefore speak of a discourse “concerning madness”.

The same logic applies to specific instances of madness such as neurosis or melancholia, so that it does not make sense either to speak of a “discourse concerning neurosis”, or a “discourse on melancholia”. These objects come into being only through the discursive activities around the objects – the objects are in turn constituted by the discourse. Again, like in the above statement about human subjects, it is discourse that gives rise to something called objects.

Consequently Foucault rejects the idea that there could be a “prediscursive subjectivity”, or experienced subjectivity that is “murmuring beneath the surface” and then taken up by scientific observation and research. Very much in agreement with the structuralists who were dominant in France at the time, Foucault asserts, “before all human existence, all human thought, […] there must already be a knowledge, a system, that we are rediscovering” (quoted in Paras 2006: 29). For him, the history of knowledge was “the unfolding of an anonymous process; a process of the formation and transformation of bodies of statements according to isolable rules” (Paras 2006: 34–5).Footnote 5

Foucault points out that this choice of terminology is for heuristic reasons, convenience, and to demarcate against the connotations of other established terms, such as science, ideology, theory or domain of objectivity (cf. Foucault [1969] 1972: 41). One of the crucial tasks is to map the surface of the emergence of an object of discourse. It should be noted that it is not an object that is stable and therefore gives rise to a stable or unified discourse. It is the discursive practice that constitutes the object. Such a mapping will “show where these individual differences […] will be accorded the status of disease, alienation, anomaly, dementia, neurosis or psychosis […]” (Foucault [1969] 1972: 45). He continues: “These surfaces of emergence are not the same for different societies, at different periods, and in different forms of discourse.”Footnote 6 There are two more elements that are necessary for a discursive formation: specialized institutions (Foucault designates them “authorities of delimitation”) and what he calls grids of specification (for example, the body, soul, or life history).

In the early 1970s Foucault took on board several political and theoretical concerns of some of his Neo-Marxist friends and colleagues. The resulting orientation was described as a move from archeology to genealogy. Here he fully engages with the problem of power on a theoretical level. In a discussion with his inspirator and interlocutor Gilles Deleuze, he told him: “We still don’t know what power is […]. And Marx and Freud are perhaps insufficient to help us to know this deeply enigmatic thing, at once visible and invisible, present and hidden, invested everywhere that is called power. The theory of the State and the traditional analysis of the State apparatus do not, undoubtedly, exhaust the field of exercise of power’s functioning” (quoted in Paras 2006: 64).Footnote 7

Discourses are linked to power and depend on knowledge. Foucault tried to escape a traditional conceptualization which interpreted power either in terms of violence or in terms of persuasion and ideology: “Now, power is not caught in this dilemma: either to be exercised by imposing itself by violence, or to hide itself, and to get itself accepted by holding the chatty discourse of ideology. In fact, every point of exercise of power is at the same time a site of transformation: not of ideology, but of knowledge. And on the other hand, every established knowledge permits and assures the exercise of power” (quoted in Paras 2006: 113).

What we see here is a concept of both power and knowledge that acknowledges their generative potential (for good or bad). Foucault says that it does not make sense to see power as separate from or opposed to knowledge. Knowledge enables power to be exercised and power transforms knowledge. There is no basis for a distinction between knowledge and ideology; there is no true knowledge behind the veil of ideology.

Power and Science

One of Foucault’s central theses was that modern industrial society emerged with the emergence of the social and human sciences. These provided the knowledge base for disciplining the workforce into a system that depended on their collaboration. The prison and the hospital, surveillance and madness were the crucial places and issues to analyze. The books, entitled Birth of the Clinic (Naissance de la clinique, 1963) and Discipline and Punish (Surveillir et punir: Naissance de la prison, 1975), are programmatic. The term discipline is understood in a double sense: on the one hand, it is the practice of disciplining workers and citizens; on the other hand it is the discipline-based knowledge that enables power holders to discipline workers, which means that it does not make sense to separate knowledge and power. Both are fused, one cannot be exercised without the other. There is no truth that speaks to power, only knowledge that has been created by the powerful to serve their purposes. Admittedly, this is a slight overstatement not necessarily borne out by Foucault’s texts, although he did make statements to this effect in comments and interviews. Foucault preferred to speak of the simultaneous emergence of concepts and practices, or discourses. Nevertheless, the constellation he describes by the power/knowledge nexus has strong functionalist overtones, something he would later abandon.

Eventually Foucault was to replace the concept of power/knowledge with the concept of government, a move that, according to some observers, was connected to a more general reorientation towards recognizing subjectivity and agency (Paras 2006). Using the power/knowledge terminology during the early 1970s, Foucault was led to an “extremist denunciation of power […] hence the question of government – a term Foucault gradually substituted for what he began to see as the more ambiguous word, ‘power”’ (Pasquino 1993: 79, quoted in Dean 2001: 325).

Thus he came to see power as a creative force that enables subjects to act upon each other in flexible relationships. He reserves the term domination for the repressive, unidirectional and rigid form of power. Domination leads to a limitation of possible action since the margin of liberty is extremely narrow. What Michel Foucault now calls domination is akin to what Max Weber and others had called power.

In an interview with Paul Rabinow we find a description by Foucault of his own work. He says that to a certain extent he tries to analyze the relations among science, politics, and ethics or, more precisely, “how these processes may have interfered with one another in the formation of a scientific domain, a political structure, a moral practice” (Foucault 1984c: 386). Once again he refers to the example of psychiatry:

I have tried to see how the formation of psychiatry as a science, the limitation of its field, and the definition of its object implicated a political structure and a moral practice: in the twofold sense that they were presupposed by the progressive organization of psychiatry as a science and that they were also changed by this development. Psychiatry as we know it could not have existed without a whole interplay of political structures and without a set of ethical attitudes (Foucault 1984c: 386-387).

He then goes on to explain that he followed the same methodological principle in his studies on madness, delinquency and sexuality, i.e., “the establishment of a certain objectivity, the development of a politics and a government of the self, and the elaboration of an ethics and a practice in regard to oneself” (ibid.). He calls these three dimensions “fundamental elements of any experience” – namely (1) a game of truth, (2) relations of power and (3) forms of relation to oneself and to others. Foucault makes the point that prevailing accounts were emphasizing only one dimension while the other two were screened out. With psychiatry the organization of knowledge took center stage, crime was seen as a problem for political intervention and sexuality was designed above all as an ethical problem. “Each time I have tried,” Foucault says, “to show how the two other elements were present, what roles they played, and how each one was affected by the transformations in the other two” (Foucault 1984c: 387–8).Footnote 8

Power and Knowledge as Non-identical

A further transformation in Foucault’s interpretation of the relation between knowledge and power occurs in his later work, specifically in his 1977 lectures at the Collège de France, and under the influence of the work of the nouveaux philosophes and his observations about the revolution in Iran. Michel Foucault’s reflections now center on the subject and the changing forms of subjectivity in modern societies (see Foucault 1982: 211–212). The self-determining subject that has gained in significance also for him as a person now influences his professional point of view and thus allows him to let go of the strict symbiosis of power and knowledge (see Paras 2006: 105–116).

Contrary to what many commentators continue to say about Foucault’s reflections on knowledge and power, that is, treating them as an essential identity in Foucault’s work (e.g. Kusch 1991), it is exactly the nature of the linkage between knowledge and power that remained his decisive research problem up until a year before his death, as his response in an interview reveals:

You have to understand that when I read – and I know it has been attributed to me – the thesis ‘Knowledge is power’, or ‘Power is knowledge’, I begin to laugh, since studying their relation is precisely my problem. If they were identical, I would not have to study them and I would be spared a lot of fatigue as a result. The very fact that I pose the question of their relation proves clearly that I do not identify them (Foucault in Kritzman 1988: 43)

So how can we summarize Foucault’s theorization of knowledge and power, their relationship, and the development of his thinking about these terms? For Foucault, power is a productive force, a force that shapes subjects and subjectivity. It is virtually everywhere, and “reaches into the very grain of individuals, touches their bodies and inserts itself into their actions and attitudes, their discourses, learning processes and everyday lives” (Foucault 1980: 30).

The productivity of power rests on knowledge, and Foucault, too, stresses the capacities that are conveyed through knowledge. Knowledge is thus depicted as a deeply social – and only social – entity, as something that shapes and emerges from discourse, not as that which corresponds to an objective truth. Truth is nothing but a tool for the interest-driven execution of power. It is easy to see how such a conception would lead many observers to conclude that there is in fact an intimate conceptual bond between knowledge and power, as they are functions of the social, of the will to power, and always already political. To govern means to “structure the potential field of the actions of others” (Foucault 2007: 97).

Still, the close connection of knowledge, power and government is by no means a clear-cut process or confined to governmental agencies or the large institutions. In any society, there are “multiple forms and loci of governing” (Foucault 2007: 101, our translation), ultimately suggesting that power and domination reside “outside the institutions” (ibid.: 99).

What emerges in his later writings is that Foucault eventually acknowledges the different ways in which knowledge-as-power may be used, hinting towards the agency inherent in employing what one knows for purposes beyond “normalization”:

I don’t see where evil is in the practice of someone who, in a given game of truth, knowing more than another, tells him what he must do, teaches him, transmits knowledge to him, communicates skills to him. The problem is rather to know how you are to avoid in these practices – where power cannot not play and where it is not evil in itself – the effects of domination […] (Foucault [1984a] 1987: 129).

Relations of power may thus still be emancipatory occurrences. The phrase “knowledge is power” is, then, as much Foucauldian as it is Baconian, but holds a quite distinctive meaning; rather than supposing that those who wield power do so by their privileged knowledge of the truth, this relation is inversed, rendering knowledge a deeply social category: those who have the capacity to claim what is true (knowledge), have a claim to power. In this light, Foucault may be read not only as a historian and philosopher, but fundamentally as a proponent of the sociology of knowledge. Knowledge and power mutually create each other, are intimately tied up, but can, for exactly this reason, never be the same.

Judging by the gradual, if decisive change in the development of his theory, the statement that in modern societies, liberty is a daughter of knowledge may no longer be seen as a utopian perspective (cf. Stehr 2016).