Cambridge Catalogue  
  • Help
Home > Catalogue > The Cambridge Companion to Foucault
The Cambridge Companion to Foucault
Google Book Search

Search this book


  • Page extent: 486 pages
  • Size: 228 x 152 mm
  • Weight: 0.88 kg

Library of Congress

  • Dewey number: 194
  • Dewey version: 22
  • LC Classification: B2430.F724 C36 2006
  • LC Subject headings:
    • Foucault, Michel

Library of Congress Record


 (ISBN-13: 9780521840828 | ISBN-10: 0521840821)


Introduction Michel Foucault: A User’s Manual


For all of Foucault’s reservations about modernity and authorship, his writings are typical of those of a modernist author in their demand for interpretation. Any writing, of course, requires some interpretation as part of our efforts to evaluate, refine, extend, or appreciate its achievement or to provide special background that readers outside the author’s culture or historical period may require. But certain authors – in literature, the twentieth-century modernists are among the best examples – present themselves as so immediately and intrinsically “difficult” as to require special interpretative efforts even for those well equipped to understand them. The Wasteland, Cantos, and Finnegans Wake, for example, require explanation, even for culturally and historically attuned readers, in a way that  Paradise Lost, the Essay on Man, and Emma do not. Philosophy, at least since Kant and Hegel, has also provided its share of “intrinsically obscure” writing. Although it may not be easy to formulate the precise difference, it is clear that Wittgenstein, the later Heidegger, and Derrida require a sort of interpretation that Russell, Dewey, and Quine do not.

   Foucault’s penchant, particularly prior to Discipline and Punish, for the modernist obscure explains much of the demand for interpretations of his work. But the need to interpret Foucault sits ill with his desire to escape general interpretative categories. More important, as the enterprise of interpretation is usually understood, interpreting Foucault is guaranteed to distort his thought. Interpretation typically means finding a unifying schema through which we can make overall sense of an author’s works. Interpretations of Foucault, accordingly, single out some comprehensive unity or definitive achievement that is thought to provide the key to his work. They claim to have attained a privileged standpoint that provides the real meaning or significance of his achievement.1

   Interpretation distorts because Foucault’s work is at root ad hoc, fragmentary, and incomplete. Each of his books is determined by concerns and approaches specific to it and should not be understood as developing or deploying a theory or a method that is a general instrument of intellectual progress. In Isaiah Berlin’s adaptation of Archilochus’s metaphor, Foucault is not a hedgehog but a fox.2

   Foucault’s writings tempt us to general interpretation along two primary axes. In the first dimension he appears as a philosophical historian, progressively developing a series of complementary historical methods: an archaeology of discourse in The History of Madness, The Birth of the Clinic, The Order of Things, and The Archaeology of Knowledge; a genealogy of power relations in Discipline and Punish and The History of Sexuality I; and a problematization of ethics in The Use of Pleasure and The Care of the Self. In the second dimension he appears as a historicist philosopher, offering, parallel to his methodological innovations, successively deeper and mutually supporting theories of knowledge, power, and the self. It is natural to combine these two dimensions in an overall interpretation of Foucault’s work as a new comprehensive understanding of human reality supported by new methods of historical analysis.

   One of the most intelligent and interesting general interpretations of Foucault is that of Hubert Dreyfus and Paul Rabinow.3 They present Foucault as developing a “new method” (both historical and philosophical) whereby he “goes beyond” structuralism and hermeneutics. This method they term “interpretative analytics”: analytics because it shares Kant’s critical concern for determining “the sources and legitimate uses” of our concepts; interpretative because it seeks “a pragmatically guided reading of the coherence of the practices” in which the concepts are expressed.4 Dreyfus and Rabinow agree that interpretative analytics “is not a general method,” since it recognizes that it itself is practiced within a historically contingent context and that its practitioner “realizes that he himself is produced by what he is studying; consequently he can never stand outside it.”5 Nonetheless, Dreyfus and Rabinow do see Foucault’s method as occupying a privileged position on the contemporary scene:

since we still take the problems of our culture seriously…we are drawn ineluctably to a position like Foucault’s. In a sense, it is the only position left that does not regress to a tradition that is untenable.…This does not mean that one is forced to agree with Foucault’s specific diagnosis of our current situation.…But…some form of interpretative analytics is currently the most powerful, plausible and honest option available.6

Dreyfus and Rabinow offer a general interpretation in that they read the whole of Foucault’s work as directed toward the development of a single historico-philosophical method that has a privileged role in contemporary analyses. Even if this method is not ahistorically universal, they clearly present it as Foucault’s definitive achievement for our time: the preferred instrument for current social and cultural analysis.

   I am uneasy with this and other general interpretations of Foucault because they deny the two things that, to my mind, are most distinctive and most valuable in his voice: its specificity and its marginality. It is striking that Foucault’s books hardly ever refer back to his previous works. The Birth of the Clinic never mentions The History of Madness, even though the two books share the common ground of the history of medicine in the nineteenth century; The Order of Things describes the episteme of the Classical Age with scarcely a hint of the author’s previous extensive dealings with that period in The History of Madness and The Birth of the Clinic; The History of Sexuality I, for all its conceptual, methodological, and topical similarities to Discipline and Punish, refuses to acknowledge any connection; and The Use of Pleasure and The Care of the Self, although formally the second and third volumes of a history of sexuality, acknowledge the first volume only to note their divergence from it. This lack of self-citation is not mere coyness. Each of Foucault’s books strikes a specific tone that is muffled and distorted if we insist on harmonizing it with his other books. In examining psychiatry, medicine, the social sciences, and other contemporary disciplines, his goal was always to suggest liberating alternatives to what seem to be inevitable conceptions and practices. But his analyses are effective precisely because they are specific to the particular terrain of the discipline he is challenging, not determined by some general theory or methodology. As we shall see, Foucault does not hesitate to construct theories and methods, but the constructions are always subordinated to the tactical needs of the particular analysis at hand. They are not general engines of war that can be deployed against any target. This is why each of Foucault’s books has the air of a new beginning.

   General interpretations of Foucault suppress his marginality by presenting his work as the solution to the problems of an established discipline or as the initiation of some new discipline. This ignores the crucial fact that disciplines are precisely the dangers from which Foucault is trying to help us save ourselves. His attacks are on the apparently necessary presuppositions (such as that madness is mental illness, that imprisonment is the only humane punishment for criminals, that ending sexual repression is the key to human liberation) that define disciplines. Therefore, they can be launched only from the peripheral areas where the defining assumptions begin to lose hold. To present Foucault as working within an established discipline or, even worse, as attempting to found one himself is to contradict the basic thrust of his efforts.7

   Resisting our inclination to general interpretation accords not only with the direction of Foucault’s work but also with some of his own explicit pronouncements. For example, in “What Is an Author?” and elsewhere,8 he challenges the unifying categories (author, works, etc.) presupposed by general interpretation. And in an anonymous interview, “The Masked Philosopher,” he describes his dream that books would not be subjected to totalizing judgments but would rather find “a criticism of scintillating leaps of the imagination [that] would not be sovereign or dressed in red [but would] catch the sea-foam in the breeze and scatter it.”9

   On the other hand, it is only fair to note that Foucault himself was prone to providing overall interpretations of his work. Thus, in 1969 he characterizes all his previous books (The History of Madness, The Birth of the Clinic, and The Order of Things) as “imperfect sketches” of the archaeological method for analyzing discursive formations that is explained in The Archaeology of Knowledge.10 But then in 1977 he says, “When I think back now, I ask myself what else was I talking about in [The History of Madness] or The Birth of the Clinic, but power?”11 By 1982 he is saying: “it is not power, but the subject, which is the general theme of my research.”12

   The ambivalence of Foucault’s view of his work is particularly apparent in a discussion at the end of The Archaeology of Knowledge. Foucault imagines a critic who suggests that archaeology is

yet another of those discourses that would like to be taken as a discipline still in its early stages…yet another of those projects that justify themselves on the basis of what they are not,…disciplines for which one opens up possibilities, outlines a programme, and leaves the future development to others. But no sooner have they been outlined than they disappear together with their authors. And the field they were supposed to tend remains sterile forever.13

Foucault first responds with forthright denials of “scientific” pretensions:

I have never presented archaeology as a science, or even as the beginning of a future science.…The word archaeology is not supposed to carry any suggestion of anticipation; it simply indicates a possible line of attack for the analysis of verbal performances.14

But he then goes on to emphasize the close connection of archaeology to current sciences. They are, he says, a primary object of archaeological analysis; its methods are closely related to those of some sciences – especially generative grammar; and its topics are closely correlated to those of disciplines such as psychoanalysis, epistemology, and sociology. Foucault even suggests that a “general theory of productions” would, if developed, be an “enveloping theory” for archaeology. He goes on to say that he is perfectly aware that “my discourse may disappear with the figure that has borne it so far.” But he also says, “It may turn out that archaeology is the name given to a part of our contemporary theoretical conjuncture” and suggests as one possibility that “this conjuncture is giving rise to an individualizable discipline, whose initial characteristics and overall limits are being outlined here.”15 It is clear that, at least when he wrote The Archaeology of Knowledge, Foucault was tempted by the hope of becoming the founder of a new discipline.

   General interpretations of Foucault are tempting because, for all their distortion, they can put us on to some important truths. My suggestion is not that we renounce them, but that we regard them as nonunique and developed for specific purposes. (Had Foucault lived, he would have surely continued to produce them as an accompaniment to his ever-changing specific concerns.) Without becoming obsessed with finding the general interpretation that will give us the “final truth” about Foucault’s work, we should be prepared to use a variety of such interpretations to elucidate, for particular purposes, specific aspects of his writings. For example, the methodological axis of interpretation, which sees Foucault moving from archaeology through genealogy to ethics, is useful for appreciating his contribution to historical method and hence relating his work to the Annales school, French history and philosophy of science, the “new historians,” disputes about the role of events in history, and so on.16 The topical axis of interpretation, which views him as starting with the study of knowledge, coming to see the inextricable connection of knowledge to power, and finally subordinating both to a primary concern with the self, shows how to read Foucault as contributing to recent discussions in the epistemology and philosophy of science (particularly social epistemology and “postmodern” philosophy of science) and in social theory.17

   It is, however, less risky and even more profitable to regard Foucault as an intellectual artisan, someone who over the years constructed a variety of artifacts, the intellectual equivalents of the material objects created by a skilled goldsmith or cabinetmaker. We need to take account of the specific circumstance occasioning the production of each artifact in order to understand and appreciate it. But each artifact may also have further uses not explicitly envisaged by its creator, so that we also need to examine it with a view to employment for our own purposes. Foucault was particularly adept at crafting three sorts of intellectual artifacts: histories, theories, and myths. As an alternative to a general interpretation of his work, I propose to discuss some examples of these productions.


Foucault wrote book-length histories of madness, clinical medicine, the social sciences, the prison, and ancient and modern sexuality. Although much has been made of his archaeological and genealogical methods, his approach to each topic is driven much more by the specific historical subject matter than by prior methodological commitments. “Archaeology” and “genealogy” are primarily retrospective (and usually idealized) descriptions of Foucault’s complex efforts to come to terms with his historical material. His “discourse on method,” The Archaeology of Knowledge, is a reconstruction, with a not insignificant amount of trimming and shaping, of what went on in the three histories that preceded it.18

   An appreciation of Foucault’s histories requires locating them on a finer grid than that defined by the two dimensions of archaeology and genealogy. I propose tracking Foucaultian histories along four dimensions: histories of ideas, histories of concepts, histories of the present, and histories of experience.

   Although Foucault’s explicit mentions of standard history of ideas are at best disdainful, we need to keep in mind that he frequently offers the sorts of textual interpretations and comparisons that are the mainstay of orthodox history of ideas. Central to The History of Madness, for example, is his reading of the passage in the Meditations in which Descartes dismisses the possibility that he is mad as a grounds for doubt.19 Similarly, crucial claims of The Order of Things are based on interpretations of scientific and philosophical texts from Paracelsus and Aldrovandi to Smith and Kant. Moreover, despite Foucault’s particular disdain for historians of ideas’ concern with attributions of originality, key points of his argument in The Order of Things depend on showing that, for example, Cuvier rather than Lamarck developed the basic framework for evolutionary theory and that Marx’s work in economics is really just a variant on Ricardo’s. Much of Foucault’s last two volumes, on ancient sexuality, also need to be read and evaluated by the norms of standard interpretative history of ideas. On at least one important level, they are simply explications of texts by Galen, Xenophon, and Plato, among others.

   Much of Foucault’s historiography falls in the genre of “the history of concepts,” as that had been understood by his friend and mentor Georges Canguilhem. This approach flows from an insistence on the distinction between the concepts that interpret scientific data and the theories that explain them. By contrast, the standard Anglo-American view (shared by both positivists such as Hempel and their critics such as Kuhn) is that theories are interpretations of data and therefore define the concepts in terms of which data are understood. On Canguilhem’s view, concepts give us a preliminary understanding of data that allows us to formulate scientifically fruitful questions about how to explain the data as conceptualized. Theories then provide different – and often conflicting – answers to these questions. For example, Galileo introduced a new concept of the motion of falling bodies (in opposition to Aristotle’s); then he, Descartes, and Newton provided competing theories to explain the motion so conceived. As long as concepts are regarded as functions of theories, their history will be identical with that of the development of theoretical formulations. But for Canguilhem concepts are “theoretically polyvalent”; the same concept can function in quite different theoretical contexts. This opens up the possibility of histories of concepts that are distinct from the standard histories that merely trace a succession of theoretical formulations.

   Canguilhem demonstrated the power of this approach in his history of the concept of reflex action.20 The standard view is that this concept was first introduced by Descartes in his Traité de l’homme. Such a view is natural if we do not make Canguilhem’s distinction between concepts and theories. The concept of reflex action is at the heart of modern mechanistic theories in physiology, and Descartes was the first to describe reflex phenomena and try to account for them mechanistically. But Canguilhem is able to show that, even though Descartes anticipates modern physiology in offering a mechanistic explanation of the reflex, he does not in fact have, either explicitly or implicitly, the modern concept of the reflex. His explanation is of the phenomenon conceived quite differently than modern physiology conceives it. By contrast, Canguilhem shows, the modern concept of the reflex is fully present in the (distinctly nonmodern) vitalistic physiology of Thomas Willis.21

   Foucault makes a similar use of the history of concepts in The Order of Things when he argues that the Darwinian idea of an evolution of species is implicit in Cuvier but not in Lamarck. He admits that Lamarck’s developmental theory recognizes biological change in a way that Cuvier’s fixist theory does not. But, Foucault argues, it is Cuvier and not Lamarck who introduces the fundamental idea that biological species are productions of historical forces rather than instantiations of timeless, a priori possibilities. Lamarckian “evolution” is merely a matter of living things successively occupying preestablished niches that are quite independent of historical forces, such as natural selection. For Cuvier, however, the fact that species do not change over time is itself a result of the historical forces that have led to their production. Lamarckian change is just a superficial play of organisms above the eternally fixed structure of species; Cuvier’s fixism is a historical stability produced by radically temporal biological processes. Accordingly, Foucault maintains that Cuvier rather than Lamarck provides the conceptual framework that makes Darwin’s theory of evolution possible.

   Of all Foucault’s books, The Birth of the Clinic (published in a series edited by Canguilhem) comes the closest to a pure history of concepts, the concept in question being that of physical illness as it developed from the end of the eighteenth century through the first third of the nineteenth. The Order of Things also makes extensive use of Canguilhem’s approach. Foucault’s accounts of the empirical sciences of the eighteenth and nineteenth centuries are simply histories of the relevant concepts. But The Order of Things also extends and transforms Canguilhem’s method. For Canguilhem concepts correspond to disciplines, and the history of a concept is written within the confines of the relevant discipline. But Foucault links apparently very different disciplines by showing similarities in their basic concepts. He argues, for example, that the Classical empirical sciences of general grammar, natural history, and analysis of wealth share a common conceptual structure that makes them much more similar to one another than any one of them is to its modern successor (respectively philology, biology, and economics). Even more important, Foucault maintains that such philosophical concepts as resemblance, representation, and man pervade all the disciplines of a given period, a view that leads him to the notion of an episteme as the system of concepts that defines knowledge for a given intellectual era.

   These extensions of Canguilhem’s history of concepts transform it by moving to a level where the historian is no longer required to define a discipline in its own terms. As a historian of biology, Canguilhem deals with concepts (such as reflex action) explicitly deployed by contemporary biology. Foucault focuses not only on such first-order biological concepts but also on concepts (such as representation and historicity) that are conditions of possibility for the first-order concepts.

   This analysis of the “intellectual subconscious” of scientific disciplines is precisely Foucault’s famous archaeological approach to the history of thought. Archaeology is an important alternative to standard history of ideas, with its emphasis on the theorizing of individual thinkers and concern with their influence on one another. Foucault suggests (and shows how the suggestion is fruitful) that the play of individuals’ thought, in a given period and disciplinary context, takes place in a space with a structure defined by a system of rules more fundamental than the assertions of the individuals thinking in the space. Delineating the structures of this space (the goal of the archaeology of thought) often gives a more fundamental understanding of the history of thought than do standard histories centered on the individual subject (which Foucault disdainfully labels “doxology”).

   Many of Foucault’s histories fall under the category he designated “history of the present.” Of course history is, by definition, about the past, but Foucault’s histories typically begin from his perception that something is terribly wrong in the present. His motive for embarking on a history is his judgment that certain current social circumstances – an institution, a discipline, a social practice – are “intolerable.”22 His primary goal is not to understand the past but to understand the present; or, to put the point with more nuance, to use an understanding of the past to understand something that is intolerable in the present. In this sense his characterization of Discipline and Punish as “history of the present” (30–31) applies to all his histories.

   Apart from the paradoxical language, there is really nothing extraordinary in Foucault’s project of trying to understand the present in terms of the past; in one way or another, this is what most historians are up to. But Foucault reverses a standard polarity of this enterprise. Whereas much traditional history tries to show that where we are is inevitable, given the historical causes revealed by its account, Foucault’s histories aim to show the contingency – and hence surpassability – of what history has given us. Intolerable practices and institutions present themselves as having no alternative: How could we do anything except set up asylums to treat the mentally ill? How deal humanely with criminals except by imprisoning them? How attain sexual freedom except by discovering and accepting our sexual orientation? Foucault’s histories aim to remove this air of necessity by showing that the past ordered things quite differently and that the processes leading to our present practices and institutions were by no mean inevitable.

printer iconPrinter friendly version AddThis