Michel Foucault: Beyond Structuralism and Hermeneutics 
by Hubert Dreyfus and Paul Rabinow, with an afterword by [afterword_writer].
Harvester, 256 pp., £18.95, October 1982, 0 7108 0450 4
Show More
Show More

French philosophers become notorious when, deviating from Anglo-American ‘common sense’, they appear to cast aside respect for truth, tradition, reality and reason. Michel Foucault is a case in point, for his books typify the manoeuverings that result. There is a growing body of secondary literature explaining his vagarious development, and the best study so far is a joint effort by two Berkeley scholars, Hubert Dreyfus, a philosopher, and Paul Rabinow, an anthropologist. Foucault himself lends credence to the Dreyfus/Rabinow interpretation by allowing them to include some of his recent unpublished material. There must have been considerable interaction between Paris and California, since the authors frequently indicate points they discussed with him. The reader is given the sense of hearing an ongoing dialogue.

The subtitle, Beyond Structuralism and Hermeneutics, suggests that the authors think Foucault is best understood by placing him in a context that is both historical and philosophical. Twentieth-century Continental philosophy is no longer thought of in terms of systematic programmes like phenomenology or cultural formations like existentialism. The desire for a rigorous, philosophical foundation or an uncommon world-view disappears when philosophy itself is seen as simply another interpretation, as in hermeneutics, or else is replaced by structuralist studies of man and language, disappearing into ‘human sciences’ like linguistics or anthropology. While both hermeneutics and structuralism disagree with, and attempt to destroy or deconstruct, the priority and privilege of the Grand Tradition of ‘first philosophy’, they represent contrasting attitudes about its results and aftermath. Their opposition is itself considered a basic division in 20th-century thought. Dreyfus and Rabinow think that Foucault’s work transcends this division, and believe he ‘is and has always been beyond structuralism and hermeneutics’. So they chronicle his writings, giving an exposition that explains how each work is pulled toward one pole or the other, at the same time identifying the features that escape either label.

What, however, is ‘beyond’? About this, Dreyfus and Rabinow are less certain. They remark in their conclusion that ‘Foucault himself has described his tactic as a “slalom” (personal communication) between the traditional philosophy and an abandonment of all seriousness.’ This concluding section is itself unusual for a work of this kind in being a cliff-hanger. After reminding us that ‘the work of Michel Foucault is still very much in progress’, they leave us, and him, with two pages of questions about what he means by, and how one could study empirically, his basic conceptions of truth, resistance and power.

While Foucault’s critics would take the frequent changes of course as vacillations, Dreyfus and Rabinow see Foucault’s self-corrections as a healthy sign of learning from temporary mistakes without deviating from a central direction. Foucault himself says here that the theme of his research was always the subject (and, surprisingly, not power): that is, how human beings constitute themselves as subjects and how they treat one another as objects. The apparently mistaken tacks result not so much from idiosyncratic shifts by Foucault as from intellectual tendencies dominant at the time Foucault was writing his various works. In the fast-changing world of French intellectuals it is not surprising that one person could evolve through several conflicting movements. For Dreyfus and Rabinow, then, the Archaeology of Knowledge is not Foucault’s major methodological treatise, as many suppose, but an aberration. Coinciding with the wave of structuralism in France in the Sixties, Foucault stressed the analysis of discourse rather than social institutions.

By his own admission, Foucault in his structuralist phase tended to think of language as autonomous. Discourse organised not only itself but also social practices and historical epochs. Vestiges of idealism lurk in this belief that language constitutes reality. Foucault, however, never thought of language as a subjective or mental phenomenon. He follows Heidegger and others in avoiding the Cartesian belief in the privilege and privacy of consciousness, or the Kantian belief in the constitutive powers of the ego. Foucault’s attack on the concept of man and on ‘humanism’ is a forceful way of saying that the subject is not given with permanent structures that condition reality, but produced historically from its social world.

Moving away from linguistic idealism, then, Foucault’s later histories of the prison and of the institutionalised conceptions of human sexuality show how technical discourses are themselves affected by underlying social practices. Dreyfus and Rabinow believe that Foucault now recognises that the background of social practices can never be completely articulated. If they are right, there is indeed a shift in Foucault’s approach, for his earlier, archaeological method is committed to the claim that a discourse creates its referents and that each era has its own a priori rules of formation, its ‘historical conditions of possibility’. Dreyfus and Rabinow argue that the shift is not a break, but a move to a new method, different from that proposed in the Archaeology. They label this new method ‘interpretive analytics’ and see it as preserving the best of both structuralism and hermeneutics, without repeating the mistakes of either.

The error of hermeneutics is to fall into the temptation of thinking that hidden behind appearances is a deeper meaning: one that cannot be seen by the ordinary social agent, but only by the rare thinker who penetrates to the ontological basis of everyday occurrences. Dreyfus and Rabinow think Foucault himself made this mistake in an early work, Madness and Civilisation, which he revised substantially after its initial publication. Here he began by assuming his historical method allowed him a theory-independent way of understanding the phenomenon of pure madness that successive medical theories of insanity had distorted. His apparent belief that madness would always be distorted by reason, by the attempt to apprehend it rationally, gave the impression he was a proponent of irrationality. Dreyfus and Rabinow insist, however, that ‘Foucault is eminently reasonable’ and ‘is not attacking reason but rather showing how a historical form of rationality has operated’. Foucault went on to correct the impression by giving up the idea that his method brought him into contact in a neutral way with the pure phenomenon, but not before being swept up by the wave of structuralism. The crucial weakness of his structuralist phase is that the focus on the radical discontinuities between autonomous, self-referential discourses made it impossible for him to say how change could come about, particularly social change. He appeared to be a functionalist and insofar as social life is systematically interconnected, his theory seemed politically conservative. If each part of the social system is indispensable, and individuals can have no real effect, rational social change is impossible.

The charge of conservatism, levelled against Foucault by Jürgen Habermas, for instance, is still a difficult one for Foucault to rebut. Much depends on how his new conception of power is understood. In turn, the new method of his most recent books must be discerned without Foucault’s help, for since the Archaeology, he has avoided writing methodological tracts. The Dreyfus/Rabinow formulation of ‘interpretive analytics’ is now the closest thing we have to an explicit account of Foucault’s metatheory. Giving up the hope for a theory that will be immune to social and historical distortions, Foucault’s interpretive analytics is more modest and avowedly pragmatic. The methods and results of his histories of the prison and of sexuality are recognisably influenced by social practices, which can only be investigated and diagnosed from within. But we are not as rigidly fixed in these social practices as conservative functionalists think. Society is not a ‘system’ in any strong sense, and has no inherent stability.

Dreyfus and Rabinow follow Foucault’s appeal to Nietzsche in calling this new form of internal social history genealogy. Archaeology, although useful as a method for getting a degree of detachment and disengagement from current social practices, is now only a tool. With this subordination of archaeology to genealogy Dreyfus and Rabinow hope to circumvent the standard accusations that Foucault’s position is relativistic and even nihilistic. Because his earlier histories stressed the abrupt discontinuities between technical discourses, Foucault’s work was often found compatible with Thomas Kuhn’s accounts of paradigm shifts in the natural sciences. Foucault writes a history of the way various disciplines for acquiring knowledge evolved, but he avoids the ‘Whiggish’ assumption of the necessary superiority of later theories, especially our own, over earlier ones. He is particularly opposed to the realist notion that this superiority results from the deliberate replacement of the earlier, false theory by a later, true one in a cumulative way that gives a clearer picture of things as they really are.

Dreyfus and Rabinow stress, however, that the social and political character of Foucault’s studies separates them in crucial ways from Kuhn’s. In particular, there is a normative and interpretive dimension in Foucault’s histories of the human sciences. As a result, Kuhn and Foucault think of ‘normalcy’ differently. Normal science is for Kuhn a period when science is able to go about its business in a productive way, with success in prediction and without being plagued by anomalies. For Foucault, Dreyfus and Rabinow, the human sciences would stagnate without anomalies – that is, without the fundamental social and political disagreements that always underlie the study of society. Furthermore, society is itself changed by the sciences that study it. These sciences can even produce their own anomalies or ‘abnormalities’, as 19th-century penal practices increased social delinquency.

Dreyfus and Rabinow do not believe the social sciences should become ‘normal’ in Kuhn’s sense, for that would lead to what Foucault calls the normalisation of society, the imposition of a model of well-ordered human activity on all aspects of social life. Practices not conforming to the social exemplars are then identified as deviancy that needs to be forcefully normalised. As the disciplines for acquiring knowledge about human social life developed, Foucault claims, society itself became increasingly subject to discipline in another sense – namely, the control of people with homogenising techniques used by social institutions like prisons, armies and schools.

Foucault’s critics, I am sure, will want more evidence for this account of the causal interaction of society and its ‘sciences’. To press him, they could well ask what grounds he has for regretting the increasing normalisation of society in the last two centuries. Since genealogy can work only within a given social context, Foucault cannot argue either that the past was any better, or that the future should be any different. He is criticising the tendency toward increased normalisation in contemporary society, yet he appears to have no alternative to offer, and no standards on which to base his angry charge that modern society is becoming more and more like a prison, however progressive and benevolent it appears.

Foucault poses a dilemma for his interpreters since he deliberately abstains from defending any normative standards and yet his histories have normative overtones and implications. Dreyfus and Rabinow defend him against the charges of relativism and nihilism by arguing that Foucault can diagnose the organising trends of our culture only because he, too, is subjected to them. We should recognise his own anger, and the basis for his criticism, in ourselves. If we object to the vision of a totally normalised society, we are not objecting to Foucault, but to a tendency of our times. The resistance to the trends of our society must also come from within it. If the normalisation were total, then there would be no grounds for criticism – and the impossibility of social criticism is a most objectionable feature of the normalised society. ‘Clearly,’ write Dreyfus and Rabinow, ‘Foucault is not saying that all of the practices of our culture are disciplinary or confessional, or that every production of knowledge functions immediately as a power-effect. The trend toward normalisation has not succeeded in totalising all of the practices.’ Unfortunately, Foucault does sometimes say what they think he should not. In Discipline and Punish, he concludes that ‘the carceral archipelago transported this [penitential] technique from the penal institution to the entire social body.’ They are right, though, that on Foucault’s own grounds he should not totalise this tendency in our society toward the ‘totalising ordering of things’.

Nihilism does not go away so quickly, however, and Nietzscheans are pestered by its eternal recurrence. Where Foucault’s thinking becomes particularly difficult to understand is not so much in his social criticism as in his conceptions of knowledge and truth. During his earlier, structuralist phase he seemed to be relativising knowledge to whatever discourse happened to be spoken at a particular time, and to be denying the possibility of serious truth. For the archaeologist, according to Dreyfus and Rabinow, the idea that ‘there could ever be issues about which it would be worth arguing’ is a ‘mysterious, inevitable illusion’. Why is truth only illusory? When people assert something, they certainly believe in the meaningfulness and truth of what they say. From the disinterested, archaeological perspective, however, what counts as true is determined by the conceptual framework, or rather, the discursive practices of a discipline. Furthermore, since for Foucault these are subject to abrupt changes, discourses and truth seem historically relative, or even arbitrary.

The archaeologist’s utterances except themselves from this relativity, producing the contradictory attitude that Dreyfus and Rabinow characterise as a Husserlian ‘ego split’: ‘Foucault the archaeologist looks on, as a detached metaphenomenologist, at the historical Foucault who can’t, if he thinks about human beings in a serious way, help thinking in terms of the meanings and truth claims governed by the latest discursive formation.’ This split between understanding serious meaning (for instance, any moral or social theory) and pretending not to leads to Foucault’s personal slalom between being a deeply committed private individual and a public figure who must hide behind masks.

He can reveal his own values and beliefs only indirectly. Although he cannot say other theories or modes of discourse are true or false, he can to some extent evaluate them critically. Nietzsche similarly could not appeal to truth, but could distinguish sick from healthful interpretations. Dreyfus and Rabinow see Foucault’s complaint that humanistic theories are ‘warped’ and ‘twisted’ forms of reflection as equivalent to the procedure whereby historians of science identify a research programme as ‘disintegrating’ – namely, as proving that ‘excitement and energy have gradually given way either to boredom and discouragement or factions and fads.’ The difficulty with their tactic is that usually the strategy for showing a programme to be disintegrative is used when no one, or only a few, could still pursue it earnestly. This strategy is really an historian’s device. Its use by a contemporary social critic who fails to specify an alternative to the current humanistic discourses would be simply whimsical.

In the more recent, genealogical phase where Foucault recognises that discourse is not autonomous, but based on social practices, he can be more frank about the engaged, purposive character of his ‘history of the present’. No longer claiming, like the archaeologist, to be outside current social practices, the genealogist takes them seriously enough to want to rectify malignancies. Foucault has given up theorising about power and settles now for diagnosing particular symptomatic manifestations of it. Physician rather than philosopher, he lets his histories tell their own story, and leaves the veracity of his diagnosis for his patients, and the future, to decide.

Of course, he cannot abstain entirely from theory. He wants to introduce new operating terms, like ‘power/knowledge’, so he must clarify these and explain which historical phenomena are relevant. Furthermore, he sees his histories as countering previous theories, and the distorted historiography they produced. For instance, his history of the prison is intended as a corrective to a similar history by two Frankfurt School theorists who relied heavily on materialist, base-superstructure explanations and an inadequate conception of power.

So power is in particular a crucial but troublesome notion. Although Dreyfus and Rabinow admire the recent histories in which power is a central theme, they admit their perplexity about Foucault’s understanding of the term. They give a credible explanation of how his conception of power differs from standard views of it as oppressive ‘power-over’. As a historian, he is suggesting that power as conceived by political theorists from Hobbes to Marcuse is not a good tool for writing social histories. He thinks traditional theorists construe power too negatively as the antithesis of knowledge and as something to be eliminated with a true vision of social relations. They therefore obscure the extent to which knowledge-gathering activities like the social and human sciences both support and are supported by power relations. Dreyfus and Rabinow caution, however, that power and knowledge are not identical for Foucault. They deny he is saying that all science, knowledge and truth is merely a product of power. His subject-matter is simply restricted to those dubious disciplines like penology or psychiatry which claim to be legitimate sciences but are intimately involved in the exercise of power. Nevertheless, like him they sometimes reason in a way that worries many about Foucault’s line: ‘Foucault argues that repression itself is not the most general form of domination. In fact, the belief that one is resisting repression, whether by self-knowledge or by speaking the truth, supports domination, for it hides the real working of power.’ Should one therefore never believe one is resisting repression or speaking the truth?

If this argument were right, furthermore, no one, not even Foucault and his friends, could ever hope to discover ‘the real working of power’ without perpetrating oppression. Since such a discovery would also be a form of self-knowledge, and self-knowledge in the interest of freedom hides the reality of power, the discovery would be impossible in the first place.

Dreyfus and Rabinow are aware of these difficulties, and it may have been at their instigation that Foucault wrote the section of the ‘Afterword’ entitled ‘How is power exercised?’ On my own reading of that essay, Foucault defends himself against the common charges of political conservatism, moral nihilism and epistemological relativism by explaining that although he abstains from the rhetoric of progress, he has not given up hope for future emancipation from some of our repressive power-relations. Many of us will disagree with his evaluation of contemporary society, but Dreyfus and Rabinow have persuaded me that at least his method is not incoherent.

On this point, however, other recent commentators on Foucault will need more persuasion. Harvard philosopher Hilary Putnam, for instance, maintains in Reason, Truth and History (1981) that Foucault is a relativist, and that relativism is logically unintelligible. Much like Habermas, Putnam thinks we must (for logical reasons) believe in the regulative ideal of reason, of a ‘just, attentive, balanced intellect’. He sees Foucault as committed to the belief that every culture, including our own, is irrational. But then, if there is no culture that is rational, then there is no standpoint from which Foucault could assert that every cultural standpoint is irrational. So his position is not just empirically false but self-refuting.

If Dreyfus and Rabinow are right, however, Putnam’s criticism would apply at most only to the earlier, structuralist Foucault, and not to his later, genealogical histories. There Foucault aptly avoids being a political theorist and settles for proposing a pragmatic attitude for historians, one that is not true or false, but either heuristic or not. His most recent stance is a pragmatist one, so his writings are not systematic treatises but histories that respond to what he sees as the needs of the present.

Putnam himself remarks in passing that Foucault is less a historian than a satirist, like Swift. If this genre assessment is correct, I should think Foucault need not fall into Putnam’s trap. Must the satirist espouse explicitly a constructive political vision? If he does, he could quickly slide into incoherence, for satire can be directed at most cultures or aspects of culture. But few satirists would assert the irrationality of all aspects of all cultures. If Swift does, then Foucault is different, for he says he still believes in the possibility of some emancipation, although not in the ideal of a totally emancipated society. Satire can simply omit consideration of the higher achievements of a culture without thereby denying that there are any beneficial elements. Indeed, the purpose of satire is often to be edifying and elevating. Like Heidegger, he may be thinking that exposing the platitudes of humanism will allow us to become more truly human in specific respects. The thought is that we would be better people if we did not assume the necessary superiority of our own historical standpoint. Foucault thus writes histories, or genealogies, that function like satires. He lets the documents speak for themselves. Presumably we the readers will find the examples so extreme that we will feel they could have been written only by a satirist. We may wonder why people in the past were not shocked by the irrationality of their practices, and when we begin to suspect from examples closer to the present that the same might be true of us, we may become more critical of our own presumed rationality, of the legitimacy and humaneness of our social practices.

What is incoherent and unintelligible, then, is not the satirical story itself, but the practices and behaviour it recounts. Satire is not itself incoherent, but like ‘good’ hermeneutics, it brings out the incoherence in practices that are so familiar and so close to us that they are taken for granted. Satire is not necessarily conservative since it can unsettle apathetic resignation to social institutions and expose reactionary, dogmatic belief in their legitimacy. Satire is a highly efficacious genre, and Foucault’s genealogy may well be its latest variant.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences