The following text has been automatically reproduced by an Optical Character Recognition (OCR) algorithm. It may not have been checked over by human eyes. For matters of precision please consult the original pdf.

The Return of the Subject in late Foucault

The Return of the Subject in
late Foucault
Peter Dews
The following essay is an initial attempt to extend the comparison of the thought of Michel Foucault with that of the Frankfurt
School, begun in my Logics of Disintegration (Verso, 1987), to
cover the work ofFoucault’s last phase. It does not claim to be a
comprehensive analysis, but simply seeks to establish two fundamental points: firstly, that the return of a self-constituting subjectivity in F oucault’ s final writings cannot be seen as merely a
shift of emphasis within a consistent probject (as suggested,for
example, by Deleuze, in his book on Foucault), but arises out of
the intractable dilemmas ofFoucault’ s earlier work, and represents a break with many of its assumptions; secondly, that the
form in which Foucault introduces the concept of the subject,
namely as an undialectical reaction to the political implications
of philosophical’ antihumanism’ , raises as many problems as it
solves. A somewhat different version of this essay is to appear in
German in the anthology Die Aktualitlit der ‘Dialektik der
Aufkllirung’: Zwischen Modernismus und Postmodernismus,
published by Campus Verlag, Frankfurt (1989). I am grateful to
the Verlag for permission to republish this material here.


In the final years of his life, Michel Foucault came to acknowledge that he could have avoided many detours and oversights in
his own research, had he been familiar with the work of the
Frankfurt School at any earlier date. Despite these recognized
affinities, however, there is clearly a significant gap in the way in
which regulating power and its dominated other, and the relation
between them, are theorized in the two cases. For the classical
Frankfurt School the oppressiveness of this relation is the result
of the preponderance in modem society of a restricted means-end
rationality, which cannot be seen as exhausting the promise of
reason as such. This means that the corporeal is not seen as
intrinsically irrational. Freud’s error, for example, is to take the
conformist ego’s view of the drives as threatening and chaotic for
an immutable truth. Conversely, instrumental reason itself appears irrational from the standpoint of a suffering nature which it
overrides. Each aspect of a divided reason experiences the irrationality of its other. Foucault, of course, rejects this perspective,
since – along with other post-structuralist thinkers – he suspects
that the promise of an undivided reason has totalitarian implications. He argues that we cannot speak of reason and its history as
such, but only of a plurality of practices, of ‘forms of rationality’ ,
which compete and overlap with each other. Our task is to situate
ourselves within this ever-shifting field of struggles, without anticipating any ultimate ‘reconciliation’.

The problem with Foucault’s position is that it deprives the
concept of rationality of any determinate content. The diversity
of rationalities is simply equivalent to the diversity of practices.

This generates two major problems. Firstly, there is the question

of the general connection between power and knowledge on
which his work of the 1970s, in particular, is based. Although his
evident intention is to present power and knowledge as internally
related (hence his use of the hyphenated term ‘powerknowledge’), this relation is in fact most frequently portrayed in
terms of the institutional preconditions for the formation of
certain types of knowledge. Foucault’s fundamental argument is
that it is the opportunities for close surveillance opened up by the
asylum, the hospital, the prison, which makes possible the
elaboration of the corresponding ‘human sciences’. Thus, in an
interview dating from 1975, he suggests that:

The archaeology of the human sciences has to be
established through studying the mechanisms of power
which have invested human bodies, acts and forms of
behaviour. And this investigation enables us to rediscover
one of the conditions of the emergence of the human
sciences: the great 19th century effort in discipline and
normalization. ‘1
But to talk in this way is in fact to make a relation between power
and knowledge non-intrinsic. Foucault does not explain how the
‘effort in discipline and normalization’ is enhanced by the application of scientific knowledge. The reason for this failure is not
difficult to discover. For, were Foucault to admit that the application of scientific knowledge increases the effectivity of action, he
would be obliged to abandon his underlying relativist stance, and
to admit the reality of ‘progress’ in at least one dimension of
rationali~y: the cognitive-instrumental dimension. Hence the
crossing of the ‘technological’ threshold by disciplines, the spiralling f{~inforcement of power and knowledge which Foucault
evokes, remains theoretically unexplained.

Secondly, there is a deep difficulty in Foucault’s accounts of
the relation between disciplinary power and the body, ‘rationalities’ and their ‘other’. Since Foucault wishes to avoid judging
power-knowledge complexes from a normative standpoint, by
assessing the force of the claim embodied in the label ‘rationality’, he is obliged to refuse any distinction between facticity and
validity, and therefore cannot denounce the human sciences as
forms of distortion or misrepresentation. For Foucault, as we
know most clearly from The Archaeology of Knowledge, the
‘objects’ of discursive formations are defined by these formations. But this abstention from judgements of validity leads to
difficulues when Foucault wishes to give his position a critical
edge. An attack on disciplinary power, for example, could only
be carried out from the standpoint of an alternative conception of
the body. But for Foucault this second conception could only be
part of another power-complex, and could not claim any greater
‘truth’ 0: normative superiority.

Foucault’s response to this dilemma remains fundamentally
ambiguous. On the one hand he is tempted to abandon his critical

claims, suggesting that
It is necessary to pass over to the other side – the other side
from the ‘good side’ – in order to try to free oneself from
these mechanisms which make two sides appear, in order
to dissolve the false unity of this other side whose part one
has taken. 2
On the other hand, Foucault is clearly unable entirely to abandon
an emancipatory perspective. But this perspective is condemned
to remain tentative and fleeting, since it seems to require, in
contrast to Adorno and Horkheimer’s exposure of the irrationality of the dominant ratio, an espousal of irrationality itself.

This difference is clearly apparent in the differing attitudes of
Adorno and Foucault to the idea of a utopia of non-regulated sensuousness. In The History of Sexuality, for example, Foucault
permits himself to evoke fleetingly a ‘different economy of
bodies and pleasures’ which would no longer be subordinated to
the confessional quest for identity,3 but this remains only an
elusive suggestion. Any more positive determination of the body
and its needs would contravene Foucault’s deep inclination
towards relativism. By contrast, in Negative Dialectics Adorno
argues explicitly that ‘all happiness aims at sensual fulfilment
and obtains its objectivity in that fulfilment. A happiness blocked
off from every such aspect is no happiness. ‘4
Even if one rejects the suggestion that its implications are
totalitarian, there is clearly a justified worry behind Foucault’s
resistance to the project of the restoration of the integrity of a

bisected reason. His suspicion is that the totalization of instrumental reason is too simple a story to account for the complexities of modernity. Paradoxically, it is Foucault – often taken to be
an archetypical thinker of ‘postmodernity’ – who can be seen as
defending a conception of the pluralism and openness of modernity, while Adorno and Horkheimer appear to be ‘postmodern’ in
their virtual extinction of the emancipatory power of reason. The
direct statement in Dialektik der Aujkliirung, that ‘Enlightenment is as totalitarian as any system ‘5 suggests the difficulties
which Adorno and Horkheimer will have in giving any coherent
account of the progressive dimension of Enlightenment univer-


salism. As Herbert Schnadelbach has written, with reference to
Horkheimer’s moral philosophy:

One could almost identify critical theory as a whole with
the conviction that the General and the Powerful cannot
be good because it is general and powerful; in other
words, the Good in this world is to be sought in the
ephemeral, the weak, in the individual impulse, in the
exception, indeed in the improbable – in the unexpected
and actually unwise goodness of individual motives and
Despite his ostensible reluctance to interpret modernity in terms
of any unilinear model of rationalization, Foucault does perceive
– and is even more hostile to – the rise of the universal claims of
bourgeois morality. This can be seen clearly from certain passages in Madness and Civilization, where Foucault argues that it
is precisely the emergence of conscience which makes the insane
legitimate targets for correction. ‘The fundamental principles of
bourgeois society,’ he writes, ‘permit this conscience which is
both private and universal to reign over madness without any
possible contestation. ‘7 Thus, there is a fundamental difficulty in
the positicns both of Adorno and Horkheimer and of Foucault, in
relation 11.) the coherent formulation of the practical consequences of their respective positions. The former retain a concept of the subject, but in a form which condemns the subject to
an inevitable, totalizing process of reification. (The equation of
the universal and the rational in the moral philosophy of German
Idealism simply perpetuates the domination of nature.) By contrast, Foucault, for most of his career, theorizes the subject as
entirely constructed through social practices, and in this respect
acquires greater freedom of interpretation, to the extent that he
sometimes denies being able to give any determinate content to
the concept of modernity.s Yet the price of this abandonment is
an inability to think the concept of emancipation coherently at
all, since, as we have seen, Foucault is deeply suspicious of what
Horkheimer and Adomo term the ‘remembrance of nature in the
subject, in whose fulfillment the unacknowledged truth of all
culture lies hidden’ .9
This cl ifficulty in part explains the abrupt theoretical shift
which characterizes Foucault’s late work. It is impossible not to
read this work both as an attempt to overcome the ambiguity of
his earlier relation to concepts of power and emancipation, and
an admis~ i)n of the limits of the ‘postmodern’ thought which
attempts to bypass the concept of the subject, and consequently
destroys any coherent notion of freedom at all. Foucault’s task, in
his late work, will be to articulate the concepts of subjectivity and
freedom in such a way as to avoid any suggestion that such
freedom must take the form of the recovery of an authentic
‘natural’ self.

This move is reinforced by Foucault’s conviction that modern technologies of power and the belief in authenticity are
intimately related: the notion of a liberation of nature, underpinned by a scientific theory of the deep self, such as psychoanalysis, leads simply to a deeper enslavement. In fact, the first
volume of The History of Sexuality can be seen as Foucault’s
attempt to provide a genealogy of ‘deep subjectivity’. Foucault
draws attention to the dissolution of the forms of group identity
which characterize traditional societies, and their replacement by
a form of identity which depends increasingly on the capacity of
the individual to reflect upon and articulate the domain of private
experience, suggesting that this transition is epitomized in the
change in meaning of the word ‘avowal’:

For a long time, the individual was vouched for by the
reference to others and the demonstration of his ties to the
comrr:onweal (family, allegiance, protection); then he

was authenticated by the discourse of truth he was obliged
to pronounce concerning himself.lo
Foucault correlates this transition with the shift from epic narrative to the modem literature of introspection, and with the rise of
philosophies of consciousness, ‘the long discussions concerning
the possibility of constituting a science of the subject, the validity
of introspection, lived experience as evidence of the presence of
consciousness to itself’ .11 Yet Foucault wishes to suggest that our
broadened access to an ‘inner world’ distinct from the external
worlds of both nature and the social is the result of a forgotten

Psychoanalysis prides itself on restoring the capacity for
pleasure, which is impaired by neurotic illness. As if the
mere (;oncept of a capacity for pleasure did not suffice
gravely to devalue such a thing, if it exists. As if a
happiness gained through speculation on happiness were
not the opposite, a further encroachment of institutionally
planned behaviour-patterns on the ever diminishing
sphere of experience.IS

One confesses – or is forced to confess. When it is not
spontaneous or dictated by some external imperative, the
confession is wrung from a person by violence or threat; it
is driven from its hiding place in the soul or extracted
from the body.I2
By linking the capacity for avowal to the inquisitions of the
confessional, Foucault is able to argue that ‘the obligation to
confess is so deeply ingrained in us, that we no longer perceive it
as the effect of a power which constrains us; on the contrary, it
seems that truth, lodged in our own secret nature, “demands”
only to surface.’I3
It is interesting to compare this argument with one of its
prototypes – Nietzsche’s account of the origins of bad conscience, in the second essay of On the Genealogy of Morals. For
it is clear that, despite his emphasis on the cruelty which is
required for moral imperatives to be internalized, Nietzsche does
not consider the emergence of an intensified awareness of one’s
own inner impulses to be simply a power-induced illusion. He

Let me hasten to add that the phenomenon of an animal
soul turning in upon itself, taking arms against itself, was
so novel, so profound, mysterious, contradictory, and
pregnant with possibility, that the complexion of the
universe was changed thereby. This spectacle (and the
end of it is not yet in sight) required a divine audience to
do it justice. It was a spectacle too sublime and paradoxical to pass unnoticed on some trivial planet. Henceforth
man was to figure among the most unexpected and
breathtaking throws in the game of dice played by Heraclitus’s great “child”, be he called Zeus or Chance. Man
now aroused an interest, a suspense, a hope, almost a
conviction – as though in him something were heralded,
as though he were not a goal but away, an interlude,. a
bridge, a great promise… ’14
Nietzsche, in other words, while intensely aware of the paralysing capacity of an excessive self-consciousness, does not consider the discovery of inner depth to be simply a power-induced
illusion. Rather, this discovery must be incorporated and transcended towards a new spontaneity. I would argue, by contrast,
that in his position Foucault has confIated two issues. His critique of the culture of therapy, and of a self-destructive cultivation of subjectivity is undoubtedly legitimate. Yet these cultural
developments need not be the only forms which a more fluid
access to inner nature can take. There is also the possibility of
more self-expressive shaping of everyday life, which would
enable the subjective and public geography of contemporary
societies to enter into a more balanced relationship.

I would argue that the Frankfurt School have a more complex
account of these problems. Adorno and Horkheimer are by no
means oblivious to the manipulative potential of psychoanalysis.

In Minima Moralia Adorno is ruthless in his exposure of psychoanalysis as a form of social control:

It is clearly a similar concern which underlies Foucault’s attack
on the ‘repressive hypothesis’ , in the frrst volume of The History
of Sexuality, and his suspicion of doctrines of liberation which
rely on a conception of a ‘deep self’ which needs to be uncovered
through some privileged form of cognitive access. Yet, at the
same time, Adorno’s position also indicates what might be
suspect about Foucault’s refusal of the deep self. Adorno does
not deny the ‘bottomless fraud of mere inwardness’I6 In Minima
Moralia he argues that ‘authenticity itself becomes a lie the
moment it becomes authentic, that is, in reflecting on itself, in
postulating itself as genuine, in which it already oversteps the
identity which it lays claims to in the same breath. ’17 But he also
sees the smoothing over of ‘terror before the abyss of the self as
equally a function of the technologies of social adaptation. IS A
similar conception was, of course, made popular by Marcuse
during the 1960s, in the form of an account of ‘repressive
desublinlation’ . Long before Foucault, the Critical Theory tradition had highlighted the possibility of a regressive form of
libidinal emancipation which simply slotted individuals even
more efficiently into the existing system of production and
consumption. However, for Critical Theory, this possibility does
not invalidate the insights of psychoanalysis into the fatality of
the dialectic between nature and society. The tendency of Critical Theory, from the late 1930s onwards, was to distance itself
from psychoanalysis as a supposedly positive science of the
mind, and from the adaptational function of therapeutic practice,
while at the same time retaining Freud’s metapsychological
insights, as a theorization, at the level of the individual psyche, of
the inherently conflictual character of culture.I9
In his late work Foucault attempts to escape these complexities, by advocating what he terms an ‘aesthetics of existence’ ,


inspired by his’ research into the ethical codes of Greek and
Roman Antiquity. He appeals to a notion of pure self-stylization,
which would not be imposed as a universal norm, but would
rather be open to the choice of the individual. However, it is
difficult to see how in contemporary society any such turn
towards an aesthetics of existence could be anything other than a
reinforcement of social tendencies towards atomization. Not
only this, but Foucault fails to appreciate the dialectic inherent in
the concept of individuality itself. For Adorno and Horkheimer:

The independence and incomparability of the individual
crystallize resistance to the blind, repressive force of the
irrational whole. But, historically, this resistance was
only made possible by the blindness and irrationality of
each independent and incomparable individual …. The
radically individual features of a person are both components in one, the factor which has been able to escape the
ruling system and fortunately lives on, and the symptom
of the injury by which the system maims its members. 20
In contrast to this dialectical conception, in much of his work of
the 1970s, Foucault describes individualization in a one-sided
manner as merely the effect of technologies of power. Then, in
his last works, he surprisingly shifts to a positive evaluation of
the individual cultivation of the self. However, what appears to
be the advocacy of an arbitrary stylization of life in these works
could easily reinforce the situation described in Dialektik der

Pseudo-individuality is rife: from the standardized jazz
improvization to the exceptional film star whose hair
curls over her eye to demonstrate her originality. What is
individual is no more than the generality’s power to stamp
the accidental detail so firmly that it is accepted as
SUCh. 19

The possibility of such an outcome is reinforced by the fact that
Foucault explicitly attacks the ‘idea of an analytical or necessary
link between ethics and other social or economic or political
structures’ .22
The problematic features of Foucault’ s conception can be
further highlighted by enquiring what the content of Foucault’s
concept of the aesthetic might be, in his invocations of an
aesthetics of existence. In one sense, this term is clearly an
anachronism when applied to the ethical codes of Antiquity,
since, as Foucault himself makes clear, such codes were deeply
embedded in a nexus of social relations of power and prestige:

the modem autonomy of the aesthetic is here nowhere visible.

Furthermore, to speak of the possibility of an ‘aesthetics of
existence’ is to describe a situation in which the aesthetic would
lose its specificity. For, as Rudiger Bubner has argued:

Familiarity with a life-world, in which we feel at home, is
to such an extent the reservoir of aesthetic effects, that its
loss would be at the same time the loss of aesthetic
possibilities. We experience unburdening, alienation,
new reflections, full illumination, pure content, only in
contrast to our everyday view of things. If this disappears,
because artistic phenomena take its place, then the fiction
begins to petrify.23
This argument suggests the curious relation in which Foucault’s critique of the deep self stands to the neo-conservative
critique of contemporary culture. On the one hand, he denounces
the cult of subjectivity and authenticity, yet at the same time his
very solution implies a breaking down of the barrier between art
and life, and a proliferation of lifestyles not primarily oriented
towards competition and achievement which, for the neoconservatives, would be subversive and socially destructive.


One final question which cannot be avoided concerns the
nature of the freedom which late Foucault invokes, both as the
basis of resistance to power, and as the freedom of self-creation.

The introduction of a concept of freedom seems to mark a
particularly abrupt break with Foucault’s earlier work, where the
subject, with its illusion of autonomy, is theorized as a construction of power and discourse. Despite this, Foucault, in his essay
on Kant’s ‘Was ist Au.fklarung’, speaks of the ‘constitution of
ourselves as autonomous subjects’ and, in an interview with
Gerard Raulet, states that his concern is ‘an analysis of the
relation between forms of reflexivity – a relation of self to self and hence between forms of reflexivity and discourse of truth,
forms of rationality and effects of knowledge’ .’lAYet this admission of reflexivity as the defining attribute of subjectivity raises
enormous problems which remain entirely unexplored in Foucault’s late work.

Only one of these can be mentioned here. The obvious
paradox of a reflexive account of self-construction is that the self
must already exist in order to construct itself. It is this problem
which comes to the fore in post-Kantian idealism and is most
earnestly grappled with by Fichte. For Fichte the self must be an
act of self-positing which posits itself as precisely this act. But
this conception of the self as act, rather than as substance, or even
as a formal unity, has two important consequences. Firstly, our
language, designed for coping with the objective world, becomes
inherently problematic when confronted with the task of explicating the structure of self-awareness. For Fichte, the fundamental orientation of language is towards objectivity, and it is for this
reason that he introduces the neologism ‘Tathandlung’ (as opposed to ‘Tatsache’) to describe the ‘givenness’ of the activity of
the self. In other words, tbe problem of the difficulty of access to
subjectivity begins here, and is not simply a construct of power.

Secondly, this activity of the self, and its tendential release from
all objective restriction, becomes the principle of morality.

By contrast, Foucault’s contention is that”ethical self-construction operates in a reflexive medium, yet at the same time he
wishes to deny that this medium itself has any ethical relevance.

It is merely the locus of ‘games of truth’.25 However, many of
Foucault’s late formulations appear to contradict this denial.

Thus, his argument against humanism, in the essay on Kant, is
fundamentally that it fixes a conception of the human being:

‘what is called humanism has always been obliged to lean on
certain conceptions of man borrowed from religion, science, or
politics. Humanism serves to color and to justify the conceptions
of man to which it is, after all, obliged to take recourse.’ To this
Foucault opposes an awareness of the contingency of all historical institutions and practices, and ‘the principle of a critique and
a permanent creation of ourselves in our autonomy’.26 Indeed, in
some of his very last interviews, Foucault argued that his role
was ‘to show people that they are much freer than they feel, that
people accept as truth, as evidence, some themes which have
been built up at a certain moment in history, and that this socalled evidence can be criticized and destroyed. ’27 Yet, once he
has made this move, there is one clear sense in which the ethical
relevance of the question of a true or false self-relation cannot be
avoided: the sense in which it is possible for the self to be
ignorant ef its own autonomy, in other words its own activity.

Foucault may deny the specific construal of that activity, inspired by psychoanalysis, which is presented in Dialektik der
Aufklarung, as a simultaneous emancipation from and perpetuation of the compulsion of nature. But it appears that, at the end
of his life, he could no longer avoid the fact that the understanding of social and historical processes is, if not a component of our
self-unde:.-standing, at the very least a contribution to our liberation from 3elf-misunderstanding.



Michel Foucault, ‘Body/Power’, in Colin Gordon (ed.), Power/
Knowledge, Brighton, 1980, p. 61.

Michel Foucault, ‘Non au Sexe Roi’, Le Nouvel Observateur,
644, 12-21 March 1977, p. 113.

Michel Foucault, The History of Sexuality; Volume One: An
Introduction, Harrnondsworth, 1978, p. 159.

Theodor Adomo, Negative Dialectics, London, 1973, p. 202.

Theodor Adorno and Max Horkheimer, Dialectic ofEnlightenment, London, 1979, p. 24.

Herbert Schnadelbach, ‘Max Horkheimer and the Moral Philosophy of German Idealism’ , T elos 66, Winter 1985 -6, p. 87.

Michel Foucault, Histoire de lafolie a I’ age classique, collection TEL edn, Paris, 1976, pp. 465-66.

See ‘Structuralism and Post-structuralism: an Interview with
Michel Foucault’, Telos 55, Spring 1983, p. 200.

Dialectic of Enlightenment, p. 40.

The History of Sexuality; Volume One, p. 58.

Ibid., p. 64.

Ibid., p. 59.

Ibid., p. 60.

Friedrich Nietzsche, On the Genealogy ofMorals, essay 2, para.





Theodor Adomo, Minima Moralia, London, 1974, p. 62.

Ibid., p. 64.

Ibid., p. 154.

Ibid., p. 65.

See Wolfgang Bonss, ‘Psychoanalyse als Wissenschaft und
Kritik. Zur Freudrezeption der Frankfurter Schule’, in
Wolfgang Bonss and Axel Honneth (eds.), Sozial[orschung als
Kritik, Frankfurt, 1982, pp. 397-405.

Dialectic of Enlightenment, p. 241.

Ibid., p. 154.

Michel Foucault, ‘On the genealogy of Ethics’, in Hubert
Dreyfus and Paul Rabinow, Michel Foucault; Beyond Structuralism and Hermeneutics, 2nd edition, Chicago, 1982, p. 236.

Rtidiger Bubner, ‘Moderne Ersatzfunktionen des
Asthetischen’, Merkur, No. 2, February, 1986, p. 107.

‘Structuralism and Post-structuralism: an Interview with
Michel Foucault’, p. 203.

Michel Foucault, L’ usage des plaisirs: Histoire de la sexualite
2,Paris, 1984,p. 13.

Michel Foucault, ‘What is Enlightenment’, in Paul Rabinow
(ed.), The Foucault Reader, Harrnondsworth, 1986, p. 44.

‘Truth, Power, Self: An Interview with Michel Foucault’, in
Luther H. Martin, Huck Gutman and Patrick H. Hutton (eds.),
Technologies of the Self, London, 1988, p. 10.



No 4

Beyond the “‘Rea llty Ef feet”‘

Postrnodernisrn and Te levision

Notes on Frornrn and Technology

Translaton of Meaning

An Enllghtenment View of the
British Const itut ion

An Interview with JOHN CRUGHIE

or £. 1 (inc p&p) from : ()1SCOl.A'”~
The Philosophy Department
GbsgoTf University
Glasgow 612800



Download the PDFBuy the latest issue