The compulsion to confess

The compulsion to confess This paper draws on the work o f Michel Foucault in order to sketch a prelim inary genealogy o f the practice o f confession in the twentieth century. The essay argues that confession has undergone major transformations, not only from a chiefly religious to a secular practice, hut to a form o f psychologised self-knowledge productively typical o f knowledge itse lf in post-Kantian modernity. In other words, we argue that confession has become diffused through knowledge practices such that it becomes imperative to confess to a particular style or use o f language in the pursuit o f such knowledge. The confession o f a style in language thus becomes a prerequisite fo r such knowledge, or the inability to arrive at it. We investigate the phenomenon in the examples o f the fac tiona l' literature o f Norman Mailer, and the human science o f ethnography.


Introduction
In confession the sinner tells what he knows; in analysis the neurotic has to tell more (Freud, 1 9 8 6 :2 8 9 ).
One of the most striking things about the late twentieth century in the West is the fact that despite the existence of a strenously secular culture, and the ongoing disappearance of formal institutions of confession, the compulsion to confess, even in the most unlikely of contexts, is still unmistakably with us. Confessional discourse continues to thrive and what might appear at first glance to be no more than an arcane or distinctively Christian religious practice has shown itself capable of a surprisingly long and varied life. Wliat is more, far from being seen as conservative, in their contemporary fomis confessional practices are most often regarded as progressive and epistemologically radical. In what follows we wish to examine the origins of the contemporary forms of confessional practices and assess their claims to epistemological power.
The classical form of the confession, that is best represented in Catholicism, has, as miglit be expected, survived -albeit in a very specific context. However, more interesting is the extent to which other confessional practices o f a mutated but recognisable form, have come to permeate many unexpected aspects of contemporary writing and culture.
It is relatively easy to understand the persistence o f broadly confessional practices in religious contexts such as charismatic evangelism and even in allied, what could probably be called neo-religious groups, such as Alcoholics Anonymous. Wliat is less easy to account for is the particular -essentially psychological -form which the practices associated with these groups take and the extent to which confessional elements appear in other settings, many o f which also have no religious connections at all. O f these 'other' contexts perhaps the most unexpected and certainly that most gemiane to this paper, is the emergence of confession in the apparently non personal field of the knowledges, especially the human and social sciences such as ethnography. Perhaps even more surprisingly, what the human scientist now confesses to is not confined to the familiar psychological sins o f passion or prejudice but has come to include the possession of a potentially problematic personal writing style.
It is this unexpected turn to confession as a linguistic or discursive event in knowledge which most concerns this paper, one which we believe it would be almost impossible to account for without the help o f Foucault. Foucault's contribution to an understanding of the nature and origins o f contemporary confessional practices is to be found in a number of his texts, both genealogical and archaeological.
On the basis o f his genealogical works, especially The History o f Sexuality Vol. 1 (Foucault, 1978) it is possible to explain the persistence of confession in modernity, the alacrity with which people today confess and the special status confession confers upon those who confess. Added to this the archaeologies, especially The Order o f Things (Foucault, 1973) make it possible to explain the prevalence of confessional practices in modem knowledge and more importantly, to account for how confession can be seen to have the significant epistemological consequence of validating, or at least going some way towards validating, truth claims.
On the basis o f Discipline and Punish (Foucault, 1977a) and The History o f Sexuality Vol. 1 (Foucault, 1978) it is possible to trace the origins o f the psychological form which modem power takes and the special role sexuality plays in its deployment while the title essay of the collection Technologies o f the S e lf (Martin el al., 1988) provides the historical underpinnings required for a critique of contemporary confessional practices. The kernel of the position presented here draws on The Order o f Things (Foucault, 1973) so as to explain the emergence of confession in modem knowledge.
Together these works provide the necessary elements to account for the emergence of the confessant from what was essentially a position of weakness occasioned by error, in which the strength was on the side of the confessor, to that in which the act of confession, be it of a personal, political or intellectual nature, is in the interests o f the penitent.

The origins of confession
In the title essay of the collection Technologies o f the S e lf (1988), Foucault's central concern is to outline tlie Graeco-Roman and subsequent early Christian confessional practices which underpin his work in The History o f Sexuality generally, and which clearly have a bearing on modem versions of confession as psychological/epistemological events.
In general terms the essay helps to account for why the "association of prohibition and strong incitations to speak [should be] a constant feature of our culture" (Foucault, 1988;16-17). What Foucault means by "technologies of the se lf' are a series of practices in which individuals act upon themselves in regulated ways designed to achieve specific effects or goals. Foucault examines these technologies of the self in two contexts.
Firstly, he outlines the confessional tradition embodied in Graeco-Roman philosophy in the first two centuries AD, followed by a description of those associated with Christian spirituality manifest in the monastic principles developed in the 4th and 5th centuries of the Roman Empire. He describes the two principal technologies of the self originating in the Graeco-Roman tradition as care of the self and knowledge of the self The Stoic tradition emphasized the care o f self whereas the Delphic was concerned with knowledge o f the self In this eariy period, Foucault points out, knowledge of the self was subordinate to taking care of oneself, but in the modem worid the relationship between the two is inverted (Foucault, 1988:22).
While popular accounts of Greek life, especially those which emphasise the Stoic tradition, would suggest that technologies o f the self are predominantly in the modes o f physical training and trials o f control and endurance, Foucault's account explains the importance of writing as a teclinique of training the self Significantly, this writing almost invariably takes the form of confession. As Foucault puts it: The self is something to write about, a theme or object (subject) o f writing activity. That is not a m odem trait bom o f the Reform ation or o f romanticism; it is one o f the most ancient W estern traditions. It was well established and deeply rooted when Augustine started his Confessions (Foucault, 1988:27).
Foucault goes on to discuss the qualities of these confessional practices which change with time and in relation to genre. Especially significant is the move from letter writing to diary writing, which marks the difference between the Stoic conscience, which is expressed and measured in the form o f deeds, and the Christian conscience, which manifests itself as a struggle of the soul. While the Stoic practice o f self-accounting is one in which faults are simply 'intentions left undone', the subsequent Christian form o f confession comes to include the ennumeration of bad intentions (Foucault, 1988:33).
The immediate result of the Christian addition is the move to encompass both actions and intentions in confessional practices. But the need to confess faults of different kinds is not the only manifestation o f the increasing variety o f techniques of the self Confessional practices themselves now come to take both internal and external or, more accurately, both imagined and real, form. To the longestablished practice of gymnasia -the training o f the self in real situations by means of traditional ascetic trials, such as sexual abstinence and physical hardship -is added melete as the careful imagining and rehearsal o f possible future trials. What is most important here is the fact that in this imagining and rehearsal the subject puts himself in an imagined situation in order to establish whether he can confront the events, and use the discourses witli which he is armed, properly and to his best advantage (Foucault, 1988:35).
According to Foucault, between this training in thought and training in reality there are a whole number of intermediary practices.
Amongst the most interesting here, because o f their epistemological implications, are those of Epictetus.
The principles o f Epictetus are perhaps the best example o f the middle ground between gymnasia and melete because they include the injunction to watch perpetually over representations.
For Epictetus, the control o f representations m eans not deciphering but recalling principles o f acting and thus seeing, through self-exam ination, if they govern your life. It is a kind o f perm anent self examination. You have to be your own censor ( Foucault, 1988:38).
If we add to this formulation the early Christian injunction that not only must the subject recall principles o f action, but must 'decipher' or assess which the bad and which the good principles arising from one's own intentions are, then we have the beginnings of a shift in the deployment and function of confession which will culminate in those 'disciplinary' practices Foucault regards as characteristic of the modem era.
The historical shift from eariy to later fonns of confession is thus imbricated in a shift in the forms of power from external to internal, a move which has as its key changes in the regimes of punishment and sexuality.
Foucault's crucial account of the implications of this shift appears in Discipline and Punish (1977a), which, when read with the essay "The Father's No" (1977b), makes it possible to show that confessional practices of a distinctly modem type are the result of the nature of power in modemity.

3, Confession and the rise of psycho power
As is by now well known, the basic subject matter of Discipline and Punish (1977a) concems a reversal in the axis of power from ascending to descending; in other words, a shift from feudal regimes in which individuation is greatest where power or privilege operate to modem disciplinary societies -where by contrast it is those upon whom power is exercised, rather than those who exercise it, that are most individualised.
In a system o f discipline the child is more individualised than the adult, the patient more than the healthy man, the madman and delinquent more than the normal and the non-delinquent (Foucault, 1977a: 193).
As a result those who vary from or who appear to flout the norm acquire focus and interest, a density and specificity that makes them prime subjects for knowledge as well as the most valued subject matter for writing.
In "The Fathers No" Foucault (1977b: 194)  The typically elliptical and elegant final proposition in this passage emphasizes the point that, in the modem era, the position of centrality previously occupied by the epic hero is now filled by non-heroic and especially fallible individuals.
The point that Foucault makes here is given its fullest form in Discipline and Punish (!977a), where he demonstrates that the shift from societies o f spectacle and punishment to modem disciplinary societies is characterised by a shift from external to internal power. In other words, he suggests that power operated in feudal societies as direct punishment imposed on the body of the offender, while disciplinary societies concentrate their energies on the production of subjects o f a certain type, ideally those capable of policing themselves.
Crucially in order to enact the high measure of self-control required, individuals are now enjoined to see themselves as constantly falling away, not so much from God or the sovereign but from their selves -those selves to which ultimate allegiance is owed. That to which modem individuals are instmcted to be constantly vigilant, is therefore not an outside principle or authority but one (necessarily potentially flawed) apparently safely housed within. Surprisingly, given the centrality of its 'psychological perception', the modem individual needs to value no particular psychological content except the possession of a unique self and the belief that this self is, and must be, his or her own to cultivate and maintain.
What lends special significance to confession then, is that it plays a cmcial role in the production o f a self believed to be a type of work or project. Because the self is always subject to error, which can be seen as the direct descendant o f Stoic weakness and Cliristian sin, it is always in need of reworking. Wliat the discourse of confession thus expresses as the celebration of the flawed, Foucault suggests, is part of that wider relation -the articulation of individuality, error and responsibility which the era of psychological power demands.

Towards an epistemology of confession
Against this general background the most important contribution to an understanding o f confession is The Order O f Things (1973) which extends the argument of Discipline and Punish (1977a) to show, firstly, why the broad structure of reflexivity in which confession participates should arise in the first place and secondly why the negative impulse which must be confessed, should home in on subjectivity and language. Both subjectivity and language now come to be seen as especially potent sources of distortion, and, cmcially inasmuch as a confessional impulse is concerned, are ofien confrised or equated.
Firstly, the broad question of the emergence o f confession in knowledge must be seen in terms o f a basic reorganization of the epistemic field characteristic of the modem episteme. More especially, Foucault sees modemity as best explained in terms of the decline or failure o f representation, which served as the overall figure in the preceding or 'classical' period -a shift which he describes as culminating in "the withdrawal of knowledge and thouglit outside the space of representation" (Foucault, 1973:242).
Kant's critique o f representation both prefigures, and is symptomatic of, the dispersion o f the epistemic field into empirical enquiry on the one hand, and mathematical and philosophical enquiry on the other. The fact that representation was once taken for granted as the spontaneous means whereby subject and object are related is reflected in the fact that as soon as representation is problematized in the nascent modem episteme the problem of the subject/object relationship soon follows suit.
This questioning of the role of representation expresses, as Gutting (1989:183) puts it: a fundamental "need for a new sort o f reflective enquiry that probes the origins and basis of the mind's powers of representing objects" . The divisions in the field of knowledge that inaugurate the modem episteme are thus due to the decline in the Classical conception of the unproblematical relation of re presentation to its objects.
Kant's critique of representation produces two new theoretical domains. Firstly, there is the theory o f "transcendental subjectivity", where the idea is that the mind constitute its own "objects of representative knowledge" (Foucault, 1973:184). Secondly, and complementarily opposed to this theory of the subject, is what Foucault points to as a transcendental philosophy of the object (or metaphysics), which emerges from the empirical sciences of life, labour and language.
However, a third position, which takes the form of an evasive detour through both of the others, also becomes possible -one in which no transcendental grounding for empirical knowledge is deemed necessary and knowledge is restricted to the field o f direct experience. This position is usually described pejoratively as positivistic.
It will be argued here that it is this position, or ratiier the space between it and the demands of the empirical which confession comes to occupy. That is, confession can now be seen as a kind of substitute grounding occupying the place which used to be quietly and confidently occupied by representation. In other words, confession plays the role of positioning the self, the empirical self, as the locus of transcendental enquiry.
Confession thus emerges as the voice o f actual experience -an attempt to unite 'positivism' (the discourse of the body, the object, and the empirical) with 'eschatology' (that o f the self, the subject, the transcendental). The discourse of actual experience is thus a 'discourse of mixed nature'. It plays the role o f an apparent mediator (because the concepts are apparently radically contesting) between "the space of the body and the time of culture" (Foucauh, 1973:321), but is actually the articulator of the two. In other words, the discourse o f actual experience enables the subject to understand itself as both an object of the empirical and to be positioned as the transcendental precondition for this objectivity.
As Foucault's renowned account rests on the understanding that man's problematical double status is based on the fact that, for modernity, man is both the foundation of all knowledges and an object of enquiry within (or for) them -both the source o f knowledge and an empirical element; a thing among things which must be known in the same way as they are.
The reason why confession thus appears in the human sciences in particular stems not, as is widely believed, from the extreme variety and density o f human experience (which is the subject matter of both confession and the human sciences), but from the essential precariousness of the epistemological form of the human sciences.
The human sciences do not, as the natural sciences do, take man in nature, that is, his biological aspect only, as their object. They are concerned, instead, with those aspects o f his existence which are inseparable from the unity o f his experience in culture: that is, from man as a living, speaking and labouring being. The difficulty in the human sciences arises because it is the same being who lives, works and speaks; who is required to know what life is; who must understand the essence o f labour and of law, and who must know how it is that man is able to speak (Foucault, 1973:353). Wliat Foucault calls "the double and inevitable contestation" between the sciences proper and the sciences of man, has its origin in an anthropologisation which continually undermines the form o f contemporary knowledges from within (Foucault, 1973:345).
In a sentence which captures tiiat whicii, in our view, links the points made so far about confession together, Foucault writes that the human sciences lay claim to the status of the foundations of the natural and physical sciences, while they, in turn, are "ceaselessly obliged to seek their own foundation, the justification of their method and the purification of their history" (Foucault, 1973:345-346).
The constant reflection of the human sciences on their own epistemological conditions is both made possible by, and manifested, in a preoccupation with the potentially problematic role of language in knowledge. The same preoccupation accounts for the problematization of language as a bearer o f truth in confessional practices themselves.
In the Classical period language occupied a privileged and unquestioned position as a unified representational apparatus, unproblematically instnunental in the organisation o f identities and differences across the grid of the classificatory table -that table which is itself the chief expression of Classical knowledge. The major shift which the modem episteme brings to this position for language relates to its new status as an object of knowledge rather than a vehicle for knowledge. This new status is manifested in two phenomena: language acquires both an empirical and a historical status. In other words, language becomes a 'positivity'.
The new grammar is im mediately diachronic. How could it have been otherwise, since its positivity was established only by a break between language and representation? The internal structure o f languages ... could be re-apprehended only in the form o f words; ... by being cut o ff from what it represents, language was certainly made to emerge for the first time in its own particular legality, and at the same time it was doomed to be reapprehensible only within history (Foucault, 1973:294).
Taking its places within history implies for language a loss of transparency and of the fimdamental position it had occupied in relation to all knowledge; it is demoted to the status of a mere object (Foucauh, 1973:296). Foucault argues that this demotion of language is, however, compensated for in tliree ways, which might be described as formalization, exegesis and literature. O f these the most important for a study of confession is the second 'compensation', that of exegesis, especially the critical value it bestows on the study of language.
Foucault points out that an important consequence o f the historicization of language is its ability to form a "locus of tradition, of ... unspoken habits of thought, of what lies hidden in a people's mind; it accumulates an ineluctable memory which does not even know itself as memory" (Foucault, 1973:297). The concomitant o f this change in the status of language is that it must also act as a site for the critical investigation o f its own workings. Tlie aim of this in vestigation would be to foreground the role o f language itself by "disturbing the words we speak, ... denouncing the grammatical habits that inform our thought and dissipating the myths that animate our words" (Foucault, 1973:298).
All of these critical activities, which make up the exegetical impulse, are not, we will argue, confined to the 'passive' realm of hermeneutics. Instead, exegesis has active consequences in fields other than the interpretative. And crucially, some o f those fields are of considerable epistemoiogical significance. If language itself is at risk from its own stale grammatical habits and potentially disruptive myths, then that variety of knowledge which only language can convey is similarly threatened. And it is this fragile condition of knowledge-in-language which requires strengthening by means of the purifying power o f the confession.

The role of language and experience
While this generalized problematization of language and representation and the double status of man acts as the precondition for the emergence o f confession in epistemological contexts, two additional characteristics of the modem episteme are required to account for the specific forms confession takes.
The first of these is modernity's commitment to actual experience, and the accompanying valorization o f its discourses. The second is, albeit indirectly, the attribution o f a specific histoiy to individuals, their experiences and their styles in language.
In other words, with the addition of the historical dimension to both language and individuals, it becomes possible to conceive of two distinct but interconnected varieties of confessional practice. The first of these has its origins in the notion of an individually inflected use of language; that is, in the recognition that an individual's style is a personal attribute like any other. But it also has the implication that this style must and can be clarified by means o f a confession that the style exists and that the particular and potentially dangerous role it plays in knowledge can be mitigated by defining it.
The second of the interconnected varieties o f confessional practice similarly focuses on language, but where the first is concerned with the consequences of an individual's style, the second is concerned with the consequences for knowledge of the genre or mode of writing which characterises that knowledge. As the titles o f the two sections o f the book "History as a Novel" and "The Novel as History" indicate, much that is central to the text's intention turns on its attempt to play upon two traditionally different modes o f writing. As a novelist or writer of fiction Mailer is of course concerned with literary or stylistic attributes -he may write in accordance with the criteria and latitude associated with a work of style -but as journalist, documenting real events, he is committed to provide a true or actual account.
What is interesting about Mailer's response to the task of fusing fact and fiction is the articulation of a solution which, we suggest, only appears to be idiosyncratic.
In meeting the challenge of melding two genres and in homing in on the differences between history and the novel, Mailer reinforces the commitment to truthful representation by attempting to purify the source or origin of those representations themselves by means of the confessional route.
In other words, as an author who is as much an historian as a novelist, he implements a critical self-examination expressed in a confession in order to explain and display the particular point of view or style believed to get in the way of the truth.
The first part of the book, entitled "History as a Novel" in which Mailer presents his experiences on the march, opens wiih an ironic play on words suggested by the tenn "protagonist" and is redolent with confessional elements. Mailer writes of his heavy drinking, his schooldays, and even his attitude to the telephone. Even the first lines which promise to bring his readers "news of your protagonist", make this confessional commitment clear and are followed by a long quotation from a Time magazine report which deals with Mailer's outrageous performance at Washington's Ambassador Theatre as part of the anti-Vietnam peace campaign.
" [I]ntimate history" (Mailer, 1968:67) of this kind continues to flavour the essentially journalistic account with the confessional ingredients. Wliat counts about this intimacy is not merely the focus on the central figure which it provides, but the simultaneous focus upon the act of writing.
Character and author merge in a double unit and soon the writer's task and that of the eye-witness, the man (literally) on the march, begin to overlap in a way Mailer hopes will bring substantial gains by retaining the best features of both fiction and journalism, of the novel and of history.
The last pages of the section explicitly spell out the confessional nature and its connection to style as personal attribute.
Then he began his history o f the Pentagon. It insisted on becoming a history o f him self over four days and therefore was history in the costume o f a novel. He laboured in the aesthetic o f the problem for weeks, discovered that his dim ensions as a character were simple; the je st had been for the novelist, for his protagonist was simple, a hero and a marvel o f a fool with more than average gifts o f objectivity. M ight his critics have as much? This verdict was disclosed by the unprotective haste with which he was obliged to write, for he wrote of necessity at a rate faster than he had eve>-written before, as if the accelerating history of the country forbade deliberation. Yet in writing his personal history of these four days, he was delivered of a discovery of what the march on the Pentagon finally meant and what had been won and what had been lost, and so found himself ready at last to write a most Concise, Short History, a veritable precis of a collective novel which here, now, in the remaining pages will seek as history, no, rather as some Novel of History that quintessentially American event (Mailer, 1968:240).
The technical, aesthetic issue of writing is solved by confessing himself as a character. Mailer (the writer) points out that Mailer (the man) is marvellously naive -the wise fool blessed thereby with 'gifts' of objectivity. Wliat Mailer does, therefore, is to solve tlie thorny issue of point o f view (and tlie threat to objectivity in knowledge that it presents) by way o f a lucky personal attribute, a distinctive and advantageous style, one which is safeguarded, what is more, by the demands made by the experience itself. It is not just uninterrupted experience, writing as quickly as experience will permit, which is the best guarantee o f truth and objectivity. When all is said and done for Mailer, however, it is the personal histoiy of those four days which makes history of another kind possible (the concise history which he is to undertake in Part Two).
Opening the second section of the book, entitled "The Novel as History", Mailer outlines what might be called the philosophical foundations of the shift which that from Part One to Part Two captures.
The Novelist in passing his baton to the Historian, has a happy smile. He has been faster than you think as a working craftsman, a journeyman artist, he is not without his guile; he has come to decide that if he would see the horizon from the forest he must build a tower, it, the horizon, will reveal most of what is significant, an hour of examination can yet do the job -it is the tower which takes months to build. So the Novelist working in secret collaboration with the Historian has perhaps tried to build with his novel a tower fully equipped with telescopes to study -to the greatest advantageour own horizon. Of course the tower is crooked and the telescopes warped, but the instruments of all sciences -history as much as physicsare always constructed in small or large error; what supports the use of them now is our intimacy with the master builder of the tower and the lens grinder of the telescopes (yes, even the machinist of the barrels) has given some advantage for correcting the error of the instrument and the imbalance of his tower. May that be claimed of many histories? In fact how many novels can be put so quickly to use? (For the novel -we will permit ourselves this parenthesis -is, when it is good, the personification of the vision which will enable one to comprehend other visions better. A m icroscope -if one is exploring the pond -the telescope upon the tow er if you are scrutinising the forest) (Mailer, 1968:259).
What is at issue for Mailer in this exemplary passage is the very possibility of Novelist and Historian working in productive collaboration and what is important for us is tlie way in which the preconditions for this productivity are seen to turn on an explicit relation to authorship. The attribute of special significance which the novelist gives to the historian is that o f a ready made clarified, because confessed, vantage point. The slow preparatory work, it seems, is not just documentation and description, but the arguing or demonstration of a particular vantage point; now quite literally a point of view.
If the historian needs to see the horizon beyond the forest, and the wood beyond the trees, he needs the extra height provided by the tower, the additional strength which the instruments provide but most important of all that unique document which only the novelist can provide.
It is not simply that this history has the advantage of having an artist as its author but that this author has provided a document of unusual worth -one that includes another history, that of its own historian and its own production, available simultaneously. Wliat Mailer sees as important and unique to this document is the self-reflexiveness of the document.
What Mailer expects this rare document to yield is revealing and, in retrospect, predictable. Wliile he recognises the necessity of both tower and telescope, an angle or vantage point which will provide a perspective on the events, it is the price paid for perspective which he finds disquieting and obviously hopes to avoid.
What worries him (like so many others) about the need for a point of view is its limitations; but, more than this, he fears its likely, if not inevitable distortions. What the "intimate" novelist gives to the "objective" historian with one hand, he could just as easily take away with the other. Mailer is preoccupied by a psychological entity, accompanied by potentially dangerous (to the truth) attributes, like personal beliefs, attitudes and values -in other words, all that is usually meant by bias.
It is not merely that he sees telescope and tower as distorting, or even that he sees these errors in the instruments as attributable to features o f their makers, but most important here is the fact that Mailer believes that a portrait of the maker -a confession o f his weaknesses and their nature -can go some way to correcting these errors at their source.
Mailer's view of writing, style and knowledge intersect around the figure o f the author so that the manner in which things are produced and known may be traced to features o f the novelist's nature or character. We may know a novelist or historian -as a self -in much the same way as we may know the lens-grinder or the tower-builder, so that both the problems and their solution may be and are thought of in the same place. The problem of the means by which content is conveyed, is equated with that o f the subject, o f the bias which inevitably accompanies the fact of having a particular character or personality. As a result a way o f doing something is closely related to the mode o f the man (the actor or author), so that the issue of subjectivity becomes linked to that o f an account of kind o f language use.
Wliat is important in Mailer's literary/historical manifesto is that it spontaneously presents a problem and solution in one place. And in this configuration he is not alone. It is this configuration, we suggest, that underlies the confessional elements common to a number of human sciences and in fact it would probably be hard to find an aspect of the human or social sciences where it does not appear.

Confession in ethnography
If most of the human and social sciences display some variety o f confessional practices, amongst the most common of these are self-reflexive approaches which focus on the role o f language and subjectivity in the production o f language. And of these it is probably in ethnography that the relations between language, writing and knowledge are most explicitly articulated.
For example in the introduction to the influential collection Writing Culture, James Clifford goes so far as to talk o f there being a sub-genre of ethnography which he calls self-reflexive field work, a genre which he describes as both "confessional and analytic", one which provides, in this "important forum for the discussion o f a wide range of issues epistemological, existential and political" (Clifford, 1986:14).
In fact the textual approach of Writing Culture (1986) itself is exemplary o f an orientation in which both the role of the author/researcher and that of the writing (both as genre and rhetoric) has moved to a central and problematized position. Clifford's substantial introduction is predominantly concerned with these issues o f subjectivity and authorship, emphasizing positively the importance o f con sidering the role writer and writing play in the production o f ethnographic knowledge. Like Mailer, Clifford argues that the recognition o f an author's point o f view results in the ceding of authority but only in order to gain epistemological strength: Ethnographic truths are thus inherently partial -committed and incomplete. This point is now widely asserted -and resisted at strategic points by those w ho fear collapse o f clear standards o f verification. But once accepted and built into ethnographic art, a rigorous sense o f partiality can be a source o f representational tact (Clifford, 1986:7).
Siinilarly, in the conclusion to his owii paper in the collection (which deals with allegory in ethnography), he writes that the recognition of the role of allegory in ethnographic writing is necessary in order to raise questions about the political dimensions of the discipline, which he suggests should be "manifest not hidden" (ChfTord, 1986:120).
Given this skepticism about the means (writing) and source (the subject) of representation, it is hardly surprising that the unknowability of the subject should become transformed into a strategic defence, in which the relativity of knowledge itself, if confessed, becomes a new fonn of veracity.

Conclusion
Wliat emerges most forcefully from this preliminary sketch for a genealogy of confession is the extent to which confession's presence within the human sciences and its imbrication with issues of language and knowledge is, retrospectively at least, entirely expected. Wliat is less expected is that the widespread persistence of confession in epistemological contexts should so oflen be seen as radical or progressive. The genealogical questions of who confesses and how, and their implied archaeological concomitant o f what discourses are privileged vehicles for confession, have been partially answered to the extent that we have shown how human scientists use confession as a form of epistemological defence.
However, questions of a more specific nature remain to be asked. Wliile we have sketched an account of confessional forms and confessing subjects, we have not examined the content of these confessions in any detail. Nevertheless, what is clear is that confessional forms represent a complicity with the relations between writing and knowledge detennined by the archive of modernity, rather than, as is so often believed, a transgression of them.