This paper explores postmodern social theory approaches to research writing. In particular I am interested in the possibilities which an emphasis on intertextuality can offer in the production of social research. Intertextuality is the process whereby one text1 plays upon other texts, the ways in which texts refer endlessly to further elements within the realm of cultural production (Barthes, 1977). Intertextuality is a feature of every text. However I shall suggest here that devices which consciously enhance and emphasize intertextuality contribute a re-reading of the relationships between the social, the writer and the reader, between researchers and researched, students and teachers, theorists and practitioners. I shall argue that intertextuality is a means to demonstrate the limits of discourse, but also, significantly, a stratagem by which it becomes possible to challenge and resist discourse - to open up the possibilities of becoming other (Bogue, 1989; Curt, 1994).

The concept of intertextuality articulates with post-structuralist perspectives on language and knowledge (Game, 1991: 18). Postmodern analyses challenge the ontological status of modernist claims to knowledgeability concerning the world. Consequently, when such approaches are applied to social theory, the privilege which has been claimed by modernist social scientific discourses is dissolved (Atkinson, 1990; Bauman, 1988; Butler, 1990; Fox, 1993b; Game, 1991; Rosenau, 1992; Silverman, 1987; Stanley and Morgan, 1993; Tyler, 1986). From within a framework of postmodern social theory, an interest in writing and intertextuality rejects distinctions between "real" and representation (Stanley and Morgan, 1993; 3). All texts, in this view, are fabrications and as such are subject to deconstructive re-writing and re-reading. Social science texts, like any others, are to be read and re-read, not as representations (accurate or flawed) of the social world, but as contested claims to speak "the truth" about the world, constituted in the play of disciplines of the social. Social research writing becomes a narrative in which, as Maines (1993; 17) has put it,

... sociology's phenomena are seen as significantly constituted by stories and in which sociological work is seen as narrative work.

Before exploring some ways in which intertextuality can contribute a richness to our writing of social research, I want to situate intertextuality more firmly within the context of some poststructuralist and postmodern positions on knowledge, power and resistance. These positions are not intended as "justifications" of an intertextual approach, but I think provide a useful theoretical context to the discussions.

DERRIDA: INTERTEXT AND DIFFERANCE

My first reading concerns Derrida's analysis of differance. Briefly, differance concerns the fundamental undecidability which resides in language and its continual deferral of meaning, the slippage of meaning which occurs as soon as one tries to pin a concept down (Derrida, 1976; 65). Differance is unavoidable once one enters into a language or other symbolic mode of representation in which signifiers (words, signs) refer not to referents (the "underlying reality") but only to other signifiers. While trying to represent the real, one finds that the meaning which one is trying to communicate slips from one's grasp. We are left not with the reality, but with an approximation which, however much we try to make it "more real", is always already deferred and irrecoverable (Finlay, 1989).

This theory of differance supplies the basis from which Derrida criticizes logocentrism. Derrida argues (1976) that claims to be able to achieve the logos, an unmediated knowledge of the world, are a feature of every discourse which seeks to explain the world, be it philosophical, religious or scientific. Such claims to have a route to such unmediated knowledge simply put being able to speak "the truth" about something or other, is the basis for all authority and authenticity. Such claims work by privileging certain aspects of the world at the expense of others. Thus, in the modern period, social and human sciences have gained legitimacy (although not incontestably) for their particular claims to knowledge of reality. But this logocentrismon the part of social science works only by a denial (or bracketing) of other competing claims. As such it is also a denial of intertextuality

Re-introducing a recognition of the intertext is implicitly a critique of sociological logocentrism. As such, it clearly challenges sociology's privilege to speak authoritatively about "the social". As Game (1991;18) has it, sociology's fiction is that sociology is not fiction. But I would suggest that at the same time, this analysis opens up the possibilities for a social theory which is no longer obsessed by efforts to attain some kind of (semi) transparent mediation of knowledge of the world by the human observer (Flax, 1990; Hutcheon, 1989). If no privilege is attached to particular discourses, social theorists may explore far more widely texts which contribute to the fabrication of the social. In short, it proffers a new richness of "data" of the social, fabricated in intertextuality: the play of text on text in novel and unending combinations of differance.

FOUCAULT: INTERTEXT IN GENEALOGY

Perhaps the fundamental difference between traditional historical accounts and the genealogical method developed in Foucault's writing (1976; 1979; 1984) and based in Nietzsche's explorations of a genealogy of morals (Lash, 1991; 260), is that while the former emphasizes continuity and the logic of events in terms of cause and effect, genealogy discloses discontinuity and the continual writing and rewriting of the world in discourse. As such, genealogy is based in the intertextuality just described. The violence of discourse, as it plays out its games of power in the claims it makes to know the world, is the subject matter of the genealogical method. Rationality is no longer the force behind the evolution of knowledge. Rather, knowledge is fabricated through the neverending struggles for power to describe the social. Medicine, psychiatry, sex therapy and other caring disciplines construct human subjects in ways which supersede earlier versions, not for reasons of rationality, but for strategic objectives concerned with power and control.

Genealogical texts can claim no privileged status for themselves. They are the latest in the line of accounts which fabricate their subject. This reflexivity which for the modernist signifies the bane of relativism emphasizes the intertextual character of genealogy. Genealogy does not provide the truth of how one discourse supersedes another. Indeed, it admits itself to be incapable of anything other than one version of this sequence, nor of being able to discern the reasoning (or its absence) whereby it became possible to speak about the world in particular ways at particular points in time and space.

Thus genealogy is a method of intertextual exploration, which while it must fail (but does not display anguish at its failure) to explain the progress of ideas, offers insights into the technologies of power/knowledge which inform discourses. The "truth" of pan-opticism (Foucault, 1976) cannot be proven from genealogy. Yet it provides us with an idea with which to toy as we explore the texts of medicine, education or industrial organization. Genealogy is thus fundamentally intertextual. Seen in this light, it is clear just how radical is the Foucauldian poststructuralist method and how great a challenge it poses to modernist social theory.

Yet, one is also struck - reading many genealogies by how their authors' selections of writings have continued to privilege particular kinds of texts: usually those which appear to possess some kind of authoritativeness or representativeness, to have been published, to be academic or "official". Such genealogical writing should perhaps document discourses which failed to become prominent, but nevertheless could still be thought/written/dreamt, to finally reject the modernist privileging of particular kinds of text (scientific, nonfiction, high culture and so on) over others.

CIXOUS: WRITING, GENEROSITY AND THE INTERTEXT

Genealogies which document the discourses which make it, document the ways that power and authority have fabricated reality, but rarely the resistances to domination and power. In this section and the next I want to look more closely at the political potential of intertextuality, firstly as illustrated in the writing of the feminist poststructuralist Helene Cixous (1986; 1990), who has emphasized the significance of writing as resistance. In Cixous' position on feminine writing (ecriture feminine), the relationship with intertextuality is clear. Ecriture feminine is a writing practice which is concerned with the openness of texts and multiplicity, in place of closure and univocality (Game, 1991; 80), and as such is closely related to the Derridean concept of differance examined earlier. Feminine texts strive to undermine oppositions, to deconstruct textuality itself.

The politics of resistance in ecriture feminine rests in it being not "just writing," but also a "writing of the body," which is the locus of resistance to (patriarchal) discourse.

By writing her self, woman will return to the body which has been more than confiscated from her, which has been turned into the uncanny stranger on display.... Censor the body and you censor breath and speech at the same time. .... To write. An act which will not only "realize" the decensored relation of woman to her sexuality, to her womanly being, giving her access to her native strength: it will give her back her goods, her pleasures, her organs, her immense bodily territories which have been kept under seal tear her away by means of this research, this job of analysis and illumination, this emancipation of this marvelous text of her self that she must urgently learn to speak. (Cixous 1990; 3201)

If this is to romanticize "the feminine", then perhaps a "feminine text" might more generally be defined as one which encourages a play of textuality, which will deconstruct its own claims to authority and authenticity. Feminine texts are not necessarily the product of women (Moi, 1985; 108), although Cixous (1990; 318) writes

I write woman: woman must write woman. And man, man .... it's up to him to say where his masculinity and femininity are at: this will concern us when men have opened their eyes and seen themselves clearly.

I would suggest that what is argued for as ecriture feminine by Cixous is concerned implicitly with the intertextual: the play of one text as de-constructive of another, itself in turn dissolved or reread. Cixous elsewhere (1986) makes a distinction between "Gift" relations based in generosity, and the those of the "Proper", based in control and appropriation. Ecriture feminine relates to the former category, offering not identity and repetitions, but trust, love, affirmation, confidence and a generosity of spirit. It challenges such "masculine" concerns as property and propriety, and can "... shatter the framework of institutions .... break up the 'truth' with laughter" (Cixous, 1990; 326).

There is a continuity here with post-structuralism's concern with difference in this feminist politics of resistance. Unlike philosophies of resistance, including Marxism and modernist feminism, which have grounded their logic of resistance in identity (membership of a class or a gender), a concern with the intertext is a celebration of difference and the possibility that things can be otherwise. Such a politics of difference goes far beyond feminism and suggests the potential for resisting discourse, knowledge and power through intertextual practices. In the context of writing social research, it replaces objectivity with indeterminacy and the search for control and closure with generosity and openness.

DELEUZE AND GUATTARI: RESISTANCE AND INTERTEXTUALITY

The conception of a feminine, openended, generous text which can be the vehicle for resisting rather than inscribing discourse resonates with the work of Gilles Deleuze and Felix Guattari whose "schizanalysis" similarly argues the possibility of resistance (Deleuze and Guattari, 1984; 1988). These writers conceive of human subjectivity as inscribed upon a philosophical surface which they call the Body-without-Organs (BwO). With Foucault, they see this surface as "totally inscribed by history" (i.e. the discourses of "the social"), yet they break with Foucault's ontological skepticism concerning the possibility of resistance to the subjectivities thus inscribed (Lash, 1991; 270). For them, the BwO can be "deterritorialized" by what they call "positive desire", and for subjectivity to become "nomadic", if only for a moment untrammeled by discourse, able to become other.

My reading of this difficult and problematic concept of positive desire 2 focuses upon Deleuze and Guattari's interest in the process of enabling another human being which we can supply in investments of care, love, trust, giving whether these be material, affective or spiritual, and leading to some kind of "becoming other", liberation or emancipation from discourses of control and domination. Understood in this way, "de-territorialization" is equivalent to the objective identified by Cixous in the investments of the Gift relationship: a breaking free, an opening up to new possibilities to act and think 3. In place of the discourses of the social, which create identity and repetition, positive desire blasts the BwO out of the subjectivity which has been inscribedupon its surface by the social.For a moment, the subjectivity is freed to wander upon the BwO the "nomad" subject becomes other. Then it settles back, but maybe in a new configuration, as the process of signification seizes upon a new patterning of the intensities on the surface of the BwO. Sometimes, this deterritorialization may move the subjectivity to a new "plateau", a more "radically other" subjectivity.

I want to suggest here that intertextuality, understood in a wider sense than hitherto developed in this paper, is at the heart of this deterritorialization of subjectivity and the freeing of the nomad subject. Deleuze and Guattari's book, A Thousand Plateaus (1988), recognizes the significance of the intertext. It is a book designed as a "rhizome", ceaselessly achieving multiplicity, refusing to follow a single chain of signification (1988; 79). As they write (1988:22),

We call a plateau any multiplicity connected to any other multiplicities by superficial underground stems in such a way to form or extend a rhizome. We are writing this book as a rhizome. It is composed of plateaus. .... Each plateau can be read starting anywhere and can be related to any other plateau.

Their objective maybe is to orchestrate enough intensities upon the reader's BwO that new connections become possible, a new plateau is achieved, so the reader can say "So that's what it's about". (Is this the secret of all good books?) Deleuze and Guattari's (1988: 11) book is also explicitly an incitement to write:

Conjugate deterritorialized flows. Follow the plants: you start by delimiting a first line consisting of circles of convergence around successive singularities; then you see whether inside that line new circles of convergence establish themselves, with new points located outside the limits and in other directions. Write, form a rhizome, increase your territory by de-territorialization, extend the line of flight to the point where it becomes an abstract machine covering the entire plane of consistency.

Read as an argument for intertextual practice, Deleuze and Guattari's notion of positive desire is shriven of its foundationalism (maybe even biologism). But it also makes the vital point that intertextuality; it is not simply some dry academic or intellectual claim concerning the power of the pen. Rather, writing is oneway to become other, and intertextuality is about the interplay of all kinds of texts, not just written, but the range of meaningful activities in which human beings engage, and which if constituted in a relationship of the Gift, celebrating love and difference - can be a force for liberation.

SOCIAL THEORY AND INTERTEXTUAL PRACTICE

Thus far I have drawn (intertextually) upon various postmodern writers to explore aspects of intertextuality and its relations to the social and to resistance. Derrida supplies the position from which to examine authenticity, authority and authorship, while Foucault's genealogies document changing authenticities. Cixous, and Deleuze and Guattari, from different theoretical positions, both explore the inscription of the body by the social and ways in which this inscription can be turned away from control and towards resistance and becoming-other.

I want to turn now to what these aspects of intertextuality mean for the practice of writing social research. Many social scientists (and maybe natural scientists too I don't know) will have had the experience of only finding the interpretation of their research data or their theoretical reasoning during the process of writing up for publication, report or thesis. Similarly, many of us will know, either as authors or referees, the subtleties entailed in fabricating authentic accounts for journal publication (I use the term "authentic" here in a postmodern way to mean credible, convincing and/or internally coherent). Ethnomethodologists might call this "doing getting published".

Postmodern writing has begun to unpack some of these fabrications. Ashmore, Mulkay and Pinch (1989) provide an interesting analysis of health economic writing which is at the same time reflexive about its own production within the framework of the sociology of science. Press cuttings concerning the arguments over the ethics of QALYs (Quality Adjusted Life Years) a health economic instrument are followed by a spoof press report about the authors' own work. Elsewhere different readings (economic, sociological, lay) of the book's conclusions are juxtaposed.

Ashmore et al (1989) also adopt the strategy of introducing a "commentator" who engages with the authors in dialogue. This strategy to enhance intertextuality was also used by Mulkay (1985). In addition, this author included a "oneact play" as part of his explorations. Curt (1994) utilizes an "interrupter" while I brought in "another voice" (Fox 1992) to introduce a fragmentation into an otherwise linear text. Curt's writing is also marked by the fact (?) that the author possesses congenital acorporeality: "Beryl C Curt" being the pseudonym for a group of writers. Beryl herself engages with the text as a further dissonant voice. Tyler (1986: 125), in postulating a "postmodern ethnography", advocates

... a cooperatively evolved text consisting in fragments of discourse intended to evoke in the minds of both reader and writer an emergent fantasy of a possible world of commonsense reality, and thus to provoke an aesthetic integration which will have a therapeutic effect.

Other authors who have reflected on the production of sociological texts include Atkinson (1990), whose deconstruction of ethnographic writing exposes the intertextuality within such reports and the efforts made by writers to obscure its social production. Atkinson (1990; 106) points to the narrative element in some such writing and while he is concerned to demonstrate how this is a strategem for fabricating authenticity, Maines (1993) sees in narrative an opportunity for sociology to explore a wide range of texts in its research. A conference and subsequent collection of articles on the topic of Auto/biography in the journal Sociology similarly identified the stories told about people and about themselves as valuable research data for sociology [see papers by Aldridge, Cotterill and Letherby, Davies, Erben and Stanley (all 1993)].

Taking the further step of accepting that all there are are stories is a hard one for modernist sociology to swallow. But this is the necessary lesson of intertextuality and one which perhaps can move writing social research into the arenas of resistance and liberation discussed above. So, to complete this discussion, and to suggest the possibilities of a self-conscious intertextuality, I will now offer a brief piece of "data" from my own work to serve both as a substantive contribution and, more importantly, as an example of how the intertextuality within writing social research/theory can deepen and broaden the experience of the phenomenon being written.

With the benefit of this illustration of a more open-ended approach to writing research, readers may wish to reflect further upon the various perspectives explored in the first part of the paper, in particular upon writing as resistance, and upon the need for trust within the research community to enable such writing. However, with the desire for open-endedness in mind, I shall not attempt subsequently to draw everything neatly together to supply closure, the rightbalance of theory and data or any other attribute of modernist social science writing.

RESEARCHING SURGERY: TOWARDS ENGAGEMENT

During the late 1980s, I undertook an ethnography of surgery at a number of UK hospitals. This was later published (Fox, 1992) as a monograph and continued to be recycled in various forms as part of my efforts to explore postmodern social theory (Fox, 1991; 1993a; 1993b). Reflecting upon this research some years on, I think that a richness can be brought to this kind of writing of the social by exploring the play of a range of texts in addition to the straight ethnographic report which I originally produced in an effort to describe events in the operating theatre or on the surgical wards.

The first text I draw on in this exercise in intertextuality is the fabricated field notes which formed the basis of the ethnographic report. For example, here is an extract which formed part of a discussion of the organization of the surgical list.

10.30 a.m. There is a delay between patients. Anaesthetic nurse: "Orthopaedic surgeons are the worst, they arrange things at the last minute, and then they're not organized properly." Researcher: "Why is that?" A.N.: "They don't communicate. It's probably because most of them are foreign they don't understand each other." The delay continues. The registrar has been sent to look for a patient who cannot be found, but then is found in a different ward. In the meantime another patient has been added to the list a 16 year old accident victim, who has had his pelvis pinned a week earlier, and now is to have the pins out. However he has not been seen that day by the house doctor and he has not signed a consent form, and may not have been starved prior to general anaesthetic. The surgeon and registrar are not happy: the registrar is sent to the ward to sort things out. Fifteen minutes later, the registrar returns. Registrar:" He had been consented. The staff nurse thought the age of consent was 18, but he's signed himself. His father has been waiting around to sign a consent." Surgeon: (to researcher)" This is the sort of thing that happens. The consultant tells the staff nurse who tells the houseman, and the houseman forgets or is too busy." 11.20. a.m. The patient finally arrives, and is very upset as he does not want a general anaesthetic after one the previous week had made him sick. Eventually he is persuaded by the surgeon, in the anaesthetic room. (Fox 1992: 378)

Extracts such as this formed the "official" text in which I attempted to represent what I perceived as the part of the social which is called "doing surgery". Methodologically, its claim to authority rested upon the representation of observations conducted during fieldwork. It is hardly necessary to make the point that the relationship between the observed and this account of it is the outcome of interpretative and representational work which renders it no more than a plausible fiction (for cogent critiques of the relationship between the social and such accounts, see Atkinson, 1990; Tyler, 1986). But, while I was writing up these notes, I was also reading a variety of texts which were relevant to the phenomena I was studying. Some, like Katz (1984) and Atkinson (1981), were social science ethnographies. But I was reading fictional accounts of surgery alongside and the televising of Colin Douglas' book The Houseman's Tale coincided with the fieldwork. Here is an extract from another of Douglas's books, Bleeders Come First(1979).

The nurse held Miss Warrender's arm back as Campbell gripped the wrist with both hands, palms downwards and pulled. There was an uneasy, crunchy feel as the ends of the fracture came apart, then he bent the wrist backwards and outwards a little and let the lower fragment of the broken bone settle gently back into place, or rather into what he hoped was its right place. There was soft tissue swelling round the site of the fracture, which made the result of his efforts harder to assess, but in general it was a betterlooking wrist than it had been. the hollow and step had gone, and if there was any resemblance not to an oldfashioned dinner fork, it was now only a very slight one. "That might be it," said Campbell. "Like to feel it, nurse, before we put the slab on?" "Yes please." The nurse ran her fingers over the straightened forearm. "I see." The gasman, holding the mask on the patient's face with one hand, and keeping a finger on the carotid pulse in the neck at the same time, was beginning to look impatient. (Douglas, 1979; 89)

The fictional patient subsequently had a cardiac arrest on the operating table and died, despite having come to casualty for a very minor procedure. Such tales, (and more recent visualizations such as the TV fictions ER and Cardiac Arrest) paradoxically bring back to me now some of the "realness" of surgery - the blood, the smells of the operating theatre, and the tragedy of individual cases which go badly. Fictional writings bring together elements which do not necessarily coincide conveniently in ethnographic observations of reality. Many of these elements which fabricate that part of the social called surgery are consequently lost in my official account. For instance, Douglas' text reminds me not to overlook the uncertainty involved in surgery and the unpredictability of patients' responses to surgical interventions. Such reminders serve as a counterpoint to the standard analyses of power imbalances and control in sociological readings of the medical, and offer further readings for the kinds of chaotic organization which I described (Fox, 1992).

Perhaps other elements are submerged or denied. Looking at a third text related to the study suggests such possible absences. This text is my own fieldwork diary which documented my impressions and thoughtsduring fieldwork and in which I reflected on my own experience of sometimes distressing events, or merely the embarrassments of being a supernumerary. Here is an extract (never before disclosed) concerning an episode in neurosurgery.

(Date label) After the fiasco yesterday I was relieved that everyone seemed to know who I was. The neurotheatre complex had a very different feel to it different layout from the others, and seems more cut off from the rest of the hospital. This needs working on. ...... The case which followed really affected me. It was a boy of about 10 with a tumour on his brain, and some injury to his neck. He was in pain and was crying because he had been hurt while being brought from the Children's Hospital in an ambulance. It was really awful, and I walked out of the room until he was unconscious. This is the first operation on a child I have seen, and it made me realize what people go through in surgery. I cannot bring myself to write this up in the log, and don't know if I want to include it in the thesis.

And indeed, this case was never written.

I have suggested that reading these various texts together supplies a new richness to the exploration of this aspect of the social. Many other texts could be introduced, and the process of intertextual reading is consequently never ending. But for the sake of this short discussion, what can be made of the play of these three texts? Certainly something further in terms of "evoking" (Tyler, 1986) surgery is made available for reading and writing. But also, I can see here my desire to produce an objective account, my inability to cope with some of the emotional aspects of surgical care, maybe suppressing or excluding exploration of how surgical staff cope, a need to explore how staff engage with their patients. These readings are all available for fabrication, and they in turn constitute new texts (for instance - this one). The potency, for the writer or the reader, inheres in the possibility always to read another text, or to reread it in a new way. This text is a version of a paper given at the British Sociological Association conference, and a further text exists in the form of a taperecording I made of my presentation at that conference. And as I write this, I can read my efforts back then to reintroduce a "lost" emotional aspect of my research as part of a more general recognition of the suppression of the emotional in my life, which I was facing up to at the time the conference paper was written. There can always be a further text.

CONCLUDING REFLECTIONS

I wish to conclude (stop writing) this paper with some brief thoughts on the implications for social theory of admitting as "legitimate" a much wider and more eclectic range of texts, in an effort to read and to write the social. I would suggest that:

1. Intertextual approaches break the distinction between researcher and researched, in as much as the researcher becomes part of the social which is being explored. The contribution of social theory to fabricating the social which it once claimed to dispassionately describe and explain, must be recognized. Distinctions between the personal and professional responses of researchers in field settings are elided. The researcher may find her/himself in situations where it is unclear if the relationship with social actors is one of research or of therapy (Darbyshire, 1991).
2. Consequently, the social theorist must adopt an ethical and political position which will structure the engagement which s/he has with the social. A commitment to intertextuality is also a commitment to difference and to becoming Other. The politics of intertextuality and the postmodern are radical and concerned with resistance and change.
3. The significance of writing research reports changes from efforts to represent or to persuade, to a reflection upon the relationship between that text and other texts, to the possibilities of deterritorialization. Social theorists may choose a fictional genre in preference to factual accounts, where this seems to offer the greatest potential to question the discourses of the social. At other times, texts in the form of social practice (teaching, therapy, protest, worship, prayer) may be chosen in place of a narrower sense of "writing". Whatever form is chosen, the text becomes the subject and the object of social theory.