Problematizing Psychotherapy: The Discursive Production of a

Bulimic

Michael Guilfoyle St Senans Hospital. Ireland

Culture & Psychology Vol. 7(2): 151-179

Abstract: This paper explores the discursive production of a psychologised bulimic subject. Two processes are highlighted through a case study, both influencing the production of the bulimic: therapeutic operations of power; and the subjugation of non-psy accounts of bulimia. Power mechanisms in therapy encourage the client to construct a complex psychological subjectivity, enabling a psychological, self-contained account of her eating disorder, thereby facilitating 'therapeutic' change. However, the condition for therapy is the disguised subjugation of client, lay and erudite non-psy accounts. The concealment of power operations reinforces psy's hegemony in defining the person-and the bulimic-in western culture. After problematizing psy discourses, a non-psychologised feminist discourse is hypothetically considered, and dialogue with this discourse suggested. A feminist discourse does not require ideals of self-containment, nor complex psy accounts, but nevertheless offers the bulimic a range of political subjectivities as a discursive priority, rather than psychological complexity.

Key Words bulimia, discourse, feminism, Foucault, psy, subject

Psychological knowledge has been inextricably linked with the production and complexification of the individual in western culture (Rose, 1989, 1994). The revelation of the individual as historically and culturally constituted (e.g. Sampson, 1989, 1993) has highlighted the need for investigations into the knowledges and strategies employed to construct ourselves, as well as others, as certain types of human beings. In this paper, I intend to address the strategies through which psychological knowledge may enable a person's construction of herself as a psychologised bulimic.

The psychologisation of the self is enabled by the cultural avail- ability of psy discourses (Rose, 1994): the vast, although disparate, body of knowledges-articulations, analyses, theories and practices produced by the disciplines of psychology, psychoanalysis, psychiatry and psychotherapy, informing the contents and strategies of self- knowledge and self-government in the western world. This follows Ingleby's (1985) labelling of the 'psy-complex': 'the network of theories and practices concerned with psychological governance and self- reflection in Western culture' (Parker, 1998, p. 68). However, I will demonstrate that the public appropriation of psy discourses is a complex process involving important transformations. Thus, we cannot expect that a diagnosed bulimic or her family will construct this problem in the same way as so-called 'expert' practitioners might. While this may be seen as a matter of degrees of knowledge, it is also a matter of politics, fostering the privileging of expert knowledge, and the minimisation of lay accounts. Thus, as I will demonstrate in a case study, a person with certain eating practices may be conscripted, via relations of power and strategic persuasions, into a subjugation of her own account of herself, allowing for a more complex psy account to emerge.

This paper attempts a constructionist, discursive account of the production of a bulimic subjectivity. In contrast, psy discourses have typically developed knowledge of bulimia in what may be described as a representational sense. Representational accounts focus on the bulimic as a pre-discursive subject, independent of the theories and knowledges developed about her. (l) We thus have knowledge about the internal (psychological) world of the bulimic, her family relations, and the influence of gender relationships and culture on her psychology (some of these knowledges will be elaborated below). While attempting to account for her behaviour in psy terms, these discourses obscure from view the impact they have on other discourses and on the persons they discuss. I propose to use a discursive account, based on some of the ideas of Michel Foucault, to problematize, or reveal as problematic, aspects of this impact. Specifically: (1) I explore how psy's development of knowledge of bulimia implicates a subjugation of lay as well as other professional (e.g. feminist) accounts; and (2) I argue that this subjugation enables (psy discursive) 'therapeutic' strategies and operations of power to persuade persons to discourse and practise them- selves in psy (rather than in such disqualified) terms. Both of these processes are seen as relevant to addressing the question posed by this paper: how might a person come to construct herself as a bulimic?

This explication moves towards questions of broader significance: What is the relationship between our strategies, as psychologist- practitioners, and mechanisms of power in psychotherapy? What social, cultural and historical conditions have made possible these operations of power? And then, arguing that aspects of feminist dis- courses have been subjugated by psy, I touch on the following question: What opportunities for empowerment and political agency might be missed by the construction in psychological terms of people and their problems of living? Further, what implications might a dialogue with alternative discourses, such as feminism, have for psy and its practitioners?

This examination is not in any way an attempt to develop a new or existing theory of bulimia. Rather, the aim is to highlight the subjugating effect of the application of psychological knowledge, by and with respect to a person who has been thus labelled in her social network, as well as by professional mental health workers.

The Discursive Production of Subjects

There has been much debate over the theoretical position of French historian and philosopher Michel Foucault. While space restrictions will not allow for a thorough explication of Foucault's complex, changing scope of study, it nevertheless becomes essential to define the Foucauldian ground on which we shall stand for the following pages. I propose to focus this study on three concepts: discourse, power and subjectivity. Foucault's arrangement of these concepts shifted over time, and so the relative discursive priorities to be employed in this study must first be elucidated.

I want to begin this clarification by invoking Foucault's interest in 'the constitution of the subject as an object for himself [sic]' (Florence, 1998, p. 461). Foucault assumes no a priori human subjectivity. Nevertheless, the individual is considered capable of being active in her or his constitution, relating to her- or himself as an object to be known, and capable of resistance against forms and forces of domination. According to Fish (1999), this was a relatively late development in Foucault's writing and interviews, which had previously suggested that humans were docile subjects, positioned in and thereby dominated by governmental discourses and practices. This turn is a crucial development, even if not fully articulated by Foucault, as Fish claims. The subject constructs her or his position within a network of power relations, and within discourse.

For Foucault, we cannot separate what people say from what they do. And so discourses may be defined as 'practices that systematically form the objects of which they speak' (Foucault, 1972, p. 49). They are not just words to describe things, but are in themselves ways of doing, allowing for certain actions and not others. In the process of their articulation and practice, they construct their objects, delineating what can be seen, spoken about and acted upon. This productivity is made possible by the strategies and mechanisms of power associated with discourse. For Foucault (1980), power 'produces things, it induces pleasure, forms knowledge, produces discourses. It needs to be considered as a productive network which runs through the whole social body, much more than as a negative instance whose function is repression' (p. 119). Thus power is not something to be held or owned, but is exercised throughout society. It can produce subjects and subjectivities, in and through discourse.

Thus, people can be positioned (and position themselves) within dis- course, affording what may be called subject positions. These positions exist within a network of power relations, allowing for the constraint or facilitation of certain experiences and actions (Willig, 1999, p. 2). However, the operations of power involved in the production (and transformation) of the subject and its position within discourse are to some extent concealed. According to Foucault (1990), the concealment of these mechanisms is a necessary condition for power's success, and makes its use tolerable. On this account, the concealment of power's techniques reduces the possibility of resistance, while its exposure might evoke resistance and confrontation. This point is pertinent to this paper, and is relevant to an understanding of the strategies and technologies of psychotherapy.

Despite this concealment, however, resistance is possible. For Foucault (1997a, p. 292), there is always a 'degree of freedom' in power relations. The potential for protest is immanent in power relations, and without it transformation could not occur. The forms taken by resistance can vary considerably, each specific to the situation against which it emerges (Foucault, 1997b, p. 168). In the case of psychotherapy, then, the client has some choice regarding which discourse to situate herself within, and is able to constitute herself via what Foucault (1997a, p. 291) called 'practices of the self' or 'technologies of the self': she is able to work on herself, think about herself, and act in relation to herself, but always from within discourse. She is able to position herself in alternative discourses to the psy discourse employed by her therapist, psychologist, doctor or psychiatrist. For example, she may assert that her being late is not a matter of unconscious or transference resistance, as suggested by her therapist, but a matter of having to find someone to take care of her children. These different discourses imply different subject positions-the one as 'psychologically resistant', the other perhaps as 'protective parent'. Her assertions are elements of practices of the self, through which she discursively constructs her experience and her identity. Foucault (1997a, p. 284) calls these points of resistance 'practices of freedom'. They constitute protests against the subjection proposed by a particular, unwanted discourse.

The client's refusal to see her behaviour as psychologically inspired or determined is important, as it is here that she begins to assert her own mode of self-knowledge in contrast to the psychologist's construction of her conduct. Such refusal, or saying no, is 'the minimal form of resistance' (Foucault, 1997b, p. 168), and it can be enhanced by its elaboration into a new self-discourse. In this way, resistance also has the potential for a more positive, creative form, enabling the generation of new discourses and practices (Foucault, 1997b). The client's assertions could then mark the beginning of a new self-discourse, elaborating, for example, on how caring for her children assumes primacy over the therapist's interpretations regarding time keeping. Such developments undermine the interpretation of psychological resistance. It is important to note that this does not remove power from the therapist-client interaction, since power is integral to the social arena. Rather, the client's refusal has the potential to produce a different set of power relations between herself and therapist, via the construction of an alternative discourse permitting different subject positions. Importantly, her refusal also threatens to expose hitherto concealed mechanisms of power in therapy.

However, psy practitioners are trained to employ strategies to address a client's refusal to psychologise her conduct, thereby limiting the possibility of shifting power relations via the subjugation of new, non-psy discourses. Thus we may obscure the political practice of refusal by investing it with psy discursive meanings. Turning political refusal into a psychological matter is an effect of power, and yet it is through this transformation that the mechanisms of 'therapeutic' power become concealed. Refusal becomes discoursed as a psycho- logical process within one person (the client), rather than a political struggle between two persons. Thus transformed, psychological resistance is both expected and well researched (see, e.g., Garfinkel, Garner, & Kennedy [1985, p. 347] on eating disorders and the 'refusal to co- operate'). Perhaps, then, as the rhetorical work of therapy is conducted, the client starts to question why she hadn't found a child-minder earlier in the week, and begins to wonder if she might indeed be' angry' (or whatever psy discursive object is made visible) with her therapist after all. In this case, her political resistance (to a psy discourse and its power mechanisms) is rendered mute, the political strategies of therapy become obscured from view, and she begins to construct herself as an object to be known through, hence becoming subject to, the psy discourses. She thereby constructs aspects of a new, altered psychological identity for herself. As her questioning continues, she develops her thinking about herself in psychological ways, and in so doing she begins to be tied, and comes to tie herself to this identity 'by a con- science or self-knowledge' (Foucault, 1982, p. 212).

It must be noted that such an outcome is by no means inevitable, and resistance to such subjectification could continue. However, this would constrain the extent to which therapy could proceed, given that psychotherapy requires 'a kind of self that would be able to appreciate that discourse and act in an appropriate way' (Parker, 1998, p. 66).

The political nature of the client-therapist dialogue must be under- stood within the context of a (psychotherapy) relationship that is itself structured by the western, contemporary institution of psy and all the operations of power it implicates. Psy enjoys a position of privilege with respect to its (therapy) clients, partly by virtue of its subjugation of alternative, non-psy accounts of the person and her conduct (the matter of psy's subjugation of other discourses via their incorporation into its domain will be discussed below). This privileged position enables a relation of power with regard to the client, which will differ from that producible within the lay non-expert community. But it is not only psy experts who have access to psy discourses. Psy discourses have permeated into common-sense, western notions of the person, providing techniques and a language for everyday individual self- knowledge and self-government, producing psychological subjectivities. This type of individual is specific to contemporary western culture, and so we must locate this study within a cultural and historical framework.

The Historical and Cultural Emergence of the Psychological Subject

The psychological subject is a western phenomenon, located in a particular point in history. Sampson (1989, 1993) has called the western person a 'self-contained individual'. This person is assumed to own her or his individual qualities and characteristics, and is contained within a boundary distinguishing self from not-self. The self-contained individual is by no means a universal ideal: in fact, Geertz (1979) has indicated that the social primacy of the self-contained individual is peculiar to western culture. There is an important sense in which this type of being has been socially constructed. Sampson (1989) has argued that in the western world the individual, distinct from communal definition and the mere players of communal roles, emerged as a political and economic focus in the 15th and 16th centuries. This meant that people were now able to break free of their communally specified roles, becoming instead 'self-determining, autonomous sovereigns, authors in charge of their own life's work. ..the central actors on the social stage' (p. 915).

Within the context of this historical development, the individual came to be known as an embodiment of various dimensions, giving rise to early developments in 'sciences' such as biology, medicine and psychology. For Foucault (1977), these are some of the disciplines that culminated from the surveillance of the individual, and the objectification and delineation of entities such as the body, the psyche, subjectivity and the self. Historically, these sciences have been 'intrinsically tied to programmes which, in order to govern subjects, have found that they need to know them' (Rose, 1990, p. 105). Psychology was thus provided the necessary historical conditions for its emergence and development, building on people's recognition of their own individuality; moreover, it provided a set of techniques and a language through which persons could practise self-government and self-regulation. And so while psychology cannot claim to have invented the individual, it was-along with other technologies, such as the confession (Foucault, 1990)-constructive in its complexification, filling the individual with learnings, needs, drives, desires, an unconscious, and a whole complex of psychic forces. Bulimia nervosa and its personification, the bulimic, is one product of this complexification, being discoursed in terms of such psychological characteristics and complexes (this will be elaborated below).

Being so tightly infused with western modernity's requirement that individuals be self-governing and self-regulating, knowledge of the individual as a psychological being had to be practised by the lay community. Thus the psychological subject became a given, a social fact. People in the western world 'recognise' themselves not simply as individuals, but also as psychological beings. (For a more extensive treatment of this issue, see Rose, 1989; Sampson, 1989.) (2). However, despite this apparent widespread assumption, a psychologist's recognition of psychological processes will differ from that of a non-expert. Historically, western culture has become associated with the development of scientific discourses and practices relating to the person, precipitated by a need for proper governance (Rose, 1990). Experts are therefore distinct from non-experts not simply in terms of what they know, but also in terms of what they can do, as persons deriving legitimacy 'from their claims to tell the truth about human beings' (Rose, 1994, p. 139). This issue requires some elaboration.

Common Sense and the Psy Discourses

According to Parker (1999), 'psychological knowledge is now part of the structure of common sense' (p. 26). But Parker and others (e.g. Moscovici, 1998) make clear that this is not a unidirectional flow of knowledge: common sense and expert knowledges mutually influence and transform each other. Thus psychological knowledge builds to some extent on common-sense knowledge, just as psychological distinctions (such as depression or bulimia) become categories for common-sense employment. Nevertheless, there are inevitably trans- formations as knowledge passes from one domain to the other. Duveen (1998) states: '[a] representation. ..is not only a way of understand- ing something, it is also always a way of not understanding something' (p. 461).3 The question arises: what is it about an expert psychological discourse that common sense does not understand? I will focus on one possible answer to this question: non-experts, by definition, should not be able to grasp the object's apparent psychological complexity.

For Moscovici, '[t]he purpose of all representations is to make some- thing unfamiliar, or unfamiliarity itself, familiar' (cited in Smedslund, 1998, p. 450). While some familiarity is indeed fostered by a common- sense psychological discourse of a problem like bulimia (knowing that it is psychological or has psychological components), its complexity is highlighted, raised for attention, by its very definition as a psycho- logical issue. A proper, comprehensive, or working understanding of bulimia is therefore excluded from the lay public, and the possibility of effective local, non-expert intervention is undermined. When a problem is located in a psy discourse, the action suggested for common-sense knowers is to refer and defer to expert knowers, rather than to mess around with something that cannot be properly understood by non- experts (as we shall see in the case below). Common sense is effectively disqualified as an appropriate means for structuring and effecting intervention for a specialised problem like bulimia. This disqualification might be rationalised as reflecting different levels of specialist knowledge, but it is more than this. Disqualification is a political strategy by which the two groups are distinguished: it reproduces, for psy practitioners, a position of privilege with respect to networks of power that subjugate and minimise common-sense knowers.

The claim is not being made that bulimia is always constructed as a psychological phenomenon by persons of current western culture. Simply to assert that it can be, and is encouraged to be, by the multitude of proponents of a psychological treatment of this problem points to the significance of these issues.

Psychology and the Psy Discourses of Bulimia

The psy discourses, to articulate a tautology, are distinguished by their 'psychological' approach. Gergen (1989) argues that even psychologists who focus on behaviour rather than the 'mind' still understand human behaviour as 'determined by or. . . dependent on psychological processes' or mental states (p. 240). But what do we mean by 'psychological'? While there are a plethora of disparate approaches and discourses available for the construction of the person, there is a sense in which they share a 'common normativity' (Rose, 1998, p. 3). Psy discourses share ideals of the individual as self-knowing, bounded and autonomous, 'the locus of thought, action, and belief' (Rose, 1998, p. 3). Again, I will draw on Sampson's notion of the self-contained individual to identify this normative standard. The construction of psychological 'objects'-concepts and distinctions to think oneself and one's relation to these ideals-is an important part of an achievement of such individualisation. These ideals constitute a normative standard against which persons such as the bulimic can be contrasted.

The very construction of the category of the bulimic depends on such individualisation: it linguistically distinguishes an individual, the 'container' of bulimia. As such, bulimia is more than the display of behaviours such as bingeing and purging, but is also a label that refers to the individual's 'interiority' (to use a term from Rose, 1998). The bulimic is thus seen to house a variety of psychological objects deemed relevant to the performance of these behaviours. This process of interiorisation is important in psychotherapy: the discursive location of psychological objects within the bulimic container facilitates her subjectification to a psychologised discourse of her conduct and of herself. In the process, she is afforded particular subject positions within a psy discourse, which may (amongst other things) undermine her self-government, and point to the necessity of expert interventions.

What psychological objects does the bulimic contain? She has been described as having low self-esteem (Peters & Fallon, 1994), emotional lability (Garfinkel & Garner, 1982), tendencies to impulsivity and depression (Kaplan, Sadock, & Grebb, 1994), a distorted body image (Peters & Fallon, 1994), ambivalence around her sexual identity and difficulties around sexuality generally (Abraham & Llewellyn-Jones, 1997), and often questions about sexual abuse are raised (Hodes, 1995; Wooley, 1994). Furthermore, 'affective illness and personality disorders' are 'psychiatric disorders' that have been associated with bulimia (le Grange & Ziervogel, 1995, p. 478). The bulimic may also be considered to have a narcissistic personality structure (Farrell, 1995). In many cases, she is considered to have poor judgement and insight (Abraham & Llewellyn-]ones, 1997).

This list offers examples of the multitude of psychological objects that allow for the construction of a complex and psychologised bulimic. While therapeutic and theoretical approaches may differ, they share an interest in the relations between such (and many other) discursive categories and their embodiment in the person of the bulimic. Together, psy discourses provide a range of possible accounts precluding notions of the bulimic as 'psychologically healthy'. The aim is to enable some variety of 'psychological health', and so (psy) therapeutic work orients around this normative standard. However, a client may deny the problematic nature and psychological properties of bulimia, and so we may encounter some form of resistance. Here I mean resistance in the Foucauldian sense of practices of freedom: a variety of protests against attempts by the therapist to define her in psy terms. These protests amount to a struggle for the retention of a particular subjectivity, located in a discourse that may be inconsistent with psy's preference for a certain kind of being (the self-contained

individual).

Parker (1998) indicates that psychotherapy relies on a 'kind of self' able to work within and act in accordance with such discourses. Therefore, for psychotherapy to proceed, a psychologically appropriate account must be rendered (see, e.g., Farrell, 1995). To address client resistance, then, a series of rhetorical practices may be employed in psychotherapy. Rhetoric may be defined here, following Billig (1987, 1990), as a set of statements that aim to be persuasive, being argumentative in structure, justifying one position while implicitly criticising others. Understanding rhetoric in this manner is essential to understanding therapeutic operations of power, which risk exposure when a therapist faces a client's refusal to discursively construct her behaviour in a manner consistent with psy. Thus, Stancombe and White (1997) ask: 'How do therapists utilise rhetoric in persuading and "moving" clients to alternative understandings?' (p. 26). For our purposes, this question might be developed thus: What rhetorical strategies might a therapist use to overcome possible resistances, conceal power operations, and persuade a client to construct herself in a

manner consistent with a psy discourse?

It must be noted that the normative standard of self-containment is being increasingly challenged, as discourses of culture, feminism, social systems and spirituality (to name a few) have become impossible to ignore in the construction of human beings. These discourses are seldom permitted to undermine psy effectively, but rather serve to reinforce and expand its parameters, extending its scope of influence. Psy discourses place outside of acceptability accounts that threaten to subjugate psy discourses themselves. Thus, as new ideas and interventions are assimilated, they risk being neutralised in their appropriation by psy (Burman, 1999, p. 162). One effect of this is that psy can now 'speak' to more people, incorporating these discourses into a new and apparently more comprehensive psy discourse. The subjugation of alternative discourses has profound implications for the therapeutic work we are enabled to do: opportunities for subjectivity offered by non-psy discourses, which may threaten the privileging of self-containment, may be lost. (Feminist discourses, and the possibilities for subjectivity they can produce, will be discussed below.)

The psy discourses prescribe a complex of psychological objects and techniques for the production of the bulimic, while simultaneously subjugating other possible accounts. As stated above, psychotherapy involves considerable rhetorical work to achieve these effects. This may involve countering resistance by promoting an account consistent with a psy discourse, while simultaneously minimising non-psy accounts (including that of the client). But more than this, as a complex social interaction, the continuation and tolerability of the therapist-client relationship may also require that the operations of power involved be at least partially hidden from view. These issues will now be considered in greater detail in a case study.

Case Study

My own therapeutic work has been influenced by the social constructionist work of Anderson (1997; see also Anderson & Goolishian, 1988), and the narrative work of White and Epston (1990). Simply stated, my strategies have evolved around the notion of therapeutic questions, as emphasized by these authors, and their use in the generation of experience (Freedman & Combs, 1996). Some of these will become evident in the case. The persons involved in this case were my client, henceforth referred to as Megan, her mother (Mrs H) and myself (therapist). These meetings were audio-recorded, and took place at a community outpatient facility in Pietermaritzburg, South Africa.

Megan was 19 years old when she was referred to me for psychotherapy. She had been living in another town (X) with friends, for approximately a year following the completion of her formal school education. One of her friends had noticed her frequent disappearance into the bathroom after meals, and confronted Megan about this. When Megan admitted to binges, and to vomiting after meals, her friends contacted her mother, and Megan was moved back into the family home over 90 kilometres away. Following the advice of a family friend, Megan's mother made appointments for her to see a doctor, a dietician and a psychologist (myself). In our first meeting, I interviewed Mrs H and Megan together. Mrs H asked if she could speak with me alone first. Megan agreed to this, and so I agreed to interview each of them separately before we decided how to take the process forward (considering the option of family therapy).

In our meeting, Mrs H constructed Megan's conduct in a variety of psychological ways, referring to her 'insecurity', 'denial' of the problematic nature of her conduct, and her behaviour as a sign of 'something wrong. ..with her mind', questioning Megan's 'touch with reality'. These constructions made possible a variety of governmental practices in relation to Megan, one of which was her referral to health and mental health professionals.

Therapist: How did it happen that you ended up coming to see me and Dr Y?

Mrs H: We were worried, and confused I guess. We didn't know what else to do. I'm worried that we may make things worse if we do the wrong thing, you know?

Therapist: Like what?

Mrs H: ...We've stopped her from going out with her friends, just to. .. I don't know. ..get some control of the whole situation.

Therapist: And that may be wrong, is that what you're worried about?

Mrs H: Maybe, I don't know. Maybe you can guide me about it?

In this manner, Mrs H constructed Megan's behaviour as a psychological phenomenon about which she had little expert knowledge and know-how. Megan was referred to expert services because psy dis- courses place problems like bulimia more or less out of reach of the layperson. Mrs H was 'worried, and confused', and expressed concern that 'we may make things worse if we do the wrong thing'. 1, on the other hand, was considered an expert in such matters, and a source of guidance. While local (family) knowledge of bulimia as a psychological phenomenon was evident to some extent, the complex, unexpected and confusing nature of its presentation required referral to an expert. This notion of therapist as expert, and the privilege it implies, cannot be separated from the dialogical interactions to be discussed below, as it is from this position that my practices derive legitimacy and influence.

Nevertheless, Megan had a different view than her mother. When interviewed later on that same day, the following transaction took place:

Therapist: So how do you feel about being here?

Megan: [laughs] ...I don't mind. They're all panicking over nothing.

Therapist: What?, what's the nothing that's worrying them?

Megan: My throwing up and stuff ...it's not a big problem

Therapist: So what is their fear, what's your mom's fear?

Megan: That I've got a problem. ..in my head or something.

Therapist: And you're saying 'not true'?

Megan: Exactly, not true. I just need to lose some weight a bit. ..I know what I'm doing.

Megan refused, initially, to problematize her behaviour in a psy discourse. This very refusal enabled her mother to evoke a psy discursive construction of a denial of a real psychological problem. For Megan, however, this was a controlled issue of attempts at weight loss, partly as a strategy to attract young men (as emerged later), and not a matter of psychological denial, or a problem with her emotions or mind. Consequently, her parents' efforts to control her did not persuade her that she had a psychological 'problem', but rather were seen as over- reactions. She stated that she was in control of the situation, while her mother felt she was 'in denial'. This contest did not effectively challenge or undermine Megan's self-construction, or her mode of knowing and governing herself and her behaviour.

Megan spoke of her eating behaviour as a matter of attractiveness and slimness. I asked: 'what convinced you that this is the best way to get what you want?' Megan responded by saying: '1 don't know, I just want to lose weight, cos I'm fat, and this is easy, you know? So what?' She refused to construct her behaviour within the psy discourse I was trying to build with her. However, in order to work with such issues in psychotherapy, they must be rendered sufficiently psycho- logical, or psychologically complex. The problem here was: how can her account be rendered psychologically good-enough to allow for our work to proceed? One rhetorical strategy to address this is to focus the dialogue on areas where minimal resistance to psychologisation will be encountered, where the stakes (in terms of potential threats to Megan's existing self-self relation) do not appear to be too high (see, e.g., Garfinkel et al., 1985). Thus, resistance to a psychologised discourse is not directly confronted, but may nevertheless be destabilised. The open contesting of discourses in sessions and the interpersonal, political nature of Megan's refusal to have her conduct psychologised thereby become concealed.

Thus, we spoke about Megan's concern that people were generally dishonest. This mattered to her when they said she was thin and attractive. She felt that this was less a matter of her own psychology (e.g. her insecurity) than a matter of the 'real-world' dishonesty (a phrase Megan suggested in sessions) of people around her. In our 3rd session, I gently countered this attribution, posing the question: 'Is it hard to trust people when they say nice things?' Megan's reply: 'Yes', opened the possibility of reconstructing the perceived lies of others into a matter of her own psychology. It was now not simply that people told lies, but that she could not trust when they said 'nice things'. The discourse could then shift from a 'real-world' (externally based) dis- course to a more internal, psychological discourse. This may be seen as a step towards overcoming Megan's resistance against attempts to define her eating conduct in psy terms.

The psy discourse was developed further. I suggested to Megan that trust was actually a matter of 'risk', and this was built upon by connecting the idea of risk with the idea of bravery and courage.

Megan: hmm. Are you just messing with me?

Therapist: I honestly feel that to accept a compliment can be a very courageous thing to do. It's risky. ..you see, I think you have a knack of being able to look at things quite deeply, and so you pick up on all sorts of little things that a lot of people just wouldn't be able to even notice. ..does that sound familiar?

Megan: [pause] ...Yes, I think about things like that quite a lot

Therapist: And a lot of people just don't think about things, about things that are more subtle and stuff ...and I think you do, and so you could probably tell things like. ..I don't know. . .

Megan: Things like. ..the other day I saw my boss, and he had a funny look on his face, and no-one else noticed. ..and then I said to Jody, what's wrong with Jay? And she thought there was nothing wrong, and she just said I was paranoid [laughs] ...but then I found out that his girlfriend broke up with him. ..

Therapist: And you were sensitive and subtle enough to see that something was up.

My suggestion that Megan could 'look at things quite deeply' was offered to her with the question: 'does that sound familiar?' This is an example of a 'story development question' (Freedman & Combs, 1996), encouraging Megan to develop a narrative of herself as perceptive, 'sensitive and subtle'. Megan provided an example (noticing her boss's 'funny look') through which she could rhetorically develop this self- discourse. This 'therapeutically' co-constructed knowledge of Megan's sensitivity was then linked back to the discourse of trust as a courageous decision a few minutes later:

Therapist: So, I'm wondering, urn. ..

Megan: About it being brave to take compliments and trust people.

Therapist: Yes, thank you, yes. It's brave when you're able to pick up on subtle

cues, cos you're more likely to see deeply. ..and so if I say 'Megan, you're great', you could see, even unconsciously maybe notice that I moved my head a certain way. ..and then because you're scared to accept the compliment. ..

Megan: I'd see that your head movement told me that you were lying. . . Therapist: two plus two equals seventeen. ..

Megan: [laughs] ...or three hundred.

In this manner, Megan's initial discourse of a dishonest 'real world' was systematically worked on, in rhetorical fashion, to the point where her ability to trust was highlighted as the 'real' problem. Furthermore, this trust was then constructed as desirable and brave, placing the problem and its solution clearly within the domain of Megan's self- contained psychology, instead of in the non-self realm of an external, dishonest world. This transformation was constructed dialogically; Megan gave examples and analogies to facilitate this alternative dis- course of her experience. Compounding examples and analogies with specific reference to one discourse constitutes a powerful rhetorical justification for the relevance of that discourse. These manoeuvres can be seen as part of Megan's 'technologies of the self', working on herself in a particular way, thinking herself as self-contained, 'owning' her fear within the boundary of her self, rather than employing external 'real- world' attributions.

In the following session, we returned to the issue of compliments. I asked Megan why I could see the compliments made to her, where she failed to see the compliments inherent in certain actions by others. The example of her being asked out on a date was used:

Therapist: What is it that stops you seeing that [compliment] I wonder? Megan: I suppose its just fear. ...It's just hard to trust people.

Therapist: Fear stops you?

Megan: Yes. ..just. ..I get nervous.

Therapist: So what does your fear say to you when someone like Lance comes along, and says in his indirect way [by asking her out on a date] ...'Megan, you're attractive and I like you'?

Megan: Urn, well, my confidence goes low when. ..when there's that fear.

Therapist: So fear says something like: 'don't believe him'? Would it be something like that?

Megan: Yes, 'he's lying, don't believe him'. So it's hard to be brave with that.

Our conversations allowed for the distinction of fear as a force or voice that interfered with her judgement and perception of social situations. Examples of Megan's 'bravery' further complemented this account (thereby developing her self-discourse): she spoke of the courage involved when she stood up to her father for the first time, prior to our 6th session. Megan saw this as significant, given her father's expectations of her 'as a girl'; that she be quiet, unassuming and non-aggressive. Megan described her mother as similarly passive in relation to her husband. She described a sense of pride in 'breaking the mould , with regard to her position in a heretofore unarticulated struggle around gender expectations and power operations. In this way, Megan began to relate certain dialogically identified psy discursive objects (such as courage and fear) to her relationship with her father and his gender-prejudiced expectations and practices. In this way, psycho- logical distinctions can make visible aspects of gender politics (see Swann, 1999; to be discussed below).

Megan's talk of herself and her interactions increasingly involved considerations of her 'fear', low 'confidence', 'trust' and nervousness. By invoking, analysing, developing and complexifying these psy objects and their interrelations, our dialogue permitted the construction of a psy discursive network within which Megan could account for her eating conduct. And so, in our 7th session, she began to analyse her body image and her eating behaviour in a psy discourse, invoking concepts of fear and courage:

Therapist: What attitude would your fear have towards accepting when people say you're attractive. ..like when Lance asked you out?

Megan: [laughs] I see where you're going you know!

Therapist: [laughs] Oh? Where does it look like I'm going?

Megan: That they might not be lying, or that he, maybe does kind of like me that way.

Therapist: Sure, they're not. He probably does. . . How much courage do you need to accept that compliment?

Megan: A lot really. I do sometimes though. . . take a chance, like be a bit brave.

Therapist: Uh . . . so does your courage. . . sometimes step in and say, 'hey, it's okay, believe it, just this once'?

Megan: Yes, it does sometimes I guess, but not always.

Therapist: That must feel great for those short few seconds, when you believe it?

Megan: Ya, before this fear takes over and says like 'OK, that's enough fun for you for one day'.

Thus Megan was able to employ the discursive category of 'bravery' to accept that some people, sometimes, found her attractive. She decided that her 'need to be liked', which she came to believe was fundamental in understanding her eating conduct, was also affected by issues of fear and courage. The following excerpt is drawn from later on in that session:

Megan: I think, urn, I worry that people won't like me, and that I have to

be. ..it's especially when we go out to clubs. The girls there are so like beautiful, some of them.

Therapist: Urn. ..if your fear wasn't so dominating, like that, would it be different?

Megan: Well ...ya, maybe. It's like my mom says, people have to like you for who you are, as a person.

Therapist: So what does courage say, I mean when your fear bluffs you into thinking you're unattractive?

Megan: Maybe. ..it's there sometimes, I mean I know that's not always true, sometimes I feel sort of, okay. ..quite good, you know.

Therapist: ...Your fear doesn't always dictate things, do you think? So that courage is coming in there. ..a bit?

Megan: [pause] ...Sometimes my fear isn't that strong. ..and when it's strong, I want to lose weight and. ..you know, get rid of my food or whatever. But sometimes that's just a little bit, just to balance it out a little bit, I can ...this courage comes along, sort of thing. .. so then I'm okay, and I wouldn't worry too much about that stuff.

Therapist: So it brings a bit of an attitude of, like '1 don't care about that' ... a little bit of that. So then you worry less, and you don't have to throw up and stuff ...the fear is like less powerful and then you just get on with things?

Megan: Just get on with things a bit more, ya.

By our 7th session, as indicated by the above dialogue, Megan began to analyse and construct, in increasingly complex ways, the relation- ship between her fear, her courage, her need to be liked, her body image and her eating behaviour. Therapeutic questioning had effectively enabled the construction of a network of dynamically interacting psychological objects, contained within and complexifying Megan's 'self'. Her eating behaviour could then be spoken about as part of this network.

The articulation and delineation of psychological objects and psychological struggles, in areas other than Megan's eating conduct, is ultimately a political and destabilising manoeuvre, eroding Megan's protests against being defined in a certain way. Destabilization is permitted and tolerable because it is concealed, disguised as a therapeutic strategy for the development of self-awareness and problem resolution. Apparently benign, helpful and gentle, talk in therapy is allowed to continue, facilitating the construction of a discursive system that eventually, but subtly, undermines Megan's refusal to render a psy account of her eating behaviour. Hence, the concealment of power operations allowed for the dissipation of the initial potential political struggle between us. This struggle involved a contest for the definition of Megan's subjectivity. Megan was conscripted into psy, becoming empowered to construct herself and her eating behaviour in psychological ways, problematizing her behaviour as psychological. In this way, she became recognisable not only to her family and to me, but to herself, as a bulimic: a person with a decidedly psychological problem with her body image and her bingeing and purging eating behaviour.

With this case, I also emphasize that my 'therapeutic' rhetorical work cannot be separated from its location: I enjoy a culturally and historically privileged position as an expert/professional/scientist. As previously indicated, psychologists and other psy practitioners derive legitimacy from their claims to tell the truth about human beings (Rose, 1994). But not only do I tell the truth, I am seen as someone trained to be helpful, 'therapeutic', someone in whom to confide and place one's trust. To some extent, my position of influence relies on my work and my profession being constructed this way, rather than being exposed as based in power operations that predetermine who should be persuaded by whom, or being revealed as manipulative, undermining, subjugating or even disqualifying non-psy accounts. With the former highlighted, and the latter concealed, my talk is invested with meaning and significance in a way that Megan's mother's-and Megan's own talk-is not. It is not simply that my rhetorical strategies may have been more persuasive than those of Megan's family, but that my words carry the weight of expertise, knowledge, and thus power.

In our 12th session, Megan announced her decision to return to city X. It seemed that her parents were less concerned about her, and had stopped their monitoring activities. Megan and her mother began to meet for lunches during the week. Megan described these meetings as pleasant and new for their relationship. She spoke of feeling 'strong and positive', and had some clear ideas about her future: she decided to embark on a teaching course, and felt she wanted to work with children. Her vomiting behaviour had significantly decreased, although she admitted to throwing up 'once in a while' (although she said she'd begun to 'forget about it').

The Possibility of Alternative Discourses

Psy's encompassment of the human being has been enhanced by the incorporation of other knowledges (such as feminism) into its dis- courses and practices. As Duveen (1998) and Burman (1999) have indicated, such incorporation is frequently accompanied by important transformations in the imported discourse. Such a transformation occurred in the case above, via the incorporation of Megan's self- knowledge by my own psy discourses (justified as rendering her eating disorder solvable), but can also occur more generally, through the incorporation of, for example, feminist discourses by psy. Bulimia has received much attention from the researches and practices of feminist authors (e.g. Dolan & Gitzinger, 1994; Fallon, Katzman, & Wooley, 1994; Swann, 1999). Taking Duveen and Burman's point, then, we must ask: What may have been obscured by an introduction of feminist ideas into psy? What may have been lost or compromised by the construction of a feminist-psychotherapeutic discourse? While this matter deserves a much more detailed analysis, I offer the following preliminary thoughts.

Feminist authors and practitioners have highlighted the relation between bulimia and gender-based oppressive practices. However, some feminist authors (e.g. Kitzinger, 1993) criticise the employment of psychotherapy to address female oppression. Rather than being highlighted, social and political action risk being understood as optional secondary aspects of individual therapeutic change. Inevitably, psychology and psychotherapy highlight psychological processes as a discursive priority, rather than socio-political processes. However, in a defence of therapy (specifically, narrative therapy) as a viable strategy for feminism, Swann (1999) disagrees with Kitzinger's view that therapy is intrinsically bad for women. Instead, she suggests that it can effectively address the psychological products of their oppression, while simultaneously pointing towards local political action. According to Swann, 'self-blame and unworthiness are examples of the power of oppression operating at a very local level, a level overlooked by Kitzinger's analysis' (p. 111). She (and other feminist therapists) views therapy as one possible means for 'politicising the personal'. Indeed, this possibility was raised with Megan when she challenged her father, and spoke in sessions of 'breaking the mould' by challenging her family's expectations of what 'girls' should be like. However, while therapy can make visible such political possibilities, it raises another question: Is the psychologisation of disempowerment, involving the construction of discursive objects such as 'self-blame', 'unworthiness' or 'bulimia' (or, indeed, 'fear', 'courage' or 'pride', as constructed in my dialogues with Megan), necessary to address the problem of female oppression effectively? Do political change and political agency need to be mediated by the degree of psychological complexity constructed in therapy? I argue that they do not.(4) We cannot assume that psy interventions are the only-or best- way of addressing problems of oppression and subjugation.

Returning then to the case of Megan, a non-psychological feminist discourse may make possible a rather different dialogue, leading to different possibilities for self-government and self-regulation. Imagine the following questions are asked within the context of a hypothetical 'political empowerment' discussion group. They are offered here as examples of a point of entry into an alternative non-psy discourse:

Megan, who might be happy/unhappy should you succeed in becoming really thin? With whom do you compete when you embark on these techniques? For what/whom are you competing? What is the prize? Who is the judge? Who sets the terms for this competition? Which parts of the community might be pleased by these attempts of yours? Who stands to gain, and who stands to lose? Is it important that women prevent themselves from becoming subject to male values and male-prescribed ideals? What are the benefits/limits of such subjection? How could you go about preventing your own domination by others? Who would you like to join you? Who would be afraid to join you? What would you do about that? What impact would this have on your interactions with men/other women? What values might be more appropriate and useful for you? Who would agree/ disagree with you? What problems might you have with the kind of questions I am asking?

These questions, while rhetorical and potentially subjugating, allude to the possibility of constructing Megan's position within the discourse of gender politics, without evoking investigations into psychological matters. This is not a call for the eradication of psy discourses, and I do not suggest that a feminist discourse is immune from problematic power relations and tactics. As mentioned earlier, power infuses social interactions. However, an approach that highlights rather than conceals the workings of power is being promoted. Further, I question the degree to which psychological complexity, constructed via intense psychological/therapeutic self-examination (as evident in the case study), is required to render a meaningful account of eating practices labelled as bulimia. At most, the above questions call for a 'naive' psychology, and propose talk around interpersonal struggle and political positionings, suggesting a politically located subject whose form of self-government takes on a different aspect. Thus, one's ethics, practices and subjectivities might be informed by different priorities: by the choices made around political situatedness and political agency, rather than by a self-contained psychological complexity. This occurs as a consequence of the elaboration of a more politically orientated discourse. As already noted, narrative (e.g. Swann, 1999) and feminist therapies (e.g. Fallon et al., 1994) do address these issues, and the questions offered above are certainly not precluded from their practice. Nevertheless, the issues raised around these questions need not be psycho- logically complex, and, unlike narrative and feminist therapies, do not necessarily evoke associations with psy discourses, practices or institutions (hence my reference to an imagined 'political empowerment discussion group', as opposed to therapy). And yet they are able to make visible an account of 'bulimia' that points us towards 'knowing' how to go on.

Thus we are able to locate 'problems' like 'bulimia' in a very different discursive domain. Gergen (1999) has indicated that there are currently almost 400 terms for psychological and psychiatric problems, of which bulimia is just one. It is not suggested that all of these might be usefully re-discoursed in feminist terms. Rather, I am simply suggesting that all of these problems of living can be discoursed in many ways, and that their containment within psy is not always necessary. In particular, psy's hegemony is undesirable. In the case of bulimia, I have argued that psy conceals alternatives, and risks de-emphasizing opportunities for political and social change with regard to gender relations. In short: psy contains and de-radicalises alternative discourses (Burman, 1999), via a subjugating circumscription and constitution of the western human being as psychological subject, and its totalizing-though often disguised-form of government. It therefore becomes important to consider the potentials of non-psychologised discourses.

Discussion

The bulimic subject, defined as a person who becomes subject and subjects herself to the psy discourses associated with the category of bulimia, cannot exist without psy. I have tried to illustrate how, during the course of therapy, Megan began to constitute herself within the dis- courses of a psychologised bulimia. Her behaviours were discursively linked to psychological objects such as 'fear', 'courage', 'the need to be liked' and 'body image'. These distinctions construct a psychologically dynamic complex designed to move Megan towards the possibility of being 'psychologically healthy' (precluded from a psy discursive account of bulimia), insightful (acknowledging and working on the problematic nature of her behaviour), accurate in her judgement regarding appraisals of social situations (such as when people like her), and accurate in her appraisal of her body image. The inevitable assumption is that her 'problems' initially reflected otherwise, both to her mother and myself, her therapist. It is important to note that Megan displayed some initial resistance to this discursive positioning of herself. However, the introduction and elaboration of a psy discourse prevented this resistance from developing into an alternative account of her conduct. Thus, the possibility of an alternative account was sup- pressed, and the means of suppression concealed.

This suppression and concealment are important for psychotherapy, since its success relies on the ability of clients to construct themselves in psy terms and act accordingly (Parker, 1998). When resistance to psychologisation is encountered, rhetorical strategies and operations of power are employed to resituate these protests discursively in a psy discourse. While facilitating therapy, these manoeuvres de-radicalise resistance, reducing its potentials to expose and effectively to challenge a psy discourse and the mechanisms of power that make it practicable. Specifically, the rhetorical, argumentative, even manipulative, nature of our work is hidden from view by the psychologisation of resistance. Instead, a view of therapy as the benign, good-natured practices of a helping profession is made possible.

The revelation of resistance as a political rather than psychological manoeuvre marks one possible beginning of the 'insurrection of subjugated knowledges' (Foucault, 1980, p. 81). Subjugated knowledges include both 'local popular knowledges' (such as Megan's account of herself) and more erudite and specialised knowledges (such as feminism). Thus, a critical look at operations of power in therapy might begin by exposing power, thereby permitting the invocation of previously subjugated and disqualified 'naive. . . local popular knowledges' (Foucault, 1980, p. 82). These knowledges can form the beginnings of protest against an unwanted discourse. In this vein, Falzon (1998) has interpreted Foucault's work as embodying a 'dialogical ethics, i.e. the ethical as an attitude of openness to or respect for the other, for that which resists' (p. 64, emphasis added). From this perspective, the discursive transformation of protest into psychological resistance does not conform to a 'dialogical ethics'.

How might such insurrection, and a respect for 'that which resists', be achieved in therapy? Certain therapies (such as narrative therapy) have attempted to address the problem of subjugation and hidden power mechanisms by exploring strategies of power, discursive positioning and the problems of gender oppression in the 'therapeutic relationship' (e.g. Morss & Nichterlein, 1999; Swann, I999). In my own practice, the tension that has arisen between my dual roles of therapist and critic/analyst has led to an emerging respect for resistance as protest, and the facilitation of client accounts that might otherwise be subverted by a psy discourse. Specifically, it has been helpful to prioritise for discursive attention hints of new stories expressed by clients, which might sometimes be inconsistent with a psy discourse. In their articulation and development, these alternative accounts can promote a different set of subject positions and power relationships within the therapy setting itself.

This is not simply replacing one story for another. The insurrection of a client's account is not intended as a means for structuring intervention in therapy (which might suggest at its extreme, for example, that a suicidal bulimic should be left to die), but involves employing resistance as a point of reference against which subjugating practices might be analysed and highlighted. And so, following Byrne and

McCarthy (1999), psy discourses need not be rejected in therapy, but can be relativized in their discursive juxtaposition with other discourses-in this case, an emergent discourse of resistance. This means that a psy discourse might be used but also questioned, in the therapist-client dialogue, by invoking and elaborating local, specific discourses of resistance to psy as they emerge in sessions, and vice versa. In this way, resistance can be retained as a political rather than psychological performance. Power and subjugation are not removed in the process. However, it is suggested that an acknowledgement of the political, rather than psychological, nature of resistance can enable a relationship in which psy and its territorialising potentials are problematised in the therapist-client dialogue itself.

However, despite these attempts to address power in the therapist-client relationship more openly, tensions remain. First, as a psy practitioner myself, I question the extent to which a psy discourse is required at all to address such issues effectively in the case of bulimia. For example, in considering the practice of juxtaposing discourses referred to above, and the narrative therapy work exemplified by Swann (1999), I question the apparent necessity of psy's retention - albeit in problematised form - as the primary discourse around which a discourse of resistance should be articulated. Certainly, therapists' training and proficiency in using a psy discourse, coupled with the cultural availability of a psy discourse on bulimia, facilitate the retention of psy as a discursively constructive force in professional 'helping' relationships. And yet, while psychotherapy practices require psychologised clients, the construction of a discourse for problem-resolution does not. Psy is just one of many discourses that could potentially make visible new practices, power relations and subjectivities in relation to people positioned as bulimic. In this regard, the example of a non-psychologised feminist discourse has been suggested.

There is a second problem, however. Using alternatives, such as a feminist discourse, in a therapy setting does little to challenge the institution of psy. While the use of feminist accounts in therapy might highlight power and subvert psy discourses at a local level (in the therapist-client relationship), such a practice reproduces the cultural prescription that problems like bulimia fall in the domain of psy, requiring the expertise of its practitioners, regardless of the discourses used within that setting. And so problems of living are retained as the territory of psy practitioners, even if the latter de-privilege and problematize psy talk in dialogue with clients. Thus, the local, in-session questioning and analysis of knowledge and power become, in them- selves, a questionable political strategy. Our work is still called 'therapy', and we continue to practise within a broader network of power relations, positioned very differently than our clients can be. But more than this, with regard to 'knowing' about human beings, western culture has situated psy practitioners in a privileged position relative to 'practitioners' (speakers, authors) of other erudite discourses. Whatever we do within therapy, our work threatens to leave the subjugation of other discourses relatively unchallenged, thereby reinforcing psy hegemony.

And so, in addition to reflecting critically on our work by analysing the subjugation of local (common-sense) knowledges of our clients in our day-to-day work, we should also confront the subversion of alternative (erudite, disciplined) discourses. For example, psy dis- courses have been able to resituate feminist concerns within psy. This appropriation threatens to undermine the political potentials of feminism through the psychologisation of gender relations and personhood. However, a non-psychologised feminism might point to different power networks and subjectivities than might a feminist psychotherapy. Specifically, it has been suggested that a feminist account could encourage political agency as a discursive priority, without psy's apparent requirement that such agency be mediated by its psychologisation. A feminist discourse does not require such discursive relocation and transformation in order to make visible an account of 'how to go on' in relation to what has been called bulimia.

In an attempt to resist the subversion of other discourses, we might engage in dialogue with practitioners of other knowledges, and together consider how networks of power and subjectivities might be formed in a manner different from psy. Through such dialogue, we (psychologists, psychiatrists or psychotherapists) should be challenged, and continually ask ourselves, when we borrow concepts from feminism, anthropology, history, sociology, political science or any other form of erudite 'knowing' , the question: What is it that we are failing to understand, or obscure from view, when we appropriate other discourses? What transformations are effected by this appropriation? Further, dialogue focusing on which discourses might potentially be employed in which settings, and in relation to which problems, might promote the visibility of discourses that prevent discursive closure and the de-radicalisation of resistance. Such negotiation might allow for, or even require, the construction of new discourses, power relations and subject positions. By investigating these issues-both in our sessions, and in our reflections on our position within broader networks of discourse - we might begin to acquire some insight into what opportunities for political agency we miss, as well as locating or constructing discourses through which such agency can be enhanced. Importantly, this could conceivably lead to a de- territorialisation of certain problems of living, which need not necessarily be accounted for in psy terms, rendering referral to psy- practitioners just one possible option amongst many. In this way, we might usefully undermine psy hegemony as we explore discursive means by which psy discourses can be prevented from effecting an ever greater encompassment and subjugation of the human being.

Notes

I would like to thank the following people for their assistance in drafting this paper: Kim Barry, Kevin Durrheim, Trudy Meehan, Floss Mitchell, Blake Stobie and Doug Wassenaar. I would also like to thank Anthony Pillay for providing an environment in which research and critical reflection are encouraged.

1, The case study presented is of a woman diagnosed as bulimic; further, the vast majority of diagnosed bulimics are female. All references to clients in this study will therefore be gendered accordingly.

2, It must be noted that alternative, non-western theories of self as socially situated have been developed by authors such as Vygotsky (1981) and Bakhtin (1984). Also, in western academia, constructionist authors such as Gergen (1999) propose a formulation of self as relational and socially constructed, and Harre (1993) argues for a discursive construction of self.

3, A note on the use of the term 'representation': Potter and Edwards (1999) have contrasted the respective roles of representations in social representations theory and discursive psychology. I use the term 'discourse' (i) in order to emphasize the constructive, rather than 'sense-making', nature of representations, and (ii) to encourage the visibility of activity or practices.

4, Incidentally, neither do personal experiences of joy, inspiration or meaningfulness, according to Rijkonen and Vataja (1999).

References

Abraham, S., & Llewellyn-Jones, D. (1997). Eating disorders: The facts. New York: New York University Press.

Anderson, H. (1997). Conversation, language, and possibilities: A postmodern approach to therapy. New York: Basic Books.

Anderson, H., & Goolishian, H. (1988). Human systems as linguistic systems: Evolving ideas about the implications for theory and practice. Family Process, 27, 371-393.

Bakhtin, M. (1984). The problems of Dostoevsky's poetics (C. Emerson, Ed.). Minneapolis: University of Minnesota Press.

Billig, M. (1987). Arguing and thinking: A rhetorical approach to social psychology. Cambridge: Cambridge University Press.

Billig, M. (1990). Rhetoric of social psychology. In I. Parker & J. Shotter (Eds.), Deconstructing social psychology (pp. 47-60). New York: Routledge.

Burman, E. (1999). Whose construction? Points from a feminist perspective. In D.J. Nightingale & J. Cromby (Eds.), Social constructionist psychology: A critical analysis of theory and practice (pp. 159-175). Buckingham: Open University Press.

Byrne, N. O’S., & McCarthy, I.C. (1999). Feminism, politics and power in therapeutic discourse: Fragments from the fifth province. In I. Parker (Ed.), Deconstructing psychotherapy (pp. 86-102). London: Sage.

Dolan, B., & Gitzinger, I. (1994). Why women? Gender issues and eating disorders. London: Athlone.

Duveen, G. (1998). The psychosocial production of ideas: Social representations and Psychologic. Culture & Psychology, 4(4), 455-472.

Fallon, P., Katzman, M.A., & Wooley, S.C. (Eds.). (1994). Feminist perspectives on eating disorders. New York: Guilford.

Falzon, C. (1998). Foucault and social dialogue. New York: Routledge.

Farrell, E.M. (1995). Lost for words: The psychoanalysis of anorexia and bulimia. London: Process.

Fish, V. (1999). Clementis's hat: Foucault and the politics of psychotherapy. In I. Parker (Ed.), Deconstructing psychotherapy (pp. 54-70). London: Sage.

Florence, M. (1998). Foucault. In J.D. Faubion (Ed.), Michel Foucault: Aesthetics, method, and epistemology (pp. 459-463). London: Penguin.

Foucault, M. (1972). The archaeology of knowledge. London: Tavistock.

Foucault, M. (1977). Discipline and punish. New York: Pantheon.

Foucault, M. (1980). Power/knowledge: Selected interviews and other writings 1972-1977 (C. Gordon, Ed.). New York: Harvester Wheatsheaf.

Foucault, M. (1982). The subject and power. In H.L. Dreyfus & P. Rabinow (Eds.), Michel Foucault: Beyond structuralism and hermeneutics (pp. 208-226). Brighton: Harvester.

Foucault, M. (1990). The history of sexuality (Vol. 1). London: Penguin.

Foucault, M. (1997a). The ethics of the concern of the self as a practice of freedom. In P. Rabinow (Ed.), Michel Foucault: Ethics, subjectivity and truth (pp. 281-301). New York: New Press.

Foucault, M. (1997b). Sex, power and the politics of identity. In p. Rabinow (Ed.), Michel Foucault: Ethics, subjectivity and truth (pp. 163-173). New York: New Press.

Freedman, J., & Combs, G. (1996). Narrative therapy: The social construction of preferred realities. New York; Norton.

Garfinkel, P.E., & Garner, D.M. (1982). Anorexia nervosa: A multidimensional perspective. New York: Brunner/Mazel.

Garfinkel, P.E., Garner, D.M., & Kennedy, S. (1985). Special problems of inpatient management. In D.M. Garner & P.E. Garfinkel (Eds.), Handbook of psychotherapy for anorexia nervosa and bulimia (pp. 344-359). New York: Guilford.

Geertz, C. (1979). From the native's point of view: On the nature of anthropological understanding. In P. Rabinow & W.M. Sullivan (Eds.), Interpretive social science (pp. 225-241). Berkeley: University of California Press.

Gergen, K. (1989). The possibility of psychological knowledge: A hermeneutic inquiry. In R.B. Addison & J.J. Parker (Eds.), Entering the circle: Hermeneutic inquiry in psychology (pp. 239-258). Albany: State University of New York Press.

Gergen, K. (1999). An invitation to social construction. London: Sage.

Harre, R. (1993). Social being. Oxford: Blackwell.

Hodes, M. (1995). Anorexia nervosa and bulimia nervosa in adolescents. Continuing Medical Education, 13(5), 481-488.

Ingleby, D. (1985). Professionals as socialisers: The 'psy complex'. Research in Law, Deviance and Social Control, 7, 79-109.

Kaplan, H.I., Sadock, B.J., & Grebb, J.A. (1994). Synopsis of psychiatry: Behavioural sciences, clinical psychiatry. Baltimore, MD: Williams & Wilkins.

Kitzinger, C. (1993). Depoliticising the personal: A feminist slogan in feminist therapy. Woman's Studies International Forum, 16(5), 487-496.

le Grange, D., & Ziervogel, C. (1995). Guest editorial; Eating disorders, including obesity. Continuing Medical Education, 13(5), 477-478.

Morss, J., & Nichterlein, M. (1999). The therapist as client as expert: Externalising narrative therapy. In I. Parker (Ed.), Deconstructing psychotherapy (pp. 164-174). London: Sage.

Moscovici, S. (1998). Social consciousness and its history. Culture & Psychology, 4(3), 411-429.

Parker, I. (1998). Constructing and deconstructing psychotherapeutic discourse. The European Journal of Psychotherapy, Counselling and Health, 1(1), 65-78.

Parker, I. (1999). Critical reflexive humanism and critical constructionist psychology. In D.I. Nightingale & I. Cromby (Eds.), Social constructionist psychology: A critical analysis of theory and practice (pp. 23-36). Buckingham: Open University Press.

Peters, L., & Fallon, P. (1994). The journey of recovery: Dimensions of change. In P. Fallon, M.A. Katzman, & S.C. Wooley (Eds.), Feminist perspectives on eating disorders (pp. 339-354). New York: Guilford.

Potter, I., & Edwards, D. (1999). Social representations and discursive psychology: From cognition to action. Culture & Psychology, 5(4), 447-458.

Rijkonen, E., & Vataja, s. (1999). Can (and should) we know how, where and when psychotherapy takes place? In I. Parker (Ed.), Deconstructing psychotherapy (pp. 175-187). London: Sage.

Rose, N. (1989). Governing the soul: Technologies of human subjectivity. London: Routledge.

Rose, N. (1990). Psychology as a 'social science'. In I. Parker & I. Shotter (Eds.), Deconstructing social psychology (pp. 103-116). London: Routledge.

Rose, N. (1994). Identity, genealogy, history. In S. Hall & P. Du Gay (Eds.), Questions of cultural identity (pp. 128-150). London: Sage.

Rose, N. (1998). Inventing our selves: Psychology, power, and personhood. Cambridge: Cambridge University Press.

Sampson, E. (1989). The challenge of social change for psychology: Globalisation and psychology's theory of the person. American Psychologist, 44(6), 914-921.

Sampson, E. (1993). Celebrating the other: A dialogic account of human nature. San Francisco, CA: Westview.

Smedslund, I. (1998). Social representations and Psychologic. Culture & Psychology, 4(4), 435-454.

Stancombe, I., & White, S. (1997). Notes on the tenacity of therapeutic presuppositions in process research: Examining the artfulness of blamings in family therapy. Journal of Family Therapy, 19, 21-41.

Swann, V. (1999). Narrative, Foucault and feminism: Implications for therapeutic practice. In I. Parker (Ed.), Deconstructing psychotherapy (pp. 103-114). London: Sage.

Vygotsky, L. (1981). The genesis of higher mental functions. In J.V. Wertsch (Ed.), The concept of activity in Soviet psychology (pp. 144-188). Armonk, NY: M.E. Sharpe.

White, M., & Epston, D. (1990). Narrative means to therapeutic ends. New York: Norton.

Willig, C. (1999). Introduction: Making a difference. In C. Willig (Ed.), Applied discourse analysis: Social and psychological interventions (pp. 1-21). Buckingham: Open University Press.

Wooley, S.C. (1994). Sexual abuse and eating disorders: The concealed debate. In P. Fallon, M.A. Katzman, & S.C. Wooley (Eds.), Feminist perspectives on eating disorders (pp. 171-211). New York: Guilford.

Biography

MICHAEL GUILFOYLE was born in South Africa, where he studied psychology at the University of Natal, Pietermaritzburg, before training at the Midlands Hospital complex. He is currently Senior Clinical Psychologist at St Senans Hospital in Enniscorthy, Ireland. His interests include social constructionist and discursive applications of therapy, and the uses of rhetoric in psychotherapy. He also has an interest in cross-cultural conceptions of self, and their relationship to therapeutic practices.

ADDRESS: Michael Guilfoyle, Department of Psychology, St Senans Hospital, Wexford Road, Enniscorthy, Co Wexford, Ireland. [email: mcgg@eircom.net]