Deconstructing Paranoia: An Analysis of the Discourses Associated with the Concept of Paranoid Delusion

David J. Harper PhD Thesis June 1999

Back to title page

Chapter 6

Analytic theme III:

The discursive construction of neuroleptic medication and other forms of professional intervention(1)

Gimme, gimme, gimme medication, gimme medication to kill this hole.

Gimme, gimme, gimme medication, gimme medication to kill this, kill this hole.

(Written by Bobby Gillespie, Andrew Innes, Robert Young and Martin Duffy of Primal Scream) © 1997 Creation Records Ltd.

Now the drugs don't work, they just make you worse but I know I'll see your face again.

(Written by Richard Ashcroft of The Verve) © 1997 VC Records Ltd.T/A. Hut Recordings

6.1 INTRODUCTION

This third chapter based on analyses of the interviews in the context of wider cultural and professional culture concerns the forms of intervention carried out by mental health professionals in the lives of users of psychiatric services. Although psychiatric diagnosis has attracted some interest by discourse and social constructionist researchers (Barrett, 1988; Hak, 1989; Harper, 1994b; Soyland, 1994b), psychotherapy has been the only treatment examined in any depth (e.g. Soal & Kottler,1996) with 'physical' treatments like psychiatric medication attracting relatively little interest. This is perhaps understandable given the tendency to privilege the linguistic in discursive methodologies. However, such research does not adequately represent the experience of most users of psychiatric services who tend to get drugs or Electro-Convulsive Therapy (ECT) as a first line treatment rather than psychotherapy (Rogers et al.,1993). Whilst critical psychologists have begun to critique the assumptions of dominant psychiatric concepts like 'schizophrenia' (eg Bentall, 1990 and Boyle, 1990), medication has received little analytic attention (although see Harrop et al. ,1996 and Holmes & Newnes, 1996) whilst commentators often seem to be divided into camps either seeing the case for medication as 'unarguable' or as always harmful.

The meanings of psychiatric interventions, like diagnosis, are contested and they too, like diagnoses, delineate certain subjectivities. Indeed, these interventions also contribute to the construction of paranoia as a diagnostic entity. The kinds of solutions offered to effect change in 'symptoms of paranoia' convey certain assumptions about the nature of paranoia, for example as primarily a matter of biology, of social and environmental influences or of internal psychological functioning.

Often such interventions can be conveyed as forms of 'treatment' which implies a range both of medical discursive positions (eg illness; diagnosis; recovery and cure) and of subject positions (eg doctor and patient). Symptoms are seen as signs of underlying pathology and the locus of treatment is seen as the particular remedy used. Alternatively, interventions can be constructed as forms of 'help' -- we talk, for example, of 'helping professions'. This again sets up a number of positions: one who is doing the helping or one who is being helped (Edelman, 1977). Such discourse constructs various interventions not only as varieties of action but also as particular objects and subjects. This occurs not simply through language, of course, but also through institutional apparatuses -- for example the prerogative of doctors to prescribe medication(2).

The two dominant forms of professional intervention which arose in the interviews and which are to be discussed here were medication and a wide variety of forms of talking with professionals -- ranging from users discussing progress at regular out-patient appointments with their psychiatrist to more formalised modes of 'talking treatments' (eg psychotherapy and counselling). Although there are obviously many other forms of professional intervention (for example day care centres, occupational therapy activities, supported employment schemes and so on) and other sites where professional rhetoric creates certain identities (eg care programme and care management meetings), both these forms of intervention have been reported as dominant in general psychiatric treatment in previous research and a focus on these two specific areas allows a more detailed analysis to be conducted. Rogers et al. (1993) reported on both these kinds of professional help but noted that medication was the primary form of treatment most users received and as a result the present analysis will focus mainly on medication since it was the main intervention most of the service-user interviewees had experienced at the time of interview. Although interventions will vary in their description from professional to professional and from user to user (with the term counselling, for example, meaning many different things) the aim here will be to explore the variety of positions set up within forms of intervention and to analyse some of the interests served by such constructions. Although both medication and talking treatments will be examined the focus here will primarily be on medication since few of the service user participants had received formal talking treatments: medication had been the primary intervention they had received.

6.2 DISCOURSE, MEDICATION AND THE BODY

6.2.1 From medicine to meaning

In much of the literature relating to medication there is a dominant received medical view which is empiricist, objectifying and positivistic and which views psychiatric drugs as given to treat medical conditions which are seen as objectively independent of observation and theory. These drugs are thought to target pathological entities (symptoms and diseases) within individuals. Those from the social sciences wishing to take a different approach to medication have also adopted a positivistic empiricist approach for example by taking issue with the reliability and validity of diagnosis (eg Bentall et al., 1988). Others have taken an approach focusing on the meaning of medication.

6.2.2 The move to meaning

Gabe (1991) has made the case for focusing on the 'social meanings which tranquillisers have for patients and their doctors' (p.33) defining meaning as the interpretation a person gives an object or event in their life. A wide range of work has developed in this paradigm. Thus Sherlock & Kielich (1991) noted how, in their interviews with patients considered to be chronically mentally ill, there was almost complete acceptance of a medical model of distress which they interpreted as a mechanism by which patients normalized their social status as patient rather than as mentally ill. Moreover they felt that a medical view enabled patients to reinterpret current and past difficulties in life as problems of prescription type and dosage. Other studies have focused on the meaning of prescribing medicines (Amarasingham, 1980; Monsonego, 1994; Uddenberg, 1990).

Rhodes (1984) explored the metaphors used by doctors and patients to describe the effects of psychiatric medication, arguing that such metaphors provided a way to communicate about issues of identity and change and were a major way in which clinical experiences were given shared meanings. Thus patients talked of how psychotropic medication 'released', 'blocked', 'straightened' and 'distorted' thoughts. Professionals talked of tranquillisers as a 'filter'. Montagne (1991) has listed a wide variety of images and metaphors for tranquillisers and has noted how such metaphors, over time, move from 'as' to 'is', ceasing to be analogous and actually coming to stand for the drug, thus structuring patients' and professionals' knowledge. Such accounts highlight how much of medical conversation is deeply metaphorical. Barham & Hayward's (1995) account has described the complexity of users' views about medication, showing how such views are not homogenous and demonstrating the wide variety of influences on users' judgement of efficacy -- this included the attitude psychiatrists had about medication -- a point noted by Rogers & Pilgrim (1993).

However, there are a number of problems with this research. First, work such as Barham & Hayward (1995) simply focuses on the views of users, rather than on those of professionals (who, in terms of medication at least, have the legal power under certain conditions to give treatment regardless of users' views). Thus there is an inadequate analysis of power in the relations between psychiatry and its patients. Even where studies did focus on professionals, it was often on what they said, rather than on what they did or what they did through saying as Atkinson (1995) has noted. Second, as the earlier quote from Gabe demonstrated, meaning in much of this work is seen as: the property of an individual; or as a mentalistic construct (eg a 'belief'); or as overly metaphorical with little connection to the material -- and thus problems would be encountered in conceptualising issues like Tardive dyskenaesia (Hill, 1992) and neuroleptic side effects (Day & Bentall, 1996). This causes difficulties for those who then go on to theorise links between individuals' accounts and society (eg Montagne, 1991; Gabe, 1991). Moreover, many of the metaphors discussed seem quite tame. Others, for example, have talked of medication often being seen as a means of control (Light, 1980), as a punishment (Newnes & Holmes, 1996). Third, many of these studies had difficulty in explaining apparently contradictory accounts. Much of the research attempted to pigeon-hole users and professionals into pro- or anti- drug categories while finding instead that contradiction and ambivalence were a recurring theme (Gabe, 1991). A final problem is that, whilst some researchers saw a focus on the subjective as a way of critiquing objectivism its actual effect was to set up a dualism by remaining agnostic with respect to the physical effects of drugs. Yardley (1996) has noted the danger that in 'constructing a psychosocial discourse about health and illness in which physical being has no real presence or force, [that] social scientists themselves surrender the physical realm to medicine' (p.492).

6.2.3 Stubborn materiality

Paralleling the focus on meaning in the social science literature on medication is a debate about the reality of diagnosis and the real effects of drugs. This debate concerns a number of issues. Firstly, there was a changing understanding of drug action. Research attempting to answer the question of variability in drug effects and side effects saw psychopharmacologists increasingly accounting for such variation with recourse to behavioural factors like 'reinforcement history' (eg Iversen et al., 1987) or cognitive factors like 'expectancy effects' (eg the so-called 'placebo' effect see Straus & Cavanaugh, 1996). Such research began to question the received wisdom that drug actions were only 'chemical' reactions (which usually implied a crude, reductionistic and overly simplistic chemical model) bypassing such factors. Another line of research concerned the actual practice of prescription since there appeared to be much variability in prescribing practices (Davis, 1994). Researchers like Rogers et al. (1993) and Lacey (1991) described 'irrational' and 'production-line' prescribing, 'polypharmacy' and 'megadosing' (ie prescribing drugs above the limits noted in the British National Formulary). Mihill (1994) reported a statement from MIND which noted that one death a week in the UK was caused by powerful tranquillisers and other drugs given for psychiatric treatment. Bradley (1997) notes that prior to the introduction of NHS indemnity in 1990 an analysis of 50 consecutive claims against psychiatrists dealt with by the Medical Protection Society (a professional insurance scheme for doctors) showed that 32% arose from alleged negligent use of drugs. Holmes & Newnes (1996) cite Breggin (1983) describing major tranquillisers as 'neurotoxic', leading to a wide variety of neurologic disorders including 'acute dyskinesias or dystonias, chronic involuntary movements or continuous dyskinesias and parkinsonism' (Holmes & Newnes, 1996, p.14). Hill (1992) and Breggin (1996) warned of the dangers of Neuroleptic Malignant Syndrome. Indeed Newnes & Holmes (1996) have argued that many of the side effects of these medications could be misdiagnosed as further symptoms of schizophrenia. Such concerns have been expressed not only by critics but also by psychiatrists conducting empirical research (Bollini et al., 1994; Carpenter, 1996; Carpenter et al., 1987; Chaplin & McGuigan, 1996) and by reviewers (Day & Bentall, 1996; Healy, 1997). Thus Warner (1985) noted that anti-psychotic drugs were unnecessary or harmful in the treatment of a substantial proportion of 'schizophrenics'. There had been extensive critical work arguing against the dangers of a medicated society (eg Illich, 1977) which also noted the role of multinational pharmaceutical companies (part of what Breggin has termed the 'Psychopharmaceutical complex'). Their power has increasingly come under the spotlight. Newnes & Holmes (1996) have disclosed that 18 of the 21 members of the Committee of Safety of Medicines have declared financial interests in the pharmaceutical industry and more than half received consultancy payments from drug companies. Prather (1991) has noted there is one pharmaceutical sales representative for every eighteen physicians in the UK. Greenwood (1994) has estimated that there are over 4500 such representatives, making one for every 6 GPs -- with over £5000 per year being spent on promotional material -- and that such representatives do influence prescribing behaviour. Shaughnessy & Slawson (1996) have pointed out how much of the information supplied by them is inaccurate -- they describe one study which reported that one in ten statements (all of which were in favour of representatives' products) were at odds with the company's own literature.

Another issue concerns the lack of philosophical and psychological sophistication in biological psychiatric theorising(3). Harrop et al. (1996) argue that most biological research assumes that physiological changes cause psychological events (eg delusions, hallucinations etc). They make the case that psychological events may cause physiological change. This redresses the startling lack in most traditional biological texts of explanation of the causes of supposed chemical imbalances in the brain. They argue that reported physiological differences between those diagnosed as schizophrenic and those considered normal should be explained not by a simplistic biological cause psychological effect model but rather by one which highlights how 'the physiological and psychological work in a reciprocal and iterative fashion' (1996, p.643). Suggesting a Copernican shift in biological psychiatry, they illustrate a number of processes by which 'psychological' factors can and do affect neurochemistry and brain structure. They also take issue with the sloppy theorising whereby researchers attempt in a post hoc fashion, on the basis of the effects of neuroleptic medications, to explain the neurochemical deficits that caused the problem the medication attempts to redress. Headaches, they note, are not caused by an aspirin deficiency!

In maintaining, for the most part, an agnosticism about such important structural and 'material' questions, research on the meaning of medication has run the risk of perpetuating dualistic accounts. However, at the same time, many attempts to examine the reality of drugs have also been flawed. For example, the groundbreaking study by Rogers et al. (1993) failed to include an adequate explanation for why users seemed ambivalent and contradictory in many of their responses concerning medication.

6.2.4 Towards a discursive reconciliation

In chapter 3 I proposed a discursive approach as one which opened up a possibility to transform dualistic research. One interesting development is that of discursive-materialist accounts. Yardley (1996) has proposed this in relation to physical health and Stoppard (1998) has used this approach in her analysis of depression in women. Taking a discursive perspective on the body does not mean simply agreeing with, for example, everything doctors say, since we also look here for how accounts of the body function, especially in terms of power.

Talk about medication exists in a matrix of other discourses and institutional relations, for example of medical power, the regulation of the self, the relationship between mind and body and so on. An important part of this matrix are drug companies and Breggin (1996) terms the link between these companies and mental health institutions the 'Psycho-pharmaceutical complex'. One needs to have relatively little exposure to mental health practice to see that certain discourses are privileged by this complex in adverts in doctor's journals, the sponsoring of conferences and so on. Often professionals (and, indeed, many users and relatives) get most of their information about medication from pharmaceutical representatives, so-called 'drug reps'. Discourse about medication is, therefore, a contested area with accounts vying for exposure and credibility and is thus an attractive arena within which to use DA.

Following a discursive approach leads us to ask certain questions. For example, how is medication used as a linguistic resource by both doctors and users to explain particular events? What kinds of accounts are drawn on to explain the effects of medication (eg psychological, biological and so on)? Again, this is not to say that the factual status of accounts are not important but that they are not the focus here. Rather, the analyst's task is to ask what the effects achieved by these accounts would be if they were not true? How are factual accounts designed to perform particular actions? And how are they designed to accomplish their factual status? What rhetorical devices (cf Edwards & Potter, 1992) are used?

Such a focus on accounts can help us to understand apparently contradictory findings about medication. Thus Gabe & Thorogood (1986) have described how many users of tranquillisers were 'ambivalent' about medication but found it difficult to theorise this. Day et al. (1996) instead focused on users' accounts and found there were a number of different narratives employed by different users: an 'unquestioning, uncomplaining, dependent' factor which accepted medical authority and saw medication as essential for prevention of relapse; an 'autonomous, sceptical' factor which viewed neuroleptic medication negatively; a 'balanced appraisal' factor which weighed up the pros (reducing fear of relapse) and cons (side effects) of medication; and an 'autonomous, responding' factor noting real benefits of neuroleptics but with little fear of relapse and little dependence.

Professionals too, of course, draw on a number of narratives in their practice and Atkinson (1995) has noted how doctors use a number of different strategies to manage the uncertainty and ambiguity present in their work. He has noted how three different kinds of account may be used: a rhetoric of science, of text-books and of experience. For instance, in his earlier work, Atkinson (1981) noted 'pharmacology is a topic where "experience" is frequently drawn on in justifying or condemning the use of particular drugs or dosages' (p.112). Light (1980) has noted how this is present too in the work of psychiatrists. Similar rhetorical strategies have been reported in other work, for example on case presentations (Anspach, 1988; Soyland, 1994b), and in diagnosis (Barrett, 1988; Berg, 1992; Harper, 1994b; Hak, 1989). Indeed a number of researchers have argued that discourse about theory and discourse about practice should be treated as two very different kinds of accounts. Thus, in his analysis of asthma, Gabbay has asserted 'the striking thing is that therapeutic practice was largely unaffected by the comings and goings of theories' (1982, p.26), a point reinforced by Potter (1982).

6.3 TABLET TALK AND DEPOT(4) DISCOURSE: RHETORICAL STRATEGIES IN THE CONCOURSE OF MEDICATION DISCOURSE

In the interviews conducted here, professionals and users of services positioned themselves in a wide variety of ways, drawing on a range of discursive resources to construct various accounts of medication. As in the previous chapters, the aim here will be to outline and briefly describe some of the positions and resources available within the concourse of talk about medication and so I have attempted to map out general positions from the interviews, drawing on my analysis of wider material from popular, critical and professional literatures. It is not the intention here to suggest that this concourse exhaustively lists all of the positions and stories possible about medication -- rather it serves a hermeneutic value in providing a way into an analysis of the interviews: they are an entry point for 'troubling' (Curt, 1994) medication talk. Nor am I suggesting that each story is unrelated to the others -- instead we will see that each is intimately and iteratively linked. The aim is to illustrate some of the 'rules' of discourse followed by users and professionals in these interviews: what can or cannot be said from their positions? We will see that there is much flexibility in this discourse and that many times professionals and users have access to similar discourses. However since users and professionals are in different positions in relation to power, their use of similar narratives will have different meanings and will be heard differently because of their differing institutional locations. Again here I have described discursive positions through the use of binary oppositions.

6.3.1 Some comments on rhetorical features and devices in medication talk

Before describing the binary oppositions some preliminary comments are in order to set the scene for some of the examples below. As a mental health professional who has worked in mental health services for some time it is easy to forget the strange power of talk about medication. Gabbay (1982) has talked of the 'recuperative' power of medical language, of how it can seem to speak a literal and non-metaphorical truth. We will see that in talking about medication certain realities are purveyed through the creation of certain objects and subjects and the construction of certain kinds of agency and subjectivity and reality including what Potter (1996b) has called 'out-there-ness'. As with my earlier analyses I have found the notion of rhetorical strategies useful. However, since I have argued that a discourse analysis needs to be politically-informed to be useful and thus requires both an analysis of culture and of institutional power (cf. Gavey 1992), a focus on rhetorical strategies alone cannot be enough since strategies may be used differently to achieve similar effects or may be changed -- witness the rise of the 'new sexism' and 'new racism' Thus, in my second layer of analysis I wish to be sensitive to the ways in which talk about medication constructs a number of subject positions for users and professionals whilst constructing a range of objects. Before moving onto that, however, there are a number of devices which occured frequently in the interviews since they formed part of a background discourse about medication and are worthy of note here.

6.3.1.1 Drug names

One obvious feature is the use of the somewhat obscure-sounding names of the medications. Names may be used in a variety of ways: to describe generic pharmacological compounds or as a trade name; for example the anti-depressant Fluoxetine is known more widely by its trade name (which is therefore a trademark) Prozac. For the academic or professional reader, it is easy to forget how strange these names are but some users have difficulty recalling the names of the drugs they are prescribed. Take the following example for instance:

Dave: What, what medication are you on?

Peter: Er <pause> Dofepin, is it Dofepin? <Dave: Dothiepin?> Oh aye, yeah, and something else, I can't think of it, it's a big name, like, blue pills. <Pause> Only for sixteen, seventeen [inaudible]

Peter Shaw (Service-user)

Here Peter cannot recall the name of the drug and so resorts to describing the size and colour of the pills. In a recognition of this fact MIND's The Complete Guide to Psychiatric Drugs (Lacey, 1991) includes descriptions of the size, colour and markings of psychiatric medication. Newnes & Holmes' interviewee states 'I think I've been misled about the drugs I've been taking because of all the different chemical names they have' (1996, p.3). What exchanges like this seemed to signify was that information about medication was constructed as a form of expert knowledge(5). Moreover drug names sound scientific and convey an image of a confident and optimistic medical solution to a difficulty. Here Peter does not note for what purpose the medication is prescribed -- it is unclear whether he thinks it is an anti-depressant, an anti-psychotic or some other medication. Again, as Newnes & Holmes (1996) have pointed out, if Peter cannot even accurately recall the name of the drug he is on he is unlikely to recall the known side effects of this drug (this does not emerge during the rest of the interview either). Thus knowledge of side effects is marked out as specialist and expert knowledge possessed only by professionals. Further features of interest emerge from the interview with Dr Williams who was involved in Peter's care for a time:

delusions were, you know, cleared off with the medication. He's on Stelazine 5mg bd and Dothiepin. So <Dave: Right> [inaudible] he suffered from an episode of psychotic depression <Dave: Uh-huh> and erm later on he went in remission so <Dave: Right> he's doing quite well I suppose.

Dr Williams (SHO in Psychiatry)

6.3.1.2 Dosage and quantificatory rhetoric

The use of numbers in this extract is a particularly powerful way of constructing an empiricist account. Edwards & Potter (1992) and Potter (1996b) define such an account as one where objectifying scientific language is used and phenomena are treated as having agency in their own right whilst people are seen as passive agents. The use of a somewhat mysterious drug name is combined with a measure of quantity which again sounds scientific. The use of "5mg" conveys that there is a highly technical scientific procedure going on here (although dosage decisions may be much more arbitrary in practice as will be seen below). Of course, quantification is a common rhetorical device in much science talk (Potter et al., 1991).

6.3.1.3 The use of Latin words for dosage information

The term "bd" in the above extract is medical shorthand for the Latin bis in die sumendum which means that the medication is to be taken twice daily. Here again is another item marking the drug out as scientific and in the realm of experts -- Latin and Greek terms are often connected in the popular mind to science and scholarship. Lacey (1991) even includes a guide to translating these Latin terms into English. Not even the full Latin is used which might be a little less opaque, rather the use of the shorthand term suggests this is a cultural marker too: knowledge about the medication is restricted to the medical and para-medical professions. I am not trying to argue that Peter may not have been told when to take the medication in English but simply that this type of discourse is available and permeates much talk about medication between professionals.

6.3.1.4 The conflation of professional intention and drug action: the construction of 'side effects'

down to. <Dave: Right> Erm <pause> but he was well for, I think it was, it may have even been eighteen months or something like that but he was out for a considerable amount of time <Dave: Uh-huh> er problem-free. <Dave: Uh-huh> Erm <pause> but then <pause> the side-effects of the drugs, I mean, he came down and we noticed his jaw was going, we said, you know, sort of "Tardive dyskinesia" which, which we were concerned about. We discussed it with the family, discussed it with him and it was, it wasn't all Ian's choice to come off it, I

Terry Reid (CPN)

The term 'side effect' is a fascinating one since the effects Terry describes are direct effects of the drug. However, they are not part of the intended effect of the drug and the occurrence of these effects is thought to vary across people -- although the reasons for this are not clear. The description of this effect as a 'side effect' constructs certain effects of the drug as intended and proper and minimises other effects of the drug as unintended and somehow wrong. This is reminiscent of the description of civilian casualties in the Gulf War as 'collateral damage' in two ways. Firstly it allows some emotional distancing and objectifying of the phenomenon: 'side effects' does not convey the horrors of tardive dyskenesia just as 'collateral damage' does not convey the same as 'thousands of dead civilians'. Secondly it allows speakers to avoid responsibility by claiming these effects are not intended, rather like the Desert Storm allied military claiming it had not intended there to be civilian casualties.

6.3.1.5 Maintenance, monitoring and the management of risk

Uh-huh> you know the best sort of care. <Dave: Uh-huh> And on the other hand, I don't feel pressurised to give er medication but a lot of our job is to give people maintenance medication. I mean that's quite a large part of it <laughs> and mon-, to monitor that. Erm there are patients that, that we see who erm <pause> we suggest that they actually are reduced, you know, a gradual thing we, we don't and in fact most of the patients that we see are on the lowest doses for erm I worked in general psychiatry and it's, and I've seen people on far higher doses than the patients that we're seeing now and I think it's because we do review them regularly, we do review the medication regularly <Dave: Uh-huh> and erm you

Terry Reid (CPN)

Another common feature of interviews concerned assumptions about medication. Important here was the concept of 'maintenance medication' to describe the need for a user to continue taking their medication, possibly for the rest of their life, in order to prevent a relapse (ie a recurrence of their symptoms). Thus Carpenter (1996) talks of 'maintenance treatment' and 'maintenance therapy'. Most people are used to the notion of having 'a course' of tablets of a reasonably short and fixed duration in order to achieve a particular resolution of their symptoms, what they might call a cure. Maintenance implies a chronic, long-term problem that is not being resolved in any fundamental way but is simply being maintained with the danger that it might unexpectedly begin again. In the extract above, Terry explains how giving maintenance medication is an important part of his role. He also explains the need to monitor the medication and to try to reduce the dosage. What is not mentioned here is the reason for monitoring and reducing dosage. What is not being said here is that these drugs are risky and dangerous. Indeed, we saw in the previous extract, Terry's description of some of the 'side effects' generated by these drugs. However, the term 'monitoring' serves a useful function in allowing this risky treatment to take place since risk is being minimised through monitoring.

6.4 OPPOSITIONS AND DILEMMAS IN MEDICATION TALK

6.4.1 Medication works/doesn't work

This binary opposition includes accounts that are characterised by either establishing that medication 'works' (ie that it is effective) or in establishing that medication does not 'work'.

6.4.1.1 Medication works

In this first half of the binary, the use of medication is seen to be largely unproblematic and to be straightforwardly effective. This talk is characterised by an empiricist form of accounting. In this case objectivity is achieved by focusing on users' symptoms with the symptoms being treated as having agency rather than the users who are often regarded as passive experiencers of these symptoms. Thus certain aspects of people's experiences are reified into symptoms according to a diagnostic schedule (Barrett, 1988) and changes in those aspects of experience tend to be linked only to the medication rather than the rest of the user's life. As a result, as Sherlock & Kielich (1991), have noted, the complexity of everyday life may come to be transformed into accounts of the prescription and dosage of medication. The effectiveness of medication is judged according to whether the symptoms disappear or are decreased in some way. Note that here there is an element of quantification in talk about symptoms. These themes also emerge in oppositions 6.4.2 and 6.4.6. Often such accounts assume an obviousness about medication as being the only appropriate treatment (and here there are links to binary 6.4.2.1). Another feature of these accounts are that they are optimistic -- thus even if there are problems at the moment, future medical advances will iron them out (this is similar to binary 6.4.3). Such optimism is reflected in the narrative which claims that drug treatments were responsible for decarceration, a narrative strongly rebutted as a myth by historical researchers like Andrew Scull (Day & Bentall, 1996; Light, 1980, p.9).

Such kinds of accounts exist in the professional literature. For example, Munro & Mok (1995) aim to provide an overview of treatment for paranoia/delusional disorder but include only drug treatments in this discussion without even a note indicating other possible treatments (eg social support, psychological interventions and so on) and so making it seem obvious that treatment = medication. Even here they focus only on Pimozide since it appeared in most reports surveyed to be the 'commonest drug of first choice' (p.616). They measure success in terms of 'recovery' or 'partial recovery' from symptoms (although there is no detail on how this is measured) and offer an up-beat account noting that delusional disorder, previously regarded as 'treatment-resistant' is 'if well-treated, [a] disorder with an optimistic outlook' (p.619). They suggest that Pimozide is 'probably the first-line treatment at this time'(6) and note that since delusional disorder is chronic 'whatever the neuroleptic used, it will require long-term and possibly permanent administration' (p.619).

Dave: Right, right and his symptoms at the moment? Do-, does he er/

Dr Smith: /He hasnt got, he hasnt got any. He has been on treatment continuously at least for the last two years. <Dave: Right> None whatsoever. <Dave: So what erm/> /He now understands erm that his beliefs were false.

Dave: Right. And erm the reason that he doesnt have any symptoms now? What would you put that down to?

Dr Smith: He is on anti-psychotic er depot erm injection.

Dr Smith (Consultant Psychiatrist)

Here most of the features of this form of accounting are present. First, there is the use of what we shall call symptom-talk which is first introduced in the extract by me. We might define symptom-talk as a form of narrative where focus is placed on a narrow range of observable and inferred phenomena (symptoms) and which views these phenomena as both possessing agency and being 'surface' signs or markers of 'deeper' pathology (ie illness or disease) within the individual. Whereas my question leaves open a possible range of answers that the symptoms had changed in some way, Dr Smith says that there are none and repeats this. There is an implied optimism here -- there is no qualifying remark like 'gone for the time being' but a simple statement. This kind of objectifying talk allows an experience to be characterised in such a manner that it can either be present or absent. Moreover, as well as the symptoms being removed, the user (Mike Sullivan) is described as having insight (line 248), which constructs him not just as a passive recipient of symptoms but as an agent in his own right. There is an implied authority which appears to draw its influence from the operation of a form of category entitlement (Edwards & Potter, 1992): Dr Smith offers no evidence as to how he knows the symptoms are gone. Finally, there is the linking of this removal of the symptoms with treatment (lines 246-247). Where I have simply asked about the symptoms Dr Smith volunteers information on treatment. Medication is not used as an explicit explanation but, in this context, it appears to function as one. Indeed when I ask for his explanation Dr Smith again gives a statement (line 251) 'he is on ...' rather than 'because he is on ... [anti-psychotic medication]'. There is also a conflation of medication and treatment, highlighting the obviousness that medication is the only possible treatment.

6.4.1.2 Medication doesn't work

The other side of this opposition encloses a space for accounts about medication not working. This half of the binary covers a variety of positions: those noting that a particular medication has not worked; those noting that medication does not work in a simplistic symptom-removal way in many cases; and those arguing that medication is given too much emphasis as the major, or only, treatment. Here then there are links to binaries 6.4.2.2 and 6.4.6.2. These different positions produce different effects: the argument against symptom-removal warrants the need for further intervention. There are traditions within psychiatry that hold that medication cannot 'work' since it only leads to suppression of symptoms rather than addressing more fundamental problems within the individual. Light has observed that from this perspective drugs may impede therapy (1980, p.84) or at best only provide 'surface relief to symptoms ... leaving untouched the root of the symptoms, which could only be treated by psychodynamic therapy' (p.7, edited).

Within the current psychiatric literature, much attention is focused on addressing symptoms per se rather than any underlying disease. But when medication does not even appear to affect symptoms, the use of this narrative provides a warrant for other kinds of professional intervention like counselling or the increasingly popular use of cognitive-behaviour therapy -- the latter being claimed to be of particular utility with 'drug-resistant' disorders (Turkington et al., 1996). The psychiatric literature allows for the possibility of medication not working through the use of terms like this and 'treatment-resistant' patients (Kane, 1996a -- 'treatment' here only denoting drug treatment). However, such terms position the locus of responsibility in the disorder or within the user and not in the drug or the doctor hence deflecting any possible criticism. We will see in 6.4.6.2 a closer analysis of the strategies used to deal with the apparent failure of medication. Other lines of narrative available within this binary include a critique of the medicalisation of problems per se (eg Illich, 1977) which is discussed in more detail in 6.4.3.2.

Dr Lloyd: <Coughs> Yeah, I had hoped that with the neuroleptics his ideas about these things would change. <Dave: Uh-huh> They haven't. Ultimately, I don't really mind. <Pause> What I want to be able to do is reassure myself that Alan can have a life which gives him some quality and some degree of autonomy beyond what he's got is part of that erm and <pause> in terms of freedom to come and go as he wants from the hostel that we're sending him to erm to go and buy what he wants erm. <Pause> So <pause> if the delusional beliefs remain, as long as he agrees to continue taking the medication so the other aspects don't collapse <Dave: Uh-huh> I really don't care. <Dave: Right, right> Erm and he knows I don't care <Dave: Right> 'cos I've said "I'm going to agree to disagree with you" erm "we both have a different construction of your life and its meaning" erm "but here are the ground-rules".

Dr Lloyd (Consultant Psychiatrist)

Dr Lloyd notes how he 'had hoped' that the medication would impact on the symptoms. This could be seen as a rhetorical innoculation: if he had simply said he was not bothered I could have questioned him further by asking why he judged symptom change in other users important (and there is evidence of this in the transcript of the rest of the interview). The relative importance of the lack of symptom change is devalued through the claim that 'ultimately, I don't really mind'. This phrase constructs a hierarchical notion of change: that whilst change in the delusional symptoms are important, other changes may be more important. This resolves the dilemma here since otherwise Dr Lloyd might have had to say that there was no change whatsoever. Here then, the definition of outcome is flexible: it may relate to delusional symptoms, or other symptoms. Moreover, the other changes appear broader and less quantifiable than delusional symptoms: having 'a life which gives him some quality and some degree of autonomy'. Indeed he makes an even stronger claim here: 'I really don't care'. This assertion would be heard differently and would function differently if spoken by the user or even perhaps other professionals (especially those further down the status hierarchy). This statement could be seen as an example of a category entitlement: it has more authority when coming from an expert consultant psychiatrist - if he doesn't care then the matter must not be that important.

6.4.2 Medication is the only obvious treatment/ isn't the only thing

This opposition concerns the relative importance of medication as a treatment.

6.4.2.1 Medication is the only obvious treatment

This first half of the binary is similar to 6.4.1.1 in that it features elements of empiricist accounting and optimism. The element of the obviousness of medication as the only real treatment is highlighted either overtly by arguing against the validity (or at least importance) of other interventions or more implicitly either by the omission of other alternatives or through the use of the word 'treatment' to mean only medication. Thus Holmes & Newnes (1996) note that whilst other aspects of psychiatry can be questioned, the need for the use of drugs is still seen as 'unarguable' and they are still seen as the treatment of first recourse. We saw in the extract in 6.4.1.1 how the term 'treatment' was seen as synonymous with medication and I noted how, although Munro & Mok (1995) claim to discuss the treatment of paranoia, they discuss only medication. Kane (1996a) is similar in this respect. Texts such as these work to present medication as an obvious taken-for-granted treatment, indeed as the only possible treatment with other options ignored. The possibility of 'treatment' sets up a range of medical discursive positions. Moreover, it marks this territory out as properly medical. There is an underlying optimistic empiricism in articles like those of Munro & Mok (1995) although there is also the use of a systematic vagueness in that terms like 'recovery' are not defined in any detail and this may reflect the operation of category entitlement.

Dave: Right, right. Do you feel the kind of psychiatric treatment that he's got has been okay? Has his er?

Dr Chapman: Yeah. <Pause> Basically it's a chemical treatment. I mean <Dave: Uh-huh> he's not had any, I don't know if it does responds to counselling. I don't think it does, does it? Counselling or psychotherapy, whatever you do? I think it is a natural chemical illness and treated by chemical means.

Dr Chapman (GP)

Dr Chapman acknowledges how treatment may not just be medication-based by highlighting that the treatment in this case is a 'chemical' one. He engages in a variant of symptom talk by talking about the user's problems as an 'it'. Dr Chapman qualifies his statement about the probable ineffectiveness of counselling by using the phrases 'I don't know', 'I don't think it does' which may be explained by the fact that he is talking to a clinical psychologist. In humorously describing counselling as 'whatever you do' there may be some use of medical authority to downplay the importance of this against medication.

Dr Chapman's statement that 'it's a natural chemical illness' is interesting. It is unclear here how 'natural', 'chemical' and 'illness' go together here. However one effect is to position this problem as very firmly within the remit of a medical and specifically a medication-based intervention since the latter two suggest only a doctor could be involved. As well as presenting a medical solution as the only appropriate one, it also places responsibility for cause within 'chemicals' which are seen as having agency. Some writers have described this as having the effect of ignoring the social and political causes of distress (Newnes & Holmes, 1996)(7). The word ' natural' suggests the problem is something to do with biological structure. This is allied to what psychiatric service users say they have been told by professionals: that they are suffering from a chemical imbalance in the brain which needs to be corrected by chemical means. Thus in their appendix giving information on what causes schizophrenia, Barraclough & Tarrier (1997) state that 'it seems that chemicals in the brain are affected' (p.231).

6.4.2.2 Medication isn't the only thing

This half of the binary reflects a position where medication is not seen as excluding other forms of intervention. Here, the word 'treatment' is seen as encompassing a variety of forms of professional (and other) forms of help. For example Manschreck (1996) notes that in the case of delusional disorder 'treatment may require psychotherapy, pharmacotherapy, and possibly hospitalization' (p. 36). Note how in being termed 'pharmacotherapy' drug treatment is positioned as just another kind of therapy. MIND similarly have noted that neuroleptics 'have a place, but only as one possible aspect of a person's treatment and care' (Mihill, 1994, p.4). Barham & Hayward (1995) have noted that whilst some users found medication helpful they argued that it was not the only thing that was needed from professionals and that it needed to be a part of a broader strategy since often medication reflected a lack of professional attention.

One effect of such narratives is once again to warrant the need for other forms of professional intervention (eg a therapeutic relationship with a professional and/or more specialised psychotherapy). This narrative was sometimes used as a way of addressing the issue of medication not having worked. Sometimes it also implied that a more political response was required (see for example the quote from a user in the Barham & Hayward study noted in 6.4.3.2). Another use for this narrative was in allowing users to maintain some autonomy over their lives by seeing medication as just one out of many other resources (Gabe & Thorogood, 1986) or as a 'standby' (Gabe & Lipshitz-Phillips, 1984).

Dave: <Pause> Okay. I mean just in general again, I mean what, what is, what's the best way of helping people who've got kind of paranoid or perceutory delusions do you think?

Edward: Best way of helping them? <Pause> Oh [inaudible word] a combination of things. I think for most people medication <Dave: Right> erm it's, it can play a very important part <pause> erm [inaudible] is to actually en-, engage in a relationship with them and it's often very fragile erm <pause> if you, if you can get to the point where you can get the trust of someone who's, who's [inaudible] paranoid beliefs you've done [inaudible] <laughs> <Dave: Uh-huh> You know it really is a skill and an art. Erm <pause> and I know that there's a few people in the Team that are just brilliant at it whereas I may, I struggle <Dave: Right> I mean, some people seem to have a way of actually just being able to, to hold onto that. <Pause> Erm I think if you can get someone to actually work with you probably the first thing you need to look at is to what the degree of conviction they actually have. How much room have you got? If it's a hundred per cent conviction you have a lot of difficulty in making any significant changes. All you can do, I would imagine is look at all the issue of support and try and check out some of the things that they feel quite threatened by. <Dave: Uh-huh> If the conviction is more where you can start to actually challenge directly and start to look at ways of erm challenging the beliefs erm themselves to check out whether or not they're true <pause> in whatever kind of format, it would depend on what the belief is and what's <Dave: Yeah, yeah> what you could work on to do that.

Edward Jackson (CPN)

This extract contains a number of features worthy of interest. First, medication is given primacy in that it is mentioned first but then the need for other aspects of professional involvement is stressed. Thus the importance of a relationship (and trust) is highlighted but this has the dual status of requiring professional expertise (a 'skill') and also the necessary personal qualities (an 'art'). Later in the extract a number of markers suggest that some kind of counselling is important. The phrases 'work with you' (line 253) and 'work on' (line 262), 'look at' (lines 254, 257 and 260), 'support' (line 257) and 'check out' (lines 257 and 260) are common in counselling parlance. One effect of this narrative is to warrant the kinds of intervention made by professionals like Edward in addition to medication.

6.4.3 Pharmaceutical faith and medical optimism/ pessimism about pills

This opposition concerns the extent to which medication is seen as a total solution. It is therefore closely linked to 6.4.2 but reflects more of a world-view about the place of medication in culture and the way that people construe difficulties (eg as biological faults or as social and human problems).

6.4.3.1 Pharmaceutical faith

In this half of the binary, 'psychological difficulties' tend to be portrayed as biochemical in origin and as if the main or best solution is a biochemical (ie pharmaceutical) one. Thus chemical and biological features are focused on and given agency. People's problems are seen as arising from chemical imbalances -- here chemicals have agency. A narrative form of accounting is used which leads from the proposition that there is a chemical cause to then argue that a chemical solution is the most appropriate. At the extreme this becomes reflected in phrases like 'a pill for every ill'. In this position there is much use of empiricist accounting of problems as symptoms which tend to be disussed both as if they were disconnected from people and their lives and as if they had agency in and of themselves. Medication is seen as working directly on these symptoms. However, alongside a narrative of science is a narrative of faith or ideological commitment. Many critics have elaborated on psychiatry as a faith (eg Szasz, 1970)(8). Indeed, Newnes & Holmes describe how drugs have 'attained a sacred status in psychiatry' (1996, p.1) and have argued that if schizophrenia was Thomas Szasz's sacred symbol of 1970s psychiatry then 'medication has become its holy water' (Holmes & Newnes, 1996, p.15). They note for example that there is no logical link anyway between a problem having a biochemical cause and needing a biochemical solution (see also Harrop et al., 1996; Ross & Pam, 1995). Indeed, Manschreck in a review of treatments for 'delusional disorder' notes that pharmacotherapy may be appropriate for these patients, [but] relevant data are limited' (1996, p.36). Speakers may also talk in terms of a preference for drug treatment (Virji & Britten, 1991), seeing it as a benefit of modern medicine (Britten, 1994).

These devices have a number of effects: they warrant the use of biochemical rather than other kinds of interventions. They allow flexibility so empirical challenges can be dealt with by recourse to narratives both of empiricism and of faith. Light (1980) for example, has noted how proponents of medication link decarceration to the arrival of neuroleptics and cast pre-neuroleptic interventions as medieval with an almost missionary rhetoric. For the most part such faith is likely to be implied in professional talk through the assumption of medication as an obvious treatment. Users too, may use a narrative of faith and, at times may be more explicit about this. Thus Gabe & Lipshitz-Phillips (1984) note how many tranquilliser users described their drug as a life-line, one that gave them peace of mind (Gabe, 1991).

Geoff: Well I have my tablets during the day as well, you see <Dave: Right, I see> <pause> the Largactil and the injection, Clopixol, does help <Dave: Uh-huh> together but on their own they're no good. See I've been on drugs now for eight years <Dave: Uh-huh> and it's mainly been Largactil but my injections they've changed em and changed em and changed em. <Dave: Right> They keep trying me on different things, you know. But this drug, this Clopixol <pause> has bad side effects, I believe, but it's like a wonder drug <Dave: Uh-huh> with Largactil it numbs the voices, you can't hear em and it makes you feel relaxed <Dave: Uh-huh> <pause> but lately, I've just had an injection I mean I'll contradict myself now 'cos it didn't work but only because I wasn't taking the Largactil as well. I've recently been to see my GP and she prescribed my Largactil and it seems a lot better now.

Dave: Right, right. When you say it's a wonder drug is that something you feel yourself or/

Geoff: /Yeah, I believe in it, you see. <Dave: Right, right> If the voices get really bad, I just want to have my injection, you know <Dave: Yeah>. Sort of depend on it, you know.

Geoff Nelson (Service-user)

This extract includes a number of the features noted above in that the only solution discussed here is one relying on medication. Symptoms are objectified and medication is seen as working on them directly ('it numbs the voices') -- see 6.4.5. The account includes notions of faith and belief which do not sit easily with an empiricist account. Thus Geoff talks of how the medication is 'a wonder drug' and how he believes in it. Both these narratives then construct an account which expresses some faith in medication as a solution and indeed, as an obvious one. When it is seen as not working he gives only drug-related reasons (lines 173-174) -- there will be further discussion of this aspect in 6.4.6.

6.4.3.2 Pessimism about pills

The other side of the binary reflects a lack of faith in medication and/or an ideology that sees psychological problems as social and political. Thus one of Barham & Hayward's service user interviewees noted 'I believe that [psychiatrists'] heads are full of chemistry and my head is full of politics and social things. So that's a conflict that perhaps won't ever be resolved' (Barham & Hayward (1995, p.68). A feature of this kind of account is that social, psychological and political influences are given agency and biochemical influences are either minimised or seen as caused by these other influences -- this can be seen in texts written by critics of medication, biomedicine and biopsychiatry (eg Illich, 1977; Newnes & Holmes, 1996) . Moreover, in this kind of narrative, speakers may note the influence of personal ideological commitment. Thus the contrast between this narrative and the previous one is similar to Gilbert & Mulkay's (1984) contingent and empiricist repertoires. These kind of extracts constructed a narrative that asserted that 'psychological problems' had social and psychological causes and therefore required social and psychological solutions. Chemical solutions are, in this kind of account, seen as crude attempts at control of symptoms rather than solutions to problems seen as more 'fundamental' or 'deeper' within the person (Britten, 1994; Light, 1980). This kind of account was not particularly well-represented in interviews with doctors and users. This is, perhaps, not that surprising since most of the users were seen as 'complying' with their medication regimes. Gabe & Lipshitz-Phillips have noted that 'those who talked about [benzodiazepines] only as a life-line were far more likely to express mixed feelings whereas those referring to them only as a standby were more likely to express consistently negative feelings about the drugs' (1984, p.531).

Another feature of this kind of account was that there was a lack of faith in and pessimism about medication as a solution. Thus a number of commentators have noted the existence of an anti-drug culture amongst users of medication with accounts characterised by a strong antipathy to using drugs (seen as 'unnatural chemicals') of any kind and fears of dangers associated with their use, like addiction. (Britten, 1994, 1996; Gabe & Bury, 1988; Gabe & Lipshitz-Phillips, 1982). Montagne (1991) also notes how drugs may be viewed negatively as crutches (p.51) or as a straitjacket or a prison (p.54).

Since an implication of using medication is that it leads users to 'deny or ignore the social concomitants of their distress, thereby helping to sustain "strained social systems" and minimize pressure for social change' (Gabe & Bury, 1988, p.327), one effect of this kind of account is to warrant some form of political analysis an intervention. A variant drawing on psychoanalytic and psychodynamic narratives warrants the use of non-pharmacological alternatives like talking treatments to move beyond control of symptoms to resolving what are seen as more fundamental psychological problems. Light notes:

to the extent that drugs 'worked' residents were constantly told by their seniors that drugs and other devices merely relieved symptoms. Even if the patient showed dramatic recovery and appeared to be 'normal' he [sic] was not considered to be much improved, and a relapse was predicted. Until one had 'worked through the dynamics' it was argued, substantial improvement was not possible.

(1980, p.263)

Others have noted that 'negative' accounts of medication may warrant 'positive' accounts relating to a preference for self-care (Virji & Britten, 1991). Breggin (1996) suggests that psycho-social interventions like psychotherapy and small volunteer-run residential settings are better alternatives to medication since a medicalised focus constructs recipients as defective objects rather than people struggling with emotional and social problems Thus, he argues 'people, not pills, are the only source of real help' (1996, p.65). This is then a powerful warrant for not using or reducing the use of medication.

I can't really enjoy my life the same because of this. It's a bit like, a little bit like being disabled, you know erm, only it's nobody can, nobody else can see it so they can't understand it 'cos it's inside my head. I think if it was something like some kind of, if I was handicapped or I was in a wheelchair, they could see it, they could do something about it but <pause> erm inside your head, it's not as easy. <Pause> 'Cos people outside, just see me outside wouldn't, wouldn't know me, know what's wrong with me. <Pause> That's why you feel so isolated with it, it's not something you can take a tablet for it to go away really <Dave: Uh-huh>.

Sharon Harrison (Service-user)

In this account Sharon describes her difficulties in social terms rather than biological ones. Here then social issues are given agency; thus understanding by others (line 455), the visibility of her problems (lines 458-459) and her isolation (line 459) are emphasised. A narrative accounting strategy, highlighting these non-chemical and non-biological factors, implies that a tablet is not the most effective way of addressing these problems. One effect of this account is to warrant the use of more social and psychological interventions to address these issues and to minimise the importance of medication.

6.4.4 Non-compliance as not following medical advice/ as an assertion of agency and choice

This opposition reflects diverging views about how to construe service users not taking their medication in the manner directed by their doctor.

6.4.4.1 Non-compliance as not following medical advice

The first half of the opposition construes medication not being taken as prescribed as 'non-compliance'. In such accounts users' failure to take medication is constructed by professionals both as an effect of their illness (ie due to a lack of 'insight') and as a sign of moral irresponsibility. A number of rhetorical features are noticeable in this kind of account. First, the prescription regime (drug, dose and so on) is constructed as 'out there' in some sense rather than as the product of a negotiation between doctor and user. Second, compliance is seen as an inherent characteristic of the user rather than of the doctor or of the doctor-patient relationship. Third, compliance is constructed as an either/or binary (compliant or not compliant) rather than as a continuum.

Similar to the way in which we saw the interaction between professional and user and the role of the professional's judgement obscured in chapter 5, the construction of compliance as an internal characeristic of the user has the effect that responsibility for 'non-compliance' is not placed on professionals or on a breakdown in the doctor-patient relationship. This account also provides a useful explanatory resource when professionals are accounting for the apparent failure of medication (see 6.4.6.2). Finally, by explaining such lack of compliance as due both to the illness and the user's desire, responsibility for it is placed on the user.

Dave: Right, right. With, with both her and Mike er at various points, they stopped treatment. <Dr Smith: Yes> Do you know why, why they did that?

Dr Smith: Well obviously in general I think people <pause> er dont like having any treatment erm and erm they believed that they shouldnt be on any treatment on a longer-term basis er quite often they have difficulty in accepting that they have got an underlying illness and they need the treatment to remain well.

Dave: Right. <Pause> Do they ever stop it because they say its not having any effect? <Dr Smith: Sometimes> Would that ever have been true for Sharon or Mike?

Dr Smith: I dont think that Mike Sullivan erm <pause> obviously when they have treatment they, they get better and they say "Well Im alright now, why should I have the treatment".

Dave: Right, right. <Dr Smith: Erm> <Pause> So they feel then that if they stop/ <Dr Smith: /Stop/> itll be okay. <Dr Smith: Yeah> And you, you would say what to that? You, you would/

Dr Smith: Well I would say "Your treatment is only controlling your symptoms. Without treatment erm er what I can tell you is er you are very likely going to relapse".

Dr Smith (Consultant Psychiatrist)

In this extract I introduce the notion of stopping treatment through which I set up a binary of stopping/not stopping which is part of the non-compliance narrative. My use of it here places responsibility on those 'stopping treatment', conveying an image of unco-operative users and localising compliance in the individual user. Dr Smith describes several reasons why users do this: they don't like treatment (lines 358-359); they believe they shouldn't be on it (line 359) and they don't accept there is a problem (lines 360-361). All these are constructed as being internal to the users described. Indeed at lines 365-366 the apparent changes in Mike and Sharon's views about this (which might be termed 'insight') are ascribed to the therapeutic effect of the medication itself. The non-compliance account is asserted most strongly in lines 371-373 where it is suggested, through the employment of a form of symptom talk, that medication only controls symptoms (rather than removing them) with the implication that they will return. Here agency is granted to symptoms which are seen as phenomena the user cannot control in any other way. The prescription regime is simply termed 'treatment' here which could be seen as a way of constructing the prescription as disinterested and 'out there'. The implication is that Mike and Sharon objected to treatment per se.

What are the consequences of such an account? First, through giving symptoms agency it provides the basis for prescribing 'maintenance' medication noted earlier. Second, by locating compliance as an internal characeristic of Mike and Sharon neither the doctor nor the doctor-patient relationship are seen as loci of responsibility. Third, through the linking of compliance with insight, what may be valid objections to drug treatment are de-legitimised. Finally, the extract works to construct non-compliance as an explanatory resource for apparent medication failure. Thus elsewhere in his interview Dr Smith says of Sharon 'she hasnt sort of co-operated er with treatment erm er so her ideas are still there' (Dr Smith: 307-308).

6.4.4.2 Non-compliance as an assertion of agency and choice

The second half of the binary contains a variety of alternative views about 'non-compliance'. In these accounts, prescription regimes are not seen as 'out there' but as the result of negotiation between doctor and user, a negotiation at times that might be seen as a battle or a power struggle (Barham & Hayward, 1995). Such accounts have a variety of features, seeing 'non-compliance' as: a defiance of medical authority; a breakdown in the relationship between doctor and users (Barham & Hayward, 1995); a self-protective response to negative side-effects (Rogers et al., 1993; Virji & Britten, 1994); a valid decision not to use such 'exceedingly dangerous drugs' (Breggin, 1996); or simply as 'self medication' -- an alternative use of medication which is more flexible than the one prescribed by the doctor, perhaps only taking medication when symptoms arise (Virji & Britten, 1994). Within this half of the binary then, the compliance narrative could be seen as an ideology that assumes and justifies medical authority since it defines adherence to a regime from the point of view of professionals (Conrad, 1985; Trostle, 1988). Here, adherence to a prescription regime is not seen as an internal characteristic of the user. Conrad has argued on the basis of in-depth interviews with patients that the issue of self-regulation is paramount and that this can be affected by a desire for testing dependence, aiming for destigmatization and having a more structured regimen by asserting control over their disorder or their life (Conrad, 1987). Moreover, the use of drugs is seen here as on a continuum from non-use to occasional, to permanent and frequent use. These accounts could be seen as serving a number of functions. First, by locating responsibility for non-adherence in the doctor-patient relationship or in the doctor's actions (eg over-reliance on medication by doctors -- Barham & Hayward, 1995) doctor-user negotiation is warranted. Second, such accounts warrant the withdrawal of users' co-operation if such negotiations are not satisfactory, through reducing their dose or refusing to take medication. Third, users' expertise and rights are seen as valued. Fourth, users are able to assert their own agency through deciding their prescription regimen themselves which may help to minimize the stigmatization caused by the use of medication and which may, in turn, be seen as a sign of moral weakness (Britten, 1996). Thus Montagne has noted that users may feel there is 'a need to take the drug due to symptoms, but only one-half of the normal therapeutic dose is taken so as not to appear bad or morally weak' (1991, p.56). As with opposition 6.4.3.2 this side of the opposition was heard infrequently in the interviews and may well be related to the fact that all the service-user interviewees were currently in contact with psychiatric services and could be regarded as 'compliant' -- the influence of this issue is discussed to some extent in chapter 7.

Sharon: Erm, well I'm on two, I think it's twenty-five milligrams a day Prozac, and then it's one, one milligram of Stelazine <Dave: Right, and> so, <Dave: Right> and that's every day, well/

Dave: /And they seem to help you?

Sharon: I don't take Stelazine that much actually, I only take it if I'm going anywhere or I feel paranoid, 'cos that one of the paranoid symptoms but, I get these depression erm I do feel a lot more relaxed, I think it's the Prozac really that's done that, yeah I used to have problems with my family erm I used to be like really angry with my family, I used to argue and everything, and <pause> I think because I was so highly strung I used to be, have like quite a bad temper, take everything out on them. But now since I've been on this medication they've found that I'm easier to get on with 'cos I'm more laid back than I was <Dave: Uh-huh>. So, I'm a lot happier than I was then <Dave: Right>.

Sharon Harrison (Service-user)

There are a number of features here. Sharon is not against all medication all of the time. Thus, she refuses a simple binary opposition between use and non-use or preference for drug treatment versus dislike of drug treatment. Rather, this is a contextualised account: Shaon makes distinctions between different medications (possibly due to their 'side' effects -- Rogers et al., 1993). In this respect, following Conrad (1987), Sharon might be said to be asserting control through choosing which medication to take and when. She says that she only takes the Stelazine under certain conditions. I noted in 6.4.4.1 how Dr Smith said that Sharon had not 'co-operated'. Here, then, there is a divergence of view. Here, medication is seen as something that a person chooses to take at certain times for certain purposes. Indeed at lines 46-47 Sharon talks of how medication has led to her being more 'laid back'. One effect of this account therefore is to position Sharon as an autonomous rational decision-maker with regard to her medication regime. This enables her to maintain contact with services but not necessarily to accept all the instructions she is given. Such an account warrants a use of medication that might lead professionals, like Dr Smith, to construct her as 'unco-operative', 'inconsistent' or even unknowledgeable about medication. Others, however, constructed her differently. Thus her GP, Dr Howard, noted in neutral terms how Sharon 'goes between one and two Prozac <Dave: Right> and varies it herself' (Dr Howard, line 269). Her account, by contrast, allows her to resolve a dilemma: to not be forced to position herself as a 'mental patient' (which might occur if she were to go on 'maintenance medication') whilst at the same time acknowledging the need for specific help at specific times, for example if she is 'going anywhere' (lines 40-41).

6.4.5 Medication as targeted/ non-targeted and crude

This binary focuses on whether medication is able to target particular symptoms or whether its effects are broader and blunter.

6.4.5.1 Medication as targeted

The first half of the binary views medication as targeted where it might be described as a 'magic bullet' or a 'magic pill' (Montagne, 1991) -- a notion which has some history (Healy, 1991; 1997). Professionals might talk of how the medication's effects are highly focused. The focus in such accounts often relates to symptoms (like beliefs or voice-hearing) and linguistic resources like symptom talk which provide easily identifiable and quantifiable concepts for demonstrating such targeting abilities may be drawn on. Such descriptions, once again, utilise a positive and optimistic discourse of scientific success and progress and may be combined with a view of emotional distress having a biochemical cause. Thus Kirkpatrick & Amador hope to find an 'anatomical site of abnormality related to suspiciousness' (1995, p.496). The targeting abilities of medication are seen as the latest benefits of increasingly sophisticated psychopharmaceutical technology(9). Indeed, this leads to the description of neuroleptics as 'anti-psychotics' with the implication that this is by design rather than by accident(10). Once more in such accounts, symptoms and drugs are given agency whilst other aspects like people and their social context are not. Another effect then is to warrant the ignoring of other aspects by focusing only on symptoms, drugs and drug action.

[Dr James:] support him, you know. This is the first thing. Drugs, nothing else.

Dave: And what effect does that have? Does that take the delusion away or does it make the person less worried about the delusion or?

Dr James: It takes the delusion away in most patients. <Dave: Right> And it's very very effective. That's the best way to do it.

Dr James (Consultant Psychiatrist)

Here Dr James conveys the targeting ability of medication in stating that it takes the delusion away (line 289). This forms part of an extremely assertive pro-medication narrative where he has stated that drugs should be a first response (line 286). This is reinforced by the statement that this is 'very, very effective' (line 290) and that this is the 'best way to do it' (line 290). There is a combination here of a number of rhetorical devices: category entitlement (Dr James's authority and experience as a psychiatrist); symptom-talk (talking only about delusions in judging effects); and the assertion of effectiveness.

The use of a relatively unsophisticated single symptom account within a narrative of effectiveness warrants the ignoring of other aspects of the user's life as indicators of outcome including, for example, the impact of any side effects, or the long term effects of such medications. Thus medication is constructed as relatively unproblematic. However, Dr James provides a qualification 'in most patients' (line 289) which would give grounds for defence if this account was challenged on this. By claiming that the medication 'takes the delusion away' this extract engages in both a simplistic and mystificatory type of accounting, constructing the medication as a technological wonder.

6.4.5.2 Medication as non-targeted and crude

The other side of the opposition encloses accounts which claim that medication does not have specific targeting abilities. In such accounts the effects of medication are not judged solely, if at all, by effects on symptoms but by a wide variety of other outcomes. Common themes here are the effects of medication on users' quality of life. Thus Breggin (1993, 1996) suggests that neuroleptic drugs, far from having specific anti-psychotic effects have the overall impact of a 'chemical lobotomy' and argues that 'despite the claims made for symptom cure, multiple clinical studies document a non-specific emotional flattening or blunting effect' (1996, p.64). Holmes & Newnes quote one user who states how their medication was 'supposed to slow me down and knock me out' (1996, p.14). Barham & Hayward note how users often described being 'knocked out' or 'kept tranquillised' (1995, p.61). They also quote one user who comments on how 'hospitals are understaffed and if they drug everybody they are easier to control' (1995, p.64). Medication may be seen then as a way of managing and controlling difficult users. Indeed Light (1980) gives control as the first reason why drugs were prescribed on the wards he observed. Medication may even be seen at a broader level as a form of social control (Gabe & Lipshitz-Phillips, 1984). Thus common features stress the broad tranquillising effects of neuroleptics and the use of drugs as controlling agents. Such accounts may draw on liberal humanist discourses to claim that such use of drugs damages quality of life and is unethical, warranting change in medication regimes.

Other accounts stress how drugs are not magic bullets which only attack a narrow range of symptoms, rather they also have a wide range of distressing side effects (Lacey, 1991). Indeed, some have claimed that the effects of some drugs may be so broad that they may be marketed as anxiolytics at one moment and as anti-depressants at the next (Healy, 1991; 1997). Rogers et al. (1993) have commented that psychotropic drugs are pharmacologically 'dirty' compounds having a wide range of unwanted effects. Such claims warrant the use of medication but in a careful and controlled manner. Barham & Hayward (1995) note how some users described their medication as useful not simply because of symptom relief but because they helped them to feel relaxed or keep them 'out of trouble' (p.61). Healy (1997) notes that a major effect of neuroleptics is a feeling of detachment, a 'who cares' feeling. Such accounts emphasise the broad effects of neuroleptics, constructing them as beneficial and relaxing. This kind of narrative may then be used not only to signify the dangers of neuroleptics, but also the need for moderate and balanced use and their positive benefits. Although there were a number of extracts noting the negative impact of the non-targeted effects of medication in the interviews, the following extract raises the issue of how this can be made to signify medication as positive.

John: Probably, yeah. Started feeling alright once I was on the medication.

Dave: <Pause> What was the main effect that the medication had on you, to make you feel better?

John: Just made me calm down, that's all.

Dave: Made you less worried about things?

John: Yeah, less anxious, no panic attacks.

Dave: <Pause> Did it make any difference to the thoughts themselves? Were the thoughts still there but just bothering you less or...?

John: Well the thoughts were there but they just didn't bother me.

Dave: So it's not that they actually make the thoughts go away as such?

John: To some extent they do, yeah. But if it comes across my mind I just ignore it. <Dave: Right. And you/> /I have no feelings towards it whatsoever.

John Stewart (Service-user)

John highlights the broad effects of the medication: it has made him 'calm down' (line 192); made him 'less anxious' with 'no panic attacks' (line 194); to the extent that thoughts just don't bother him (line 197). Indeed, the more specific my questions become, with questions apparently drawn from a highly specific symptom repertoire, the more variable John's account becomes. In this account broad effects are constructed as positive with one effect being to warrant continued use of the medication as requested by professionals.

6.4.6 Accounting for medication working/accounting for medication not working

This opposition concerns the way in which the apparent success or failure of medication is accounted for.

6.4.6.1 Accounting for medication working

The first half of the opposition is concerned with accounts which explain both the reasons for thinking medication is working and mechanisms of action. Gabbay (1982) and Potter (1982) have stressed the need to look at the discursive strategies used in explaining mechanisms of therapies and the discursive effects of those explanations. Here the focus is on the rhetorical resources drawn on to account for the working of medication and, indeed the resources drawn on to account for medication having been seen to have worked (ie the effects of the medication) -- where medication is seen as an explanatory resource.

There was a variety of descriptions of medication effects in the interviews. Medication: 'gets rid' of delusions (Mike Sullivan, line 449) but leaves them in the back of the mind as a memory (Mike Sullivan, line 451); stops belief in delusions (Mike Sullivan, line 455); makes delusions 'go' (Dr Smith, lines 254-255) or 'clear off' (Dr Williams, line 143); or makes them go away 'to some extent' (John Stewart, line 199); makes delusions 'less apparent' (Terry Reid, line 240); or is said to have 'dulled' them (Terry Reid, line 241); makes users less 'bothered' or 'troubled' by delusions (Terry Reid, line 242; John Stewart, line 197); 'less fearful' (Edward Jackson, line 211); and 'less anxious' (Edward Jackson, line 220; John Stewart, line 194)); and helps users 'calm down' (John Stewart, line 192).

There was thus a wide variety of effects culturally available to professionals and users in accounting for whether medication had worked. Such variety means that speakers have a range of accounts to draw from and allows a great deal of flexibility in deciding whether or not a medication can be said to have worked. Thus in the extract in 6.4.1.1 we saw how an apparent lack of effect of medication on 'positive' symptoms (like delusions and hallucinations) can be counteracted by an apparent effect of medication on other symptoms seen as 'negative' (like withdrawal or motivation) or on other issues like a user's quality of life. Accounts of whether and why medication has worked linked with those in 6.4.1.1 in that many of them drew on a narrative of optimism, empiricism and obviousness. Such a flexibility in judging the effectiveness of medication may warrant its continued use in a wide variety of different situations and may be used to defend medication from a variety of challenges. Moreover, many accounts focus on only the medication as a causative agent, warranting the minimising of other factors. Johnstone (1993c) argues that a common rhetorical strategy is to attribute all improvement to medical interventions alone (and conversely ascribing progress in counselling to other factors, like 'remission'). Indeed even though there may be agreement that the causative agent is medication there may be disagreement between users and professionals and amongst professionals about which medication has led to which effects.

In the extract in 6.4.5.2 the account of medication working appears to be straightforward: the medication made John calm down (line 192). The account is characterised by symptom-talk, focusing on anxiety in particular. John does not give an account of how he links this change to his medication. A narrative account leads to the implication that progress is due to the medication. In line 189 John says that he started feeling alright 'once I was on the medication'. The link between the outcome and perceived cause is constructed as a chronological one but is also ambiguous. One effect of this is that if the account was challenged John could locate the cause elsewhere. In mentioning only the medication as a potential causal agent, the implication is that the medication is responsible and this is reinforced by the implication of obviousness (see 6.4.2.2). Thus medication here is seen as a default explanation.

One effect of this account is to warrant: the construction of medication as the cause of John's recovery; his continued use of medication; a lack of curiosity both about the mechanism of this action and the relative importance of broad versus specific effects; and thus a faith in medication. This is reinforced by the way the account grants causative agency to the medication despite the acknowledgemet of his own role in dealing with the symptoms (line 199-200).

6.4.6.2 Accounting for medication not working

The second half of the binary opposition includes accounts where the apparent failure of medication was accounted for. A number of potential sources of failure are noted in the literature. Thus Lacey notes that mental health workers are 'fallible professionals whose tools are imperfect drugs' (1991, p.13). However, in the interviews, few professionals appeared to place the locus of responsibility on themselves or on the idea of medication per se. Rather there were a variety of complex narratives drawn on. Because of the wide variety of rhetorical strategies it seemed important to examine these in more depth.

There is relatively little work on how the business of accounting for the apparent failure of medication is done(11). This mirrors the general lack of discussion of the failure of other professional interventions (see Kaffman, 1987 and Spellman & Harper, 1996 for a discussion of failure in family therapy). Failure talk has two features: first, it may be dealt with through a re-definition of terms, for example through the use of new diagnoses (like 'treatment-resistant symptoms') or of outcome; second, the locus of failure may be placed at some other point than the specific medication used (or of medication in general) -- like the dose, diagnosis or even the user. Johnstone has suggested that failure may even be denied by disqualifying the counter-evidence thus 'if ECT appears to "work", then it will be used again. If it doesn't "work", then it will still be used again in case it "works" next time' (1993c, p.31).

Light (1980) has described a range of responses by trainee psychiatrists that pass as 'reasonable' in conflictual discussions with nursing and other staff. The first response consisted of a psychiatric explanation of an event. The second involved continuing the report a nurse had started thereby demonstrating the doctor's knowledge of the case. The third response involved putting the present situation into an historical perspective (usually to demonstrate things were not as bad as others thought). Fourthly, the doctor explained the behaviour as a stage the patient would pass through. A fifth response consisted of a decision to medicate or change medication or to restrict the patient in some way. A sixth response was to explain why a particular policy was to be maintained. A seventh consisted of the doctor remaining silent or avoiding eye contact if unsure what to say. An eighth kind of reply was to comfort or reassure staff about a patient's behaviour. The final response he described as a form of evasion, diverting the thrust of the nurse's report as a 'decoy' so as to negotiate on the psychiatrist's territory rather than the nurse's. A number of these responses were seen as ways of managing ambiguity and uncertainty and he suggested a number of strategies used to manage this, for example, by deferring to clinical experience. Together with some sustained critique of such rhetorical strategies (Johnstone, 1993c) there have also been some humorous attempts, with Lowson (1994) accusing many professionals of suffering from 'professional thought disorder' symptoms of which include 'an assumption of intellectual or moral correctness or superiority, frequently held in spite of the presence of major contra-indications' (1994, p.29). Judging the effectiveness of medication is a complex and uncertain business but I would argue that it is a rare event for many psychiatrists to be open about this with users, relatives and workers and to explain to them about why this might be the case (though see Healy, 1997 and Thomas, 1997 for exceptions to this). For many psychiatrists (and users) who use a simplistic biological model the failure of medication creates a problem(12). If medication is supposed to treat an illness called schizophrenia or is supposed to be targeted at certain symptoms which are then 'removed', how is the fact that often use of medication is met with what is seen as little change in symptoms to be accounted for? My reading of the interviews suggested there was a wide range of possible ways of accounting for failure which were culturally available both to professionals and users.

6.4.6.2.1 'The patient is a non-responder'

Dave: Uh-huh. <Pause> Have the doctors tried to explain what the voices are?

Geoff: They just said it's very common <pause> erm <pause> it's not uncommon <pause> and there's not a lot they can do 'cos I don't respond very well to medication. <Dave: Right> Something that I've got to learn to adjust to and live with myself, you know.

Geoff Nelson (Service-user)

Geoff describes how he has been told by professionals that he doesn't respond well to medication. We noted earlier how the literature abounds with phrases like 'drug non-responders'. Such talk does the work of locating the source of the problem in the user rather than in the medication or elsewhere in the user's context. This has three main effects: first it removes blame and responsibility from the professionals; secondly it removes blame and responsibility from the medication; thirdly in placing responsibility on Geoff it requires him to 'adjust to and live' with it himself (lines 100-101).

6.4.6.2.2 'There are obviously odd exceptions': rhetorical innoculation and qualification

Dave: Right, right <pause> and in your experience how does that affect the, the, the paranoid delusion? What, what does it do to it?

Dr Smith: Well, yes, it er, under treatment very often the, the paranoid delusion goes. There are obviously odd exceptions here and there <Dave: Right> no matter what you do er the ideas still persist, probably the intensity goes down a bit but erm they may still er have to carry on with the idea but dont act on it erm but thats not that common.

Dr Smith (Consultant Psychiatrist)

Dr Smith describes how under treatment (taken here to mean drug treatment) the delusions go (line 255). He then goes on to note that there are 'obviously odd exceptions' (line 255). There is a continual dilemmatic movement in the text between a narrative that medication works and a more qualified narrative: the delusion goes but there are exceptions (line 255); intensity goes down (line 256) but the idea carries on (line 257) and so on. This works to innoculate any future challenge from either side. Thus if the statement that delusions go under treatment were to be challenged it could be pointed out that Dr Smith had only said 'very often' (line 254) and that he had noted the existence of exceptions.

6.4.6.2.3 'We don't know'

Dave: <Pause> Right, right. And with, with Paul er Dench what, erm <pause> erm what, what seems to have helped him? What, what, what treatment has he received?

Dr Smith: I dont think any treatment has really <pause> helped him very much <Dave: Right> er he is at the moment on a depot. He says <pause> er it is helping him to some extent <pause> er I dont think any treatment has, has done him any good so far.

Dave: Why, why do you think that is?

Dr Smith: He still holds same belief. Although he will tell you that he can argue with you and, and see the logic of it but the next moment er he will tell you he is still terrified.

Dave: Right, right. So why is it that the treatment doesnt work?

Dr Smith: <Pause> Cant answer that <laughs> erm <pause> erm he had different types of er erm anti-psychotics er at some point he probably complained of depression er the GP has given him an anti-depressant erm <pause> only partially the depot has been helpful.

Dr Smith (Consultant Psychiatrist)

Here, I ask Dr Smith at line 495 why it is that the treatment (again, assumed to be drug treatment) has not done any good. Dr Smith explains what makes him say that but does not directly answer the question. This exchange could be seen as an example of systematic vagueness (Edwards & Potter, 1992). When I re-phrase the question Dr Smith pauses and then laughs as he answers. This may suggest that at the earlier turn Dr Smith was attempting to (not necessarily intentionally) evade answering the question. This effect is increased when Dr Smith then goes on to note that he cannot explain why the medication has not worked. This is similar to the kind of rhetorical evasion noted by Light (1980). However, Light did not note whether staff ever challenged this. Given the status of professionals like psychiatrists it may well be unusual that such a move would be challenged. The fact that I challenged this move because of my power as interviewer also, then, needs to be part of this analysis. There are other features to this account such as the claim that the depot has been partially succesful (line 503) despite all the evidence of the account and the blaming of the GP (line 502), user (line 501) and previous medication (line 502) for the medication's failure.

6.4.6.2.4 Because the patient is chronic

A narrative which was present in many of the transcripts related to the chronicity of conditions. The chronicity narrative emphasised the permanence and severity of symptoms and illnesses and usually involved assumptions about the biological origin of problems within the person. This was a useful resource for locating agency within a problem (like a symptom) which was essentialised and abstracted from the person and the context of their life and relationships yet projected within the person. This had a number of effects. One was that it removed explanatory responsibility from professionals: if a symptom is chronic then we do not need to look for causes or solutions in the current environment. A second effect was that it removed responsibility from professionals for their failure. However, it did this by placing responsibility on the form and nature of the constructed problem. Since this problem was seen as lying within the user there was some discursive variability about whether responsibility and agency lay with the problem or the person. Holmes & Newnes (1996) have noted how users may be blamed for drug failures. Rowe (1995) has noted some interesting rhetorical moves:

Ever since the late Fifties, psychiatric treatment for depression has been with drugs and ECT. If a substantial proportion of those people so treated either don't get better or have recurrent bouts of depression it is not, so the psychiatrists say, because the drugs and ECT don't work but because they have a chronic illness. Psychiatrists are experts in blaming the victim.

Rowe (1995, p. 10)

The next extract is a good example of a chronicity narrative combined with symptom-talk:

Dave: D-, do you find that on the whole delusions tend to respond to the medication that, that's prescribed or, or does it tend to be more a kind of symptom that's quite difficult to, to deal with with medication?

Dr Williams: Depends on the condition. In depression and, you know, with bi-polar affective <pause> er they recover completely <Dave: Uh-huh> after an episode <Dave: Uh-huh> and everything clears off <Dave: Right> I mean they're like, in between episodes they're like normal people. <Dave: Uh-huh> But, you get in schizophrenia, there's gradual deterioration and at times the s-, symptoms persist, you know, they've got residual symptoms. <Dave: Uh-huh> So the delusions or hallucinations kind of ease off but they persist in between episodes as well <Dave: Right> and then er er when patient have the next episode, after that it will be even worse you know. Gradually patient deteriorates and some of the residual symptoms persist. So in, in that case, with chronic er schizophrenia or this kind of schizophrenia where there's no, no complete remission <Dave: Uh-huh> the pers-, the delusions can persist and they are difficult to treat there.

Dr Williams (SHO in Psychiatry)

Dr Williams uses the notion of 'residual symptoms'. These are usually defined as symptoms remaining after drug treatment and are also known as 'treatment-resistant' or 'drug-resistant' symptoms. In this account, however, the term is used as a new diagnosis: the symptoms remain because they are 'residual symptoms' (lines 253-254). The account works to obscure the tautologous nature of 'residual symptoms'. This is a common effect of much scientific and medical accounting: a thing is given a descriptor and this descriptor may be used as if it is an explanation(13). What are some of the effects achieved by the use of this narrative? First, the problem is located within the patient (the 'patient deteriorates' line 253) and the symptoms are given agency. Secondly, in implying the permanence of the symptoms, responsibility for their cause or continued existence is removed from professionals. This is also achieved through the construction of a category ('residual symptoms') which then normalises this occurrence. Thirdly it is implied that such symptoms are hard to treat but the tautology is again obscured because of an objectifying empiricist account which constructs symptoms and chronicity as real and 'out there'.

6.4.6.2.5 Because the patient is on too low a dose

Within the interviews there was a particular class of explanations for drug failure. These revolved around a focus solely on the chemical substance itself and how it was used: for example the dosage or the type of drug used. These accounts achieved a very powerful rhetorical effect. They constructed a kind of obviousness and relied on common-sense constructions of drug actions: for example the notion that if a drug does not work it may be because not enough is given and that potency will increase as the dose increases. Thus Barnes et al. (1996) list 'inadequate dosage' as one possible reason for an 'inadequate response to Clozapine'. Some empirical work suggests this is not the case with many neuroleptics -- see Bollini et al., 1994; Healy, 1997). Light notes that 'most patients considered first medication or increased dosages as a clear sign of their own deterioration' (1980, p.151)(14). It is clear from the literature that there is a great deal of variation in prescribing practices (Chaplin & McGuigan, 1996; Davis, 1994; Rogers et al., 1993) and so the focus on drug dosage, quantity and mixture is a widely culturally available linguistic resource.

Several features are present in most of these accounts: they involve empiricist accounting through the use of quantification (as in discussing dosage) and symptom-talk; decisions about medication tend to focus only on the drug and symptoms to the exclusion of other factors which either might influence the decision or which might influence whether the drug was said to have worked -- although other factors are mentioned these are not given prominence. Such an account has a number of effects: a focus on symptoms means that the effects of a drug on other aspects of the user's life are not highlighted; a focus on the drug means that other influences (eg changes in the user's life circumstances, relationships or their own efforts to change) are not highlighted. An account is created where changes in symptoms are therefore only related to medication.

Edward: I think he came with 500 a month. <Dave: Right> And that seemed, that was what he wanted when he got back here but erm the Consultant wouldn't prescribe it. I think the feeling was prescribe low and work your way up so <pause> I think it was erm quite a few weeks before he [inaudible] the 500. So, of course, Geoff was sort of badgering <Dave: Uh-huh> and asking for [inaudible sentence]. But I think when his medication was at the top it did, what it, what it was when he was in [name of town], it did seem to have quite a, quite an effect sort of getting towards the time that he actually disappeared and went back to [name of town]. That, that wasn't having the impact it could have or should have had.

Edward Jackson (CPN)

This extract consists of a claim that the dosage of the medication was too low. It conveys this through employment of a narrative of about '500 a month' being the right dosage for Geoff. Several sources are used to back up this claim: first that it was the dosage which Geoff came to the area with (line 206); second that it was what Geoff wanted (line 207), indeed he 'badgered' for an increase (line 210); third it was only when the dose reached this level that there was an 'effect' (line 212) and an 'impact' (line 214). These latter terms refer to a change in symptoms and utilise a version of symptom-talk. However, throughout this account which is ostensibly about the 'real' effects of a drug there are a variety of interests and conflicts. For example, although Geoff apparently made his wishes clear, the Consultant 'wouldn't prescribe it' (lines 207-208). Edward here gives the account of a disinterested observer but gives no account of his own position. Although these issues are noted however, they are seen as a simple background to the real effects of the drug when it eventually began to have an impact. This disclosure of conflicts and other influences on drug decision-making then paradoxically strengthens a drug-focused account. Another effect of such an account might be to deflect a challenge that either the wrong drug was being used or that medication per se was not useful by claiming that if the dosage was increased, therefore increasing its potency, then its effect would be achieved.

6.4.6.2.6 Because the patient is on too high a dose

There is a further example of a drug-focused account however, in the claim that dosage is too high. Once again, research indicates that given current prescribing practices this is a widely available resource -- so much so that authors talk of 'megadosing' (Rogers et al., 1993). Whilst the previous narrative tended to be employed as a justification for increasing drug dosage to have an impact on symptoms, in contrast this account tended to warrant a decrease in dosage by placing a focus both on effects of the drug (constructed as negative, like side effects) and on aspects other than solely symptoms, using a liberal humanistic quality of life narrative to achieve this effect.

Mike: Chlorpromazine was the first one, I was on 400mg of that initially. I was sort of like a zombie er gradually got it reduced to about 100mg. And I forget the other one that he put me on after that, it was an injection <pause> <Dave: People/> Depixol's been better.

Mike Sullivan (Service-user)

In this account, Mike Sullivan describes how the high dose made him 'sort of like a zombie' (line 462). This is a powerful rhetorical device which conveys a strong image of how the dose was affecting him. Here again though, the account works by deploying popular notions of psychiatric intervention and medication. There is also a focus on the drug alone here separated from the context of Mike's life and other possible influences on his state (and a psychiatrist might also claim that Mike's feelings were influenced by his illness or that his body was adjusting to the dose). There is no mention here of symptoms and we might expect that symptom-talk would not be employed here (or only if the increased dose had not been felt to have had an impact on symptoms). There is also use of a liberal humanist narrative which would claim that the drug was restricting his quality of life.

6.4.6.2.7 Because the patient is on the wrong drug

Light (1980) notes how previous drug history is a basis for many prescription decisions. In this account, it was claimed that the reason for medication failure was because the wrong medication was being used. This kind of account was another one which solely focused on the drug and was used in a variety of contexts. It might be used to claim that a different anti-psychotic drug be used. Or it could be used to argue that a different family of drugs might be useful (eg to move from anti-psychotics to anti-depressants). Once again, one effect was to deflect criticism of pharmacotherapy by claiming that it was simply a case of finding the right drug -- it could also then draw on accounts relating to diagnosis. By focusing on the drug as a lone active agent, other aspects were ignored or minimised.

Dr Howard: She really didn't stay on the treatment. Well I didn't know whether it was the right thing to [inaudible word] because she did respond to anti-depressants when I started her on them. <Dave: Uh-huh> So she came back to me and said "I don't really want to be under them any more" <Dave: Right> so I started her on Prozac which I wasn't sure was the right one for her. <Dave: Right> But we'd also tried a few other ones before that. Prozac was the best one for her. That's really the only one that's helped her. <Dave: Right, right> But erm she goes between one and two Prozac. <Dave: Right> And varies it herself. But she thinks that's the best one for her. And there's no side-effects you see.

Dr Howard (GP)

Dr Howard does not account for her medication decision-making solely on drug grounds but notes the influence of a variety of other factors: Sharon Harrison's compliance; her previous response to anti-depressants; her wishes; and the lack of apparent side-effects. Thus deciding on correct medication appeared to be a flexible, pragmatic and situated process. One effect of this kind of account is, once again, because of its flexibility, to be open to change depending on circumstances. Moreover, because the narrative is not tied to a particular theoretical model and is more pragmatic its flexibility is increased. Here again though, although there are other influences noted, paradoxically, the account still conveys that this drug worked by affecting the symptoms.

Here there is less use of a symptom-focused narrative and little empiricist accounting. Indeed Dr Howard continually stresses how unsure she was about the correct medication: 'I didn't know' (line 262); 'I wasn't sure' (line 266). Rather, the impression of collaboration is given. Such an account positions her as a liberal professional trying to empower her patient and warrants a movement between medications which might not have been justifiable using only an empiricist account -- although this might still have been possible through the use of a strategy drawing on notions of experimental trial and error.

6.4.6.2.8 Because the patient is on too many different kinds of drugs

This was a further variant of the wrong dose/wrong drug narrative. Here again the focus was mainly on the chemical action of the drug and its effects on symptoms. Newnes & Holmes' interviewee, for example, states 'at the beginning of the illness they try out so many different drugs. You're like a guinea pig' (1996, p.3). This kind of account could be used in this way as a strategy for arguing against polypharmacy -- the use of 'cocktails' of drugs or alternatively, as a way of arguing against a particular combination of drugs. Once again, given the importance of drug-history in medication decisions (Light, 1980) and the extent of polypharmacy (Chaplin & McGuigan, 1996; Rogers et al., 1993) we know this is a widely culturally available account. Barnes et al. (1996) list 'comorbid drug use' as another reason for inadequate response to Clozapine.

Geoff: Well I have my tablets during the day as well, you see <Dave: Right, I see> <pause> the Largactil and the injection, Clopixol, does help <Dave: Uh-huh> together but on their own they're no good. See I've been on drugs now for eight years <Dave: Uh-huh> and it's mainly been Largactil but my injections they've changed em and changed em and changed em. <Dave: Right> They keep trying me on different things, you know. But this drug, this Clopixol <pause> has bad side effects, I believe, but it's like a wonder drug <Dave: Uh-huh> with Largactil it numbs the voices, you can't hear em and it makes you feel relaxed <Dave: Uh-huh> <pause> but lately, I've just had an injection I mean I'll contradict myself now 'cos it didn't work but only because I wasn't taking the Largactil as well. I've recently been to see my GP and she prescribed my Largactil and it seems a lot better now.

Geoff Nelson (Service-user)

Geoff talks only about his medication and symptoms. The Largactil and Clopixol when taken in combination are seen as helpful (line 168) whereas when taken on their own (line 168) or in other combinations (line 170) they are not seen as helpful. This account conveys a picture where a wide variety of drugs are tested in combination. Even where the drugs are not seen as working, this is constructed as a difficulty with the medication (lines 175-176). Such an account has a number of effects: it focuses attention away from other issues which might have an influence on whether Geoff is better (such as social context) and it also focuses attention away from other issues in Geoff's life. Side effects, for example, are minimised when Geoff states that although he believes the effects to be 'bad' (line 172) he feels that Clopixol is a 'wonder drug' (line 172). As Sherlock & Kielich (1991) have described: users may increasingly come to account for their lives in terms of medication, dosage and symptoms.

6.4.5.2.9 Because the patient has not been compliant with their medication regime

Another narrative used to account for medication not appearing to work was for professionals to complain that users had not been compliant by taking their medication as prescribed. This is a widely available explanation. Thus Kane (1996b) and Barnes et al. (1996) see poor or non-compliance as an explanation for inadequate response to neuroleptics. Moreover they note that 'adverse effects' may occur which may then lead to poor compliance. Barnes et al. (1996) note how 'lack of insight' may lead to non-compliance. Kane (1996b) also notes that 'maintenance failures' may be to blame -- it is unclear how this is different from non-compliance.

In the interviews, there was often some disagreement about compliance, with professionals often suspecting this rather than having proof if the issue concerned oral rather than depot injection (in much the same way as they might suspect a user of having used illicit drugs). If users had previously not followed medication advice this account was more likely to be used. Once again, such an account explained the apparent failure of medication by placing the locus of blame elsewhere (on the user). For example, in this discussion of Sharon Harrison:

her. Erm and she is convinced, you cant really shift it. Erm she hasnt sort of co-operated er with treatment erm er so her ideas are still there.

Dave: Right, what, she stopped her medication or? Right/

Dr Smith: /Yes, never fully co-operated with treatment. <Dave: Right> Erm

Dr Smith (Consultant Psychiatrist)

We discussed the features and effects of the compliance narrative in 6.4.4 and here it can be seen that 'co-operation' is constructed as a binary either/or and is seen as internal to Sharon, thus removing responsibility from professionals and the medication.

6.4.6.2.10 Because the patient has been wrongly diagnosed

This narrative is related to 6.4.6.2.8 in that if a user was felt to be wrongly diagnosed they could be said to be on the wrong medication. One effect of this is to explain a drug's failure by claiming that it was not the right medication for the problem. However, given the flexibility of the symptom repertoire and the existence of similar symptoms in different diagnostic categories (see chapter 3) it seemed possible for a variety of diagnoses (each entailing different medications) to be entertained. The notion of a correct diagnosis also implies that there are exact drug and diagnosis matches whereas, in practice, a number of different drugs may be used for the same diagnosis whilst the same drug may be used for quite different diagnoses (Healy, 1997).

Dr Howard: /No I just wanted to clear the diagnosis up. I didn't think she'd been diagnosed properly and wasn't quite sure whether she did come into the schizophrenia pattern or whether it was, but, but I thought with the psychotic tendency <Dave: Right> that she may need some, like, either injection treatment or major tranquillisers as well as anti-depressants. But really I shouldn't really have instigated that without getting a more concrete diagnosis. <Dave: Right, right> Erm and she was reluctant to take any other medication anyway. <Dave: Right> Erm.

Dr Howard (GP)

Dr Howard explains that she felt that Sharon Harrison might need a depot injection or (presumably oral) major tranquillisers because of a 'psychotic tendency'. Here a dimensional model of psychosis is used as opposed to a categorical one. One effect of this is to warrant the use of two different diagnoses (schizophrenia and depression) and two different medications (anti-psychotics and anti-depressants). Although categorical, most diagnostic schedules allow the possibility that a person may have more than one diagnosis (this is termed co-morbidity and such users may be said to have a dual diagnosis(15)), the use of a dimensional model makes this easier. The notion of Sharon not having been diagnosed properly draws on a realist and static view of diagnosis. Although Dr Howard focuses mainly on diagnosis and medication here, however, it seems that other factors are influencing the decision such as Sharon's reluctance to take other medications (line 227).

6.4.6.2.11 Because some of the patient's problems are due to manipulative behaviour

In 6.4.6.2.5, we noted how a rhetoric of chronicity placed responsibility for drugs not working on the illness entity within the user. The issue of compliance placed responsibility more firmly with the user. Another kind of account where this occured was when medication was seen as not working because the symptoms were not caused by a real disease but instead were caused by the user.

Dave: What do you think has helped him? Er do you feel that you've helped him or treated him in any way or? Is anything appeared to have helped/

Dr James: /Really we are not helping very much apart from giving the medication otherwise he will become, he does become very ill. He has become very, he gets the medication regularly which is good/ <Dave: /Right, right, right> But the rest is manipulative behaviour.

Dr James (Consultant Psychiatrist)

In this account responsibility for the medication not producing any change is placed on Geoff Nelson's 'manipulative behaviour' (line 182). This is a narrative that has a number of effects. First it maintains a defence that the medication does treat a disease and that the only reason there is no change is that a proportion of the symptoms are not related to the disease. Second through the claim that without the medication Geoff will become ill (line 180) the implication is that the medication is, if not making changes, at least preventing the situation worsening. This is a very powerful device for maintaining compliance -- Day et al. (1996) noted how fear of relapse was a motivating factor in adhering to prescription regimes. Third, through the suggestion that the symptoms which are not changed by the medication -- what might be termed drug-resistant symptoms -- are 'manipulative behaviour' (line 182), a tautological argument is drawn on. In other words, drugs treat the disease and therefore symptoms of the disease will go. If a drug is given and symptoms remain they cannot be related to the disease. Such an account once again defends the apparent limitations of drug success, warrants the continued use of medication, and places responsibility for other symptoms elsewhere. The use of the term 'behavioural' to identify a realm not affected by drugs draws on a notion of organic disease states which are seen as separate from behaviour. However the use here of 'manipulative' also introduces a moral element of blame in that there is an implication that Geoff is, to some extent, deliberately engaging in such behaviour to manipulate others. It might be argued that the manipulation is judged by effects (eg reinforcement) rather than by intention but the ambiguity serves to question Geoff's intentions. This even more firmly places responsibility and blame on Geoff and not on Dr James or the medication. In these respects, formulations of manipulative behaviour have similar discursive effects to those of personality disorder (ie in providing a rationale for the limitations of treatment).

6.4.6.2.12 Because it is just that the drugs are not working 'fully'

A final narrative drawn on in accounting for drugs not working was to claim that they were working but were just not working to their full potential. Such accounts both explained the limitations of the drug and justified its continued administration.

Dave: Right. And you referred to psychiatry then? <Dr Howard: Yeah> Was that, what was the thinking behind that? Was that/

Dr Howard: Because she wasn't responding fully to the anti-depressants <Dave: Right> and was reluctant to go on any other medication <Dave: Right> and the family couldn't cope. <Dave: Right> I thought "Well she can't carry on like this". <Dave: Right> Erm so she did agree reluctantly to see the psychiatrist again.

Dr Howard (GP)

In this extract, it is noted that Sharon was not 'responding fully' (line 213) to the medication. There is an ambiguity here about why she was not responding fully: because of the medication; because the wrong diagnosis had been made; or because of some factor in Sharon? Also some responsibility is placed on Sharon because of her reluctance (line 214). Such an account maintains the need for continued use of the medication but accounts for their limitation by appealing to the need for other medication. Since it is Sharon's reluctance that prevents this happening, some responsibility is thus placed on her.

It was almost as if there existed a fluid network or a concourse of possible explanations. Rather than get into a debate about whether these are 'true' in general or in specific cases I want to argue here that certain strategies are drawn on to construct these explanations and that, when employed, they have certain effects. For example, they warrant certain actions (eg the increasing of the dose of a drug in the case of 6.4.6.2.5) and they lead to the construction of certain kinds of identities (eg of the patient as 'resistant' in the case of 6.4.6.2.9). In the British Journal of Psychiatry supplement noted earlier the suggested alternative interventions in the case of inadequate response to neuroleptics were all pharmacological. Morrison (1996) makes the following suggestions which would be warranted by the use of these explanations including: increasing or reducing the medication dose; using more than one drug; and using 'adjunctive treatments' (other drugs and ECT).

6.4.7 Talking treatments are unnecessary compared to medication/Talking treatments are important

In both 6.4.2.2 and 6.4.3.2 we have described an account about talking treatments like psychotherapy and counselling. Recent years have seen cognitive-behaviour therapies, for example, encroaching into treatment of psychosis, previously regarded as requiring pharmacological approaches only (Parker et al., 1995). This chronology has not been the same everywhere, however, and Light (1980) has described how in his place of study, psychoanalytic therapy was seen as the best way of addressing problems with drugs only providing symptom cure but that this had changed over time so that increasingly drugs were seen as the essence of therapy whilst psychotherapy was relegated to a supportive role. This opposition, then, is organised around the importance given to talking treatments in addressing paranoid delusions.

6.4.7.1 Talking treatments are unnecessary compared to medication

In this half of the binary, speakers suggested that talking treatments were unnecessary compared with medication. Such accounts drew on elements of 6.4.1.1 and 6.4.2.1 in that medication was taken for granted as an obvious treatment that worked. Here talking therapies were either completely ignored or mentioned but minimised in some way. Thus Johnstone (1993c) has described how psychiatrists ignored or discounted non-medical approaches or attributed all improvements to medical interventions alone rather than psychotherapy. This narrative had a number of effects. First it constructed the prime treatment as pharmacological and, as we saw in the extract in 6.4.2.1, this often necessitated the construction of the problem as having a biochemical cause which was then used to warrant a biochemical solution. Second, in so doing it positioned psychiatry as the prime discipline in treating such problems. Third, as Breggin (1996) has noted, such a focus minimises psycho-social interventions like psychotherapy and small volunteer-run residential settings by constructing recipients as defective objects rather than people struggling with emotional and social problems.

Dave: Right. <Dr James: Yeah> Okay, erm. Could I just ask you one, one last question. What, what in general do you feel helps people with paranoid delusions? What kind of form of treatment is best for them?

Dr James: The best treatment is listen to them, don't argue with them about his delusions. Don't er, sort of er, you know, you just listen. The other thing is, really, I think erm er the best value for money or, you know, to is to give neuroleptic. <Dave: Uh-huh> If you had more time, you'd talk to them more, support him, you know. This is the first thing. Drugs, nothing else.

Dr James (Consultant Psychiatrist)

Here it is stated that listening and not arguing is the best treatment (line 282) and this partly achieves its effect because of category entitlement. The account draws on a narrative of cost-effectiveness to state that giving neuroleptics is the 'best value for money' (line 284). A second element in the narrative is related to time: 'if you had more time, you'd talk to them more' (line 285). This builds up, through a naarative form of accounting to the inevitable conclusion that 'drugs, nothing else' is 'the first thing' (line 286). Here then, medication has primacy in terms of being the first, the best value for money and most time-efficient treatment. Also, medication is constructed as being the active agent here with the user constructed as passive, almost an obstacle.

The account contains some variability because Dr James has also said that the 'best' treatment is listening. Thus medication is not judged as the best treatment, merely the most time and cost-effective one. Such an account creates an effect of pragmatism, thus positioning Dr James as reasonable and this account, paradoxically, is thus able to respond to challenges. If someone were to say Dr James was anti-talking treatments, one could point to the comment that he sees them as the best treatments if only there was time and money. Indeed since I, as a clinical psychologist, was conducting this interview Dr James might be expected to give an account that was at least sympathetic to talking treatments. However, the effect here is to warrant 'pharmacotherapy' as opposed to psychotherapy and thus to warrant the primacy of psychiatry as opposed to other disciplines in the management of psychosis.

6.4.7.2 Talking treatments are important

In this half of the opposition, it was argued that talking treatments were important. These accounts had a number of features similar to 6.4.2.2 -- the assertion that medication was not the only thing users needed from services (Barham & Hayward, 1995) and that they needed some form of relationship and counselling. It also shared some features with 6.4.3.2 in that use of such an account often reflected a lack of faith in medication (or at least in the use of medication alone) and an ideology that problems were primarily social and political, not biochemical. Here agency was given to the professional-user relationship and tended to construct users as active agents. However, the uses of this kind of account were various and could be used to justify: the sole use of talking treatments and the abandonment of medication (Breggin, 1996); the use of both together but with psychotherapy seen as having primacy (Light, 1980, p.263); a use of both (Manschreck, 1996); and a use of both but with medication seen as having primacy (Light, 1980, p.130).

Paul: /No <Dave: No, right>. <Pause> There's, there's one thing missing in psychiatric service these days and you're psychologist so <pause> [inaudible] like Julie does with me and listens to me <Dave: Right>. That's a good way of sorting <pause> paranoid feelings out and what's, what's to be paranoid and what you shouldn't be, <Dave: Right> whatever.

Dave: Right. That's the kind of thing Julie's been doing with you, is it?

Paul: Yeah, not just pumping you with drugs <Dave: Right> and <pause> I hate, I hate being pumped with drugs.

Paul Dench (Service-user)

Paul describes how listening is one thing 'that's missing' (line 374) from the psychiatric service. This is contrasted with 'pumping you with drugs' (line 380). Through the use of 'listening' rather than talking, professional involvement is constructed as passive and benign whilst Paul is constructed as more active and empowered since listening is a 'good way of sorting ... paranoid feelings out' (lines 376-377). In contrast, the notion of pumping Paul with drugs constructs him as a passive object and professional involvement as active, punitive and perhaps more interested in social control. One effect of this is to warrant both an involvement with psychotherapy and a reduction in dosage or abandonment of medication. Once again, however, there is likely to be an influence here since Paul was aware that, as a psychologist I was more likely to be sympathetic to talking treatments.

6.5 CONSTRUCTING THE OBJECTS AND SUBJECTS OF MEDICATION TALK

The previous sections have focused both on the binary oppositions implicit in much of the talk about medication and on the rhetorical strategies used in these oppositions and in dealing with various challenges (eg on the validity and effectiveness of medication). However, theorists like Parker (1994) have argued for the need to move beyond a description of these features of talk to look at what kinds of subjects and objects are constructed in this talk and we have drawn on the notion of paranoia as a discursive complex throughout Parts I and II. From such a standpoint the focus is not so much on spotting rhetorical devices but on describing the kinds of object and subjectivity constructed in talk. In my second layer of analysis I wish to be sensitive to the ways in which talk about medication constructs a number of subject positions for users and professionals whilst constructing a range of objects. The meanings of interventions like medication are contested and, like diagnoses, they delineate certain subjectivities. The questions the analyst asks here, then, might be 'what kinds of things have agency in talk about medication?' and 'what kinds of spaces are opened for subjectivities to inhabit?'.

Although I described seven separate binary oppositions (with fourteen poles) above, one could argue that these revolve around two central poles. The first pole of each opposition could be characterised as exemplifying an optimistic, positivist, empiricist and biological account which is institutionally dominant in the practice of Western psychiatry. The second pole of each opposition maps onto an account which reflects a range of alternatives to this dominant account. The second pole is more varied, fragmentary and contradictory than the first and this is because the first pole reflects a culturally available and dominant account whereas the second pole reflects all those accounts which are subjugated and seen as alternatives. In other words, all that the accounts clustered around the second pole have in common is that they are devalued, at the margins and are spaces of resistance to the dominant account.

Much of the critical work on madness has been located in the context of an analysis of the institutions engaged in its regulation, the asylum of course having a central role here (eg Foucault, 1967). However, increasingly, with the growth of 'community care', commentators have attempted to theorise how power is distributed spatially and conceptually across different sites. Rose (1986) has described some of the different sites available within the contemporary practice of mental health. Nettleton has noted that 'whilst care and control in institutions were described by Foucault as Panoptic power, care and control of people in the community have been characterized by what Armstrong [1983] has called "Dispensary power" '(1995, p.247). Whilst surveillance continues through out-patient appointments and visits by community workers like CPNs, in this form of organisation, it could be argued that medication plays a significant role. When used within the mental hospital, medication is one of the forces which mark patients out and regulate them. Outside the confines of the hospital, however, medication is one of the social forces which effects this regulation by fixing patients in a diagnostic space which is not so much geographical (as in the physical separation of people in asylums) as conceptual and ideological. Medication also continues forms of physical regulation which are not perhaps as overt as they may have been in the past (eg through the use of confinement and physical restraint) but are more covert, internal and pharmacological. Medication becomes a wing of the hospital -- an internal asylum -- together with visits (now in the home) by health and social service workers. The use of medication constructs its subjects in particular ways and they then have particular powers and duties. Often the subjectivity, as we have seen, is passive and the only construction available for active agency is to resist and refuse treatment. The use of medication also constructs particular kinds of objects with particular effects. In the previous section we focused only on single oppositions whilst we noted in chapter 3 the importance of examining how such oppositions work together. Here, however, I will be attempting to analyse some of the interview extracts in more detail to map some of these oppositions and describe how they work together to construct these kinds of objects and subjects. I will also be attempting to locate myself more clearly in the production of these extracts.

6.5.1 A middle-of-the-road position on talking treatments

Not all accounts discussing the relative importance of treatments drew solely on one side of the talking treatments binary (6.4.7). As talking treatments have become increasingly popular in mental health and, having achieved some dominance in dealing with 'neurosis', are being introduced more into dealing with 'psychosis' (Parker et al., 1995), accounts have drawn on both sides of the binary. Hammersley (1995), for example, tries to navigate between medical/pharmaceutical and counselling/therapy narratives. She argues that there will never be a pill for every emotional problem and suggests that counselling is important. Yet she still maintains that it is equally unrealistic to believe that drugs are always harmful and that unlimited counselling should always be on offer. In combining these two accounts she argues that they are interested in different phenomena and that whilst medication produces biochemical effects, counselling can be used to address issues relating to the meaning medication has for users and prescribers through a focus on 'attitudes' and 'beliefs'.

Such accounts utilise a rhetoric of balance and a liberal assumption that all points of view have some utility (cf Billig, 1987b). However, narratives of balance and eclecticism, or middle of the road accounts, whilst appearing to acknowledge criticisms, may actually work to assimilate them and thus can paradoxically work to support the status quo of current practice. Thus Anthony Clare's (1976) book Psychiatry in Dissent managed to relativize the criticisms of psychiatry through the use of a rhetoric of 'balance'. Moreover, such a narrative constructs the speaker as a thoughtful, liberal professional or knowledgeable lay-person who weighs up pros and cons in a careful, rational and balanced manner. Accounts constructed in this way have a number of effects: they maintain the importance of medication (and therefore the continued administration of drugs and the acceptance of a biochemical factor in distress) whilst allowing talking treatments some power, thus warranting a claim that care is more holistic. Emphases in these middle-of-the-road accounts may vary, with counsellors accepting the use of medication on pragmatic rather than ideological grounds and biological psychiatrists acknowledging the use of talking treatments on similarly pragmatic grounds, whilst asserting the primacy of medication. Thus Manschreck (1996) has described the use of 'pharmacotherapy' as just one out of many treatment strategies in dealing with delusions alongside other interventions like psychotherapy. Indeed, Beitman (1996) has discussed the importance of integrating psychotherapy and pharmacotherapy models and interventions. Such accounts tend to use a 'biopsychosocial' narrative to accomodate apparently conflictual accounts. The notion of balance may well be drawn on to construct different interventions as complementary. In these accounts then, medication is just another therapy and a biological model is maintained through the assimilation of other narratives.

Dr Williams: When, when a patient is floridly psychotic I suppose er he needs medication at that time. I don't think talking about delusions or confronting them will make any difference to them <Dave: Uh-huh> because they won't listen to you. <Dave: Uh-huh> Once they are a bit, you know, cured with anti-psychotics they're more, more amenable to discussion and, you know <Dave: Uh-huh> to your reasoning. <Dave: Uh-huh> But not when they first come in with florid psychotic symptoms, I don't think so.

Dr Williams (SHO in psychiatry)

Medication is noted as the first treatment to be used, thus constructing it as primary. The issue of drug treatment is then constructed as an either/or choice -- this is structured as if to defend against an implied challenge that talking replace drug treatment (this had not been suggested: the previous turns had concerned who might offer talking treatments). There is no claim to substantiate this statement, rather it operates as a category entitlement. Here, there is also a deployment of the binaries 6.4.1.1 and 6.4.2.1. In constructing such an opposition, an extreme case formulation is then used in suggesting that when someone is 'floridly psychotic' they need medication since they 'won't listen to you' (lines 290-291) and they are not 'amenable to discussion' (line 292). The phrase 'floridly psychotic' constructs the subject here as someone who is very mad, drawing on the image of the irrational mad person who is not able to hold a coherent conversation. This is allied to the notion that talking treatments are appropriate for 'neurosis' but not 'psychosis' (Pilgrim & Treacher, 1992). This subject is contrasted with the professional, implicitly constructed here as rational and balanced, since the psychotic subject, when 'florid' will not be 'amenable to discussion' or 'your reasoning' (lines 292-293). This might be a strong challenge to those cognitive approaches which stress the importance of attempting to understand delusions by assuming the position of a rational but mistaken person, trying to work out how the person has come to the conclusions they have. What is given agency in this account? Both the psychosis and the person are given agency here in that the psychosis is 'florid' (line 288) whilst the person 'won't listen to you' (lines 290-291). One effect is that the subject here is constructed both as the passive recipient of psychosis and as an obstructive individual who is not 'amenable to discussion' (line 292). Both these narratives have the effect of constructing talking treatments as not being able to make 'any difference' (line 290). If the subject was constructed as an active and co-operative agent, amenable to discussion, the opposite effect would be achieved in that talking treatments would be seen as appropriate since the subject so constructed would be that of the ideal psychotherapy client.

The 'delusions' are constructed as abstracted from context, and are seen as 'out there' rather than within the person or their world. This enables the person's beliefs to be transformed into a particular class that are seen to have something in common (eg other delusions) and have certain properties (eg conviction and so on). One property is that the delusions rather than the person are seen as having a separate identity from the person and as having agency: it is the talking about or confronting of the delusions, rather than of the service-user which is seen as not making any difference. Here then, through the way the problem is constructed (as delusional and psychotic), so too, is the solution. Constructing delusions as separate and as having certain properties and agency and as inaccessible through conversation renders talking treatments inappropriate. If, however, the delusions had been constructed as, to some extent, subject to the service-user's agency, and as open to discussion, then talking treatments might have seemed an appropriate intervention.

This account is open to challenge, but the narrative includes qualifiers like 'I suppose' (line 288) and 'I don't think' (lines 289 and 294) which positions Dr Williams as thoughtful and open-minded and which could be used to rebut such a challenge. Discussion might be more amenable but this is only after the user is already 'a bit, you know, cured' and this rhetorical strategy constructs the medication as the primary curative intervention, whereas psychotherapy here is constructed in less clear terms: as simply 'talking about delusions' or as 'confronting them', and neither statement conveys any curative powers. This account, then, warrants both the continued use of medication as a treatment of first resort and the primary position of psychiatry in administering this treatment since non-medics cannot prescibe medication.

6.5.2 Multi-factorial talk and rhetorics of chronicity

I noted in 6.4.6.2 how a rhetoric of chronicity was useful in accounting for why medication had not apparently worked by locating the cause and responsibility for this within permanent, biological and constitutional factors located within the individual user. This kind of narrative occured a number of times within the interviews. It could also be deployed together with other forms of narrative. One example was in the interview with Dr Lloyd where it was used in an extract which also included the use of a multi-factorial approach in explaining why delusions had persisted despite drug treatment.

Multi-factorial talk could be seen as a further rhetorical resource to achieve certain kinds of effects. Gabe & Lipshitz-Phillips (1984) noted how GPs had used multi-factorial explanations of their patients' problems. However, whereas their analysis seemed to suggest that such descriptions were simply more accurate, the accounts can also be analysed for the effects they achieve in talk. Multi-factorial accounts, like 'balance' and 'eclectic' talk, warrant a biopsychosocial model of psychopathology; whilst on the one hand appearing to be liberal, open-minded, eclectic and flexible, this kind of account actually may also function in a conservative manner by relativising challenges, and thus functioning to maintain current practice. The following extract is an example of how oppositions can work together to respond flexibly to challenges.

Dave: Right, why don't you feel, why do you feel the medication hasn't affected the beliefs?

Dr Lloyd: This is the interesting thing. I think because they're <pause> they're deeply ingrained thought patterns <Dave: Right>. I don't think you can any more just say they're illness-based. Their origin is illness-based but they have had <pause> their existence and their entertainment and their re-inforcement by repeated thinking has had positive effects for Alan for the reasons that I gave earlier and therefore, even if the reason for the existence of the <pause> delusional beliefs is no longer there, assuming <Dave: Uh-huh> <pause> that the neuroleptic medication that affected the other things affects what lies at the bio-chemical core of those delusions <Dave: Uh-huh> they're too important to leave. <Dave: Right, right> I mean that's, that's one possible explanation <Dave: Uh-huh> I don't think Alan doesn't have schizophrenia apart from the delusions because he's on neurolepric medication. Erm it's difficult to know now what's long-standing ingrained personality patterns and what's long-standing, ingrained, chronic, psychotic deficits and what is institutionalization from being 25 years inside <Dave: Right> erm <pause> we will never really know.

Dr Lloyd (Consultant Psychiatrist)

This account is extremely complex, combining a number of rhetorical devices. The effects which are thus achieved are varied too. First, a number of suggestions are made as to why medication does not appear to have affected Alan's beliefs. These include: (i) 'deeply ingrained thought patterns' (line 249); (ii) reinforcement through repeated thinking leading to positive effects (lines 251-252); (iii) 'long-standing ingrained personality patterns' (lines 259-260); (iv) 'long-standing ingrained chronic psychotic deficits' (lines 260-261); and (v) 'institutionalization (line 261). These various theories are presented in a five part list - such lists and contrasts have been noted to be powerful in producing factuality (Edwards & Potter, 1992). This extract combines ideas eclectically from a range of theoretical viewpoints. Such eclecticism is useful since it gives the account sufficient flexibility to meet a variety of possible challenges. All these descriptions form part of an opposition of pathology/normality in that all the terms imply pathology and deficit -- a common feature of the diagnostic language of the psy professions (Gergen, 1990). Thus the subject constructed here is pathological and deficient. Indeed, Light has noted how when psychiatrists learn to diagnose, often through anecdotal material alone, they 'learn to look mainly for pathology' rather than fuller descriptions (1980, p.162).

Terms from a wide range of theoretical frameworks are used here: sociological; behavioural; cognitive; personality and biological psychiatry. One effect of this is that even if one of these were challenged -- eg the biological model, since medication did not appear to have worked in this case -- then the other candidates could be brought in to explain the continued presence of delusions. Multi-factorial talk is flexible in another sense. Such talk has the possibility of presenting the various theories as equally valid but as fixed within a hierarchy, with biology at the 'core' (line 255) with other issues, such as cognitive or behavioural factors considered to be the mere effect of underlying biological mechanisms. There is another important point here which is how the oppositions work together. The challenge to biology posed by the failure of medication (lines 246-247), which is met by a move to thought patterns (line 249) might at first seem a progressive move from a passive biological subject to an active cognitive subject. However, in this move the oppositions pathology/normality and individual/social are not challenged so the thought pattern is still seen as in the individual and as pathological. Moreover, the speaker assimilates both biological and psychological factors by making a distinction between the origin (line 250) and current maintaining factors of the delusions (line 251) and this is aided by the failure to challenge the two oppositions just noted. Even when the social is acknowledged, as when a sociological term like institutionalization is used (line 261), it is used in an individualistic context, and so fails to challenge the individual/social opposition. In a sense then, whenever there appears to be an escape from a dominant biological psychiatry view, the alternative option is already set up still to construct a pathological individual(16).

At the same time, qualifications are continually offered, and these include 'I think' (line 248), 'I don't think' (line 249, lines 257-258), 'assuming' (line 254), 'that's one possible explanation' (line 257), 'it's difficult to know' (line 259), 'we will never really know' (line 262). This set of qualifications has the effect of introducing ambiguity, vagueness and tentativeness. This is useful, since any individual challenge can be met with the response that only a tentative hypothesis was being proposed together with a flexible move onto another such hypothesis.

The comment that the medication has affected the illness which is the origin of the symptoms and which is at the 'bio-chemical core' (line 255) appears to be an attempted solution to the problem of the medication not having affected the symptoms. However, this solution causes another dilemma: if the medication has affected the cause then why do the symptoms persist? The attempted solution to this is, drawing on a hierarchical model, that although the origin is the illness (line 250) the beliefs persist for a variety of other reasons noted above. If there had been a solely biological account, a solution to this might have been difficult. A biopsychosocial account is able to handle such dilemmas, neutralising challenges to a biological account through the use of psychosocial theories. The separation of the notions of origin/cause and maintaining factors is useful in this context.

The implication that medication does not work is addressed through two means: first, that the medication affects an underlying illness (using metaphors of depth common in foundationalist theories); second, by pointing to the effect of the medication on 'other things' (line 255), ie other symptoms. This again shows the flexibility of these kinds of repertoires. First, depth metaphors continually offer us the possibility of knowledge from 'experts' (a form of category entitlement) of an unseen realm below the surface (here, the illness) -- this knowledge cannot be verified but only supposed through the use of surface signs (ie symptoms). Second, the symptom repertoire can meet challenges based on the failure to achieve change in one symptom by pointing to change in another symptom.

A final point here concerns one of the theories suggested by Dr Lloyd in his four part list at the end of the extract where he talks of 'long-standing, ingrained, chronic psychotic deficits' (lines 260-261). This phrase is a further example of symptom-talk where the symptom (here cast as a deficit) is given agency. However, by virtue of its being seen over a period of time it is granted some level of permanence by being described as 'long-standing' and 'chronic' and is constructed as constitutional -- and we have seen already (in 6.4.6.2.4) the effects of chronicity talk. The use of the word 'ingrained' conveys a three-dimensional image of inscription which appeals to a variety of depth metaphor. The repeated use of these words here, then, convey that these symptoms are to be seen as permanent. Moreover, they are thus transformed from symptoms of psychosis into 'psychotic deficits' which are seen as affecting a person's personal abilities in an active and negative way.

One effect of the use of these resources is that the theories and symptoms are seen as having agency whereas Alan is constructed as ambiguous: on the one hand a passive victim of biology, institutionalization and so on; on the other hand an active rational agent. One effect of this is as noted earlier, that some blame can be located with Alan for the drug non-response. Another effect is that by constructing the objects of talk here as disembodied, interventions can be constructed as if they will affect only these factors rather than Alan as an embodied agentive subject.

6.5.3 Symptom-talk, maintenance and relapse

We noted in 6.3.1.5 how the taking of medication could be constructed as essential to maintaining well-being rather than aiming for cure and we saw how one effect of this kind of account was to warrant the continued use of medication, with the attendant dangers of this minimised by the use of the rhetorical device of 'monitoring'. This kind of account is worth considering in more detail because of the particular kinds of object and subject it constructs.

Nettleton (1995), in her description of dispensary power notes how one of the effects of this move was not only to address the actual pathology already known but also to look deeper into the apparently normal and non-pathological population for 'potential pathology'. The notion of relapse and the need for maintenance medication as a form of prophylaxis is related to this view. Indeed, in chapter 4, we discussed Castel's (1991) and Rose's (1996) views on risk. In talk about mental illness it often seems to be assumed that potential pathology lurks beneath the surface: professionals talk of cases having gone into 'remission' and of the danger of 'relapse'. This constructs the subject of such discourse as a high risk individual(17), a passive and helpless victim of a pathology that might rear up again, unexpected at any moment, leading to a relapse, ie a return of the previous symptoms.

The objects constructed in this kind of discourse include the 'underlying disease' and the symptoms which are considered to be the visible signs of this disease. Symptoms here are constructed as independent of subjects and their contexts, agentive and unpredictable. Since normality may only ever be apparent there can never be proof that the pathology will not return. Here medication is seen as a way of warding off this threat. 'Relapse' thus functions to continue to fix the apparently normal psychiatric subject in diagnostic space even though symptoms may not be apparent. Of course, it also functions to warrant continued medication. It is important to note that this is not the only biologically-informed account available. Thus Goldberg & Huxley (1992) have noted how many mental health difficulties may resolve over time without professional help. Moreover, notions of using medication only before or at times of relapse ('targeted medication' -- Carpenter et al., 1987; Healy, 1997) may allow some agnosticism about the existence of underlying disease states.

The notion of relapse, then, is based not only on an idea of agentive symptoms but also of a separate and agentive disease state, of which the symptoms are merely signs. These notions rely on foundationalist metaphors of depth, of reality being at a deeper, underlying level. This disease state is constructed variably within psychiatric discourse, seen as both impermeable to neuroleptic medication (which is only able to impact on symptoms) and as being treated directly by the medication. The former account draws on psychoanalytic discourse. Thus we have noted earlier Light's (1980) description of psychoanalytically oriented psychiatrists who regarded the effect of medication simply as removing symptoms without addressing fundamental underlying pathology which could only be dealt with through psychotherapy. He noted, however, that there was scant evidence for this view.

In the following extract we will see how the notions of risk of relapse and maintenance medication are constructed.

Dave: Right, right. With, with both her and Mike er at various points, they stopped treatment. <Dr Smith: Yes> Do you know why, why they did that?

Dr Smith: Well obviously in general I think people <pause> er dont like having any treatment erm and erm they believed that they shouldnt be on any treatment on a longer-term basis er quite often they have difficulty in accepting that they have got an underlying illness and they need the treatment to remain well.

Dave: Right. <Pause> Do they ever stop it because they say its not having any effect? <Dr Smith: Sometimes> Would that ever have been true for Sharon or Mike?

Dr Smith: I dont think that Mike Sullivan erm <pause> obviously when they have treatment they, they get better and they say "Well Im alright now, why should I have the treatment".

Dave: Right, right. <Dr Smith: Erm> <Pause> So they feel then that if they stop/ <Dr Smith: /Stop/> itll be okay. <Dr Smith: Yeah> And you, you would say what to that? You, you would/

Dr Smith: Well I would say "Your treatment is only controlling your symptoms. Without treatment erm er what I can tell you is er you are very likely going to relapse".

Dr Smith (Consultant Psychiatrist)

What kind of objects are constructed here? First, there is 'treatment' (line 357), constructed as homogeneous, 'out there', non-specific but assumed to be medication. It is seen as essential to the control of a second object, or cluster of objects, known as 'symptoms' (line 371) which are seen as unpredictable and agentive. Treatment is clearly contested, thus Dr Smith states that Sharon and Mike use it to get rid of symptoms, that is until they 'get better' (line 366) whereas he views it as necessary to prevent relapse, a third object. Relapse is seen as a resurgence of the original symptoms, of the 'underlying illness' reasserting itself. It is a marker of potential pathology. The underlying illness is, of course, a fourth object, shrouded in mystery (since only the symptoms can be seen) and only reachable by indirect means. However, it again is seen as agentive but perhaps more predictable since Dr Smith warns that Sharon and Mike are 'very likely' to have a relapse without medication. Of course the notion of relapse serves important functions in that the prospect of relapse regulates medication use: many users take medication to avoid the danger of a relapse (Day et al., 1996). There are other functions too. Healy, for example comments in relation to medication that 'the production of compounds aimed at minimising risk has the advantage that a greater number of individuals can be deemed to be at risk of disorders than will ever acquire those disorders' (1991, p.244).

These objects have particular qualities. Thus treatment is seen as: involving users' choice; being targeted at symptoms; and controlling rather than curing. Symptoms are seen as having power but as controlled by the treatment. The relationship between the symptoms and the underlying illness is ambiguous. However, the medication seems aimed at the symptoms rather than the illness. The aim is control not cure. At other points though, the speaker uses a 'cure'-type account in talking of users 'remaining well' (line 361) and getting 'better' (line 366). There is a complex movement between the control/cure and symptom/illness oppositions.

What kind of subjects and subjectivities are constructed here? The users are seen as inconsistent when they stop treatment (line 357, lines 366-367). This inconsistency is seen both as a result of personal opinion (lines 358-359) and also as contingent on the illness and on insight (line 360 and line 366). The subject progresses through the narrative of lack of insight treatment cure insight. Thus the user subject is seen as both responsible and not responsible for their compliance. There are three oppositions at work here: the notion of treatment as being stop or go (as opposed to a continuum); the image of users as either compliant or not compliant; and the idea that the psychiatric subject either has or does not have insight (and is therefore responsible or not responsible for their actions). Since compliance is constructed as a binary opposition (and as within the user rather than in a relationship of doctor and user), movement between them is constructed as idiosyncratic inconsistency rather than as an assertion of agency.

The users are also constructed as naïve in some sense since they are seen as at risk of feeling they are better when they are not (lines 366-367), and they require the intervention of those with expert knowledge. They are seen as passive victims of agentive symptoms, implying a mechanical view of the self. However, running alongside this thread is a moral account which views the subject as responsible for its circumstances: if only they had complied with their medication, there would be no risk of relapse. In this way the user subject is constructed as both a passive victim and as a responsible moral agent(18). The other subject constructed here is that of the professional subject. They are seen as having expert knowledge both of the users' problems and of the required intervention. In a sense they are omniscient, predicting relapse when medication is stopped.

6.6 DISCUSSION

This chapter has been concerned with examining a dominant mode of professional intervention in the lives of many people considered to have paranoid delusions, that of medication. Rather than explore these accounts from a naïvely realist perpective, asking 'is this account accurate?' we have examined some of the effects of these accounts and asked 'what does this account do?' and 'how does this account achieve those effects?'. In the course of this examination, I have argued that the talk about medication in my interviews with professionals and users involved the use of a number of rhetorical devices and can be understood to cluster around a number of binary oppositions (and, in particular I have focused on how speakers account for times when medication does not seem to have 'worked'). I have suggested that these oppositions can be seen to form around one basic opposition: that of a biological psychiatry account versus an account comprising a range of alternatives. I have then gone on to describe some of the objects and subjectivities constructed within medication discourse and their effects. I then argued that some of the apparent attempts to 'escape' an opposition fails as other oppositions are already set up. I have noted in chapter 3 how attempts to deconstruct one opposition will fail if other oppositions are not similarly challenged. We have seen this once again here as I have described how the objects and subjects of medication discourse lock together in a web or field of oppositions.

So are we to conclude that all professional intervention is flawed or that it is impossible to escape these oppositions? What are some of the rhetorical strategies I have used in constructing these accounts and how do they relate to my location? Responses to these questions will be addressed in the next, final section concerned with both a reflexion on these chapters and some proposals for practical change for users, professionals and others.

1. Parts of this chapter have been published in Harper (1998 and in press).

2. However, psychologists are now calling for similar prescribing privileges (see special

issue of The Psychologist, 'Prescribing privileges for clinical psychology', April 1995 and articles in the American Psychologist, March 1996).

3. Ross & Pam (1995) make similar comments about the 'pseudo-science' in much biological

psychiatry.

4. A 'depot' in this context refers to a long-acting form of neuroleptic medication administered by an intra-muscular injection (instead of being given in tablet form) usually by a Community Psychiatric Nurse. Since it is a slow-release form of medication it only needs to be administered every 1-4 weeks and is used instead of oral medication.

5. Of course, as Holmes (1997) notes, many users (especially those on neuroleptics) do know the name and dosage of their medication. But of course expert knowledge can be learned as a skill by non-experts. Such expertise by users, however, may be minimised by professionals who might then term such a user a 'professional patient'.

6. Munro & Mok (1995) note that 'it is quite possible that Pimozide is being promoted by a number of enthusiasts' (p.619).

7. Johnstone (1997) has noted the emergence of a more sophisticated version of the chemical cause argument which assumes that the initial causes of emotional distress are social and psychological but concludes that this leads to a fixed imbalance in neurotransmitters which then becomes independent of the original causes so the person becomes biologically 'stuck' in their distress and thus the best intervention is a chemical one. Such an argument thus admits psychosocial causes whilst retaining biochemical solutions.

8. Thus in Li's (1996) review of Ancill et al. the following paragraph is quoted: 'Nearly 20 years of CT and post-mortem brain research in schizophrenia has made it abundantly clear that there is no gross structural pathology associated with the disorder ... It may legitimately be asked if there are any convincing abnormalities of brain structure in schizophrenia. Notwithstanding the present uncertainties in the MRI literature, the answer is likely to be yes' (pp.131-132). Li goes on to comment 'I cannot but conclude that it is faith operating here rather than science or critical thinking' (p.80, emphasis in original).

9. There is an interesting symbiosis between the development of neuroleptic medication and aetiological theories in biological psychiatry. Some commentators have noted how the appearance of new neuroleptics leads to revisions of biological theories of aetiology. However, at the same time, these biological theories of aetiology are used to warrant drug treatment. This leads to a curiously tautological situation where causes are inferred post hoc on the basis of interventions which are warranted by those same causes. This is not a situation peculiar to psychiatry. Gabbay (1982) makes the point that theories of why certain medical interventions work change according to the paradigms of the age. A similar point about inferred causes-as-warrants could be made about non-biochemical interventions like psychotherapy too.

10. Indeed Day (1995) and Day & Bentall (1996) report that Chlorpromazine was originally synthesised as an anti-histamine in France and its anti-psychotic properties only discovered later. This lends weight to the view that, to some extent, interventions are accidental and explanatory theories are often post hoc rationalizations.

11. For work questioning the effectiveness of medication see Johnstone (1993b) and Kriegman (1996).

12. Papers published in a supplement of the British Journal of Psychiatry provide further evidence of the rhetorical strategies noted here. Entitled 'Management of difficult to treat patients with schizophrenia', the supplement included papers which were overwhelmingly biological in orientation (eg Sharma & Kerwin, 1996) which one might expect since treatment resistance might be seen to undermine a biological case.

13. Harré (1991) notes the common tendency for medical terms to, at first sight, appear to be informative and explanatory when, in fact, they are simply names. Thus the terms 'treatment-resistant' or 'treatment-refractory' are not particularly informative and it is hard not to infer some professional frustration in the choice of the word resistance which seems to imply motivation and intention on the part of patients.

14. Indeed, Troisi et al. (1997) have reported a positive correlation between psychiatric

patients' scores on the paranoid/belligerence cluster of the DSM-IIIR Positive and Negative Symptom Scale and daily dosage of neuroleptic treatment. This suggests that this strategy is especially available to professionals working with users diagnosed as paranoid.

15. Kane (1996b) lists 'comorbid conditions' as one reason for 'refractoriness to treatment'

16. I am not claiming that this kind of account does not provide more opportunities for alternative interventions, simply that we must recognise some of the implications of liberal accounts.

17. Blackman (1996) has given an account of the history of the notion of 'risky' personalities and classes.

18. In such accounts it is rarely explained how, if a user is taking their medication, they might lose insight. If they are taking their medication and decide to stop taking it how can this be a sign of the illness? Either way it is likely that such a choice will be pathologised as either lack of insight or non-compliance for other reasons.