Principles of Social Psychiatry

CATEGORIES AND CONTINUUMS 33 3.2. RELIABILITY AND VALIDITY At first sight there appears to be no contest between patterns from the norm’, manifest mainly as dysfunc- these two approaches; the categorical approach wins tional behaviour that is ‘pervasive’, and which leads to hands down–butitiswisetothinkagainbeforecoming distress or impairment in social, occupational and to this decision. Doctors and decision makers in all personal function. This, although more reliably rated parts of medicine prefer categorical classifications to than any of the individual personality disorders [4,5], dimensional ones as clinical decisions are categorical is still extraordinarily subjective, and it is easy to see ones. You cannot decide to 30% treat a patient or admit how the diagnosis can become, in the words of Lewis 50% of a patient to hospital, so the decision maker and Appleby, ‘the patients psychiatrists dislike’ [6]. abhors the dimensional system as unworkable in prac- Dimensions are less stigmatic, more reliable and tice.However,thisassumesthatthecategoryconcerned avoid false or arbitrary dichotomies. The reason they is a valid one. The trouble is in psychiatry that so many are more reliable is really a statistical relationship of our diagnoses are not valid because we have no between the numbers of options available [7], sum- external tests that are independent and can answer the marized in Table 3.1. true test of validity, ‘Does this actually measure what it The caveats in Table 3.1 arevery important. In basic purports to measure?’ So the identification of the language they are saying to the diagnoser, ‘Are you pneumococcus in my sputum was the clincher for my confident that the condition you are trying to classify diagnosis of pneumococcal pneumonia, and although has no clear natural boundaries?’ If it has, then these the sudden development of all the symptoms and signs will determine the number of groups that are chosen. of lobar pneumonia was almost as good, it was the Thus, for example, in dementia there is increasing identification of the bacterium that really counted. evidence that there are two main categories of dis- When there is no external measure of validity we fall order, full Alzheimer’s disease and mild cognitive back on gaining excellent reliability, and even though impairment [8], and that these can be identified with we are reminded that even 100% reliability in the increasing confidence. It would therefore be inap- detection of an alleged diagnosis adds not one jot to propriate to have a seven-point scale for diagnosing validity we are still reassured that without good relia- Alzheimer’s disease as here we do have sufficient a bility we are not going to achieve a valid diagnosis. priori knowledge to be confident about the diagnosis There is now increasing evidence from research studies of individuals using a few categories. in all areas of medicine where there is no independent However, in psychiatry this extra knowledge, measure of disease that dimensional systems are more mainly gained from independent biological and neu- reliable.Thisisparticularlytrueofpsychiatry,andinthe ropsychological evidence, is relatively rare. As David complex jungle of social psychiatry it is seen in its Kupfer, charged with a major role in the construction clearest form. Let me take one example, the vexed of DSM-V, puts it pessimistically ([9], p. xvii): question of the classification of personality disorder [2,3].Thisisahighlyunpopulardichotomousdiagnosis. Despite many proposed candidates, not one labora- According to current diagnostic practice people either tory marker has been found to be specific in identify- ing any of the DSM-defined syndromes. Epidemio- have this nasty condition that no-one in their right mind logic and clinical studies have shown extremely high could derive any satisfaction in having or you are rates of comorbidity among the disorders, undermin- squeaky clean with no personality disturbance at all. ing the hypothesis that the syndromes represent dis- How do we decide on the dividing line between tinct etiologies. Furthermore, epidemiologic studies awfulness and smug virtue? This is a very arbitrary have shown a high degree of short-term diagnostic process. Thus the ICD-10 and DSM-IV definitions of instability for many disorders. With regard to treat- personality disorder describe this condition as ment, lack of treatment specificity is the rule rather ‘deviations of personal characteristics and behaviour than the exception.

34 PERSPECTIVES AND METHODS Table 3.1 Improved reliability of dimensional continuums compared with categorical classification . A computer simulation study (Monte Carlo study) designed to investigate the extent to which the interrater reliability of a clinical scale is affected by the number of categories or scale points . All scale points between 2 (dichotomous) and 100 were examined . Results indicated that reliability (agreement between raters) gradually increased between 2 and 7 points in the scale . Beyond 7 points no substantial increase in reliability occurred no matter how many scale points were examined Caveats These findings only hold under the following conditions: (1) The research investigator has insufficient a priori knowledge to use as a reliable guideline for deciding on an appropriate number of scale points to employ (2) The dichotomous and ordinal categories being considered all have an underlying metric or continuous scale format (From Cicchetti, D. V., Showalter, D. and Tyrer, P. (1985) The effects of number of rating scale categories on levels of inter-rater reliability: a Monte Carlo investigation. Applied Psychological Measurement, 9, 31–36.) This candour is refreshing after the initial enthu- At present there is little evidence that most contem- siasm that greeted the introduction of DSM-III in porary psychiatric diagnoses are valid, because they 1980 [10]. Far too many people assumed that a reliable are still defined by syndromes that have not been diagnosis derived from operational criteria was also a demonstrated to have natural boundaries. This does not mean, though, that most psychiatric diagnoses are valid diagnosis. In truth, very few diagnoses in psy- not useful concepts. In fact, many of them are invalu- chiatry can have any claim to validity; the best that can able. But, because utility often varies with the context, be obtained is a high level of clinical utility in the statements about utility must always be related to setting of its use, nicely described in the words of context, including who is using the diagnosis, in what Kendell and Jablensky ([11], p. 8): circumstances, and for what purposes. 3.3. CHOICE OF CATEGORY VERSUS CONTINUUM From the guidance given above we can postulate a continuum from existing categories and how others can number of rules that can help to decide whether it is feel more secure in their current categorical state. prefereable to use a category or continuum for any Anorexia nervosa has support as a category as there particulardiagnosticconstructinpsychiatry(Table3.2). are independent measures (weight loss, endocrine dis- Using the information in Table 3.2 it is easy to turbance), there is clear clinical differentiation from see how some disorders might best be moved to a other eating disorders (with a little concern about a Table 3.2 Questions to be asked before deciding on a categorical or continuum scale Knowledge question Answer Choice of category or continuum Is there a truly independent measure of Yes Use category the diagnosis? No Be inclined to use a continuum Are there clear boundaries between the disorder Yes Use category and related ones? No Use continuum Does the disorder have a different natural history Yes Use category and course from other (apparently related) disorders? No Use continuum Is there a specific treatment for the disorder that makes Yes Use category it useful to identify separately? No Use continuum

CATEGORIES AND CONTINUUMS 35 fuzzy boundary with atypical anorexia nervosa) and the measures justifying its diagnosis – we are still looking natural history of the condition is different from other for a successor to the dexamethasone suppression disorders (with a very much higher mortality rate) [12]. test [13] – and its boundaries arevery difficult to define. Other conditions, such as ‘major depressive episode’ in It is therefore not surprising that ICD-10 [14] moves DSM, are much more shaky as diagnostic entities. towards the continuum option of mild, moderate and Major depressive episode does not have independent severe depression in its classification. 3.4. IMPORTANCE OF CLINICAL UTILITY Diagnosis is a useful shorthand for clinicians wishing what false dichotomy between normal personality and to communicate a commonality in a few words, and a personality disorder. Using this simple categorical sys- good diagnosis is only valuable if it is perceived tem it appears that between 4 and 10% of the total adult as useful by the user. This is especially true of treat- population of any country has a personality disorder ment. Any condition that leads to a specific treatment [16,17]. These are staggeringnumbers,asit implies that recommendation (probably the most important item 150 000–400000 people in the UK have this medical in Table 3.2) is going to be very popular in clinical condition. Subsumed within this group are around 300 practice – but is has to be backed by proper evidence. people who have alleged ‘dangerous and severe per- Thus, a suggestion several years ago that pimozide, sonality disorder’ [18], but in the present classification an otherwise unremarkable antipsychotic drug, was a structure they are indistinguishable from the others. specific treatment for monosymptomatic hypochon- It had been noticed in several past studies that driacal psychosis [15] led to a great interest in both the more severe personality disordered patients tend the diagnosis and treatment of this condition. The to have a much greater number of personality dis- evidence for this specificity was always very flimsy orders than thosewith milder disorders [19–21]. In our and has never been supported and so both diagnosis own work we found that those with single personality and its alleged special treatment have faded into disorders with apparently severe symptoms paradoxi- oblivion. cally had less associated morbidity than those with Clinical utility is also an important consideration milder symptoms but with more personality disorders when a condition is very common and yet only a that crossed clusters. This led to a new continuum of minority of people can receive, or are available, for classification [22] (Table 3.3) that has been shown to treatment. Earlier in this chapter I indicated the some- very useful in practice [23]. Table 3.3 Example of a continuum scale to record the severity of personality disorder Level of personality disturbance Features Level 0 No personality disorder (i.e. good capacity to form relationships, reasonable personal resources to draw on at times of adversity) Level 1 Personality difficulty (tendency for enduring patterns of behaviour to interfere with social functioning at times of particular stress and vulnerability but not at other times) Level 2 Simple personality disorder – particular and persistent personality abnormalities that create significant problems in occupational, social and/or personal relationships (present cutoff point for personality disorder in both ICD-10 and DSM-IV) Level 3 Complex or diffuse personality disorder – widespread personality abnormalities covering more than one cluster of personality disorders Level 4 Severe personality disorder – widespread personality abnormalities covering more than one cluster of personality disorders and leading to gross societal disturbance

36 PERSPECTIVES AND METHODS 3.5. SOME ADVICE FOR SOCIAL PSYCHIATRY This account indicates that both categories and 7. Cicchetti, D. V., Showalter, D. and Tyrer, P. (1985) continuums have a place in social psychiatry. In The effects of number of rating scale categories on the past too much reliance has been placed on cate- levels of inter-rater reliability: a Monte Carlo gorical descriptions that have often been based investigation. Applied Psychological Measurement, 9, 31–36. more on a combination of dogma and clinical con- sensus rather than evidence. The use of continuums 8. Ritchie, K. and Touchon, J. (2000) Mild cognitive impairment: conceptual basis and current nosological is preferable in the absence of evidence, as they status. Lancet, 355, 225–228. convey more information, generally lead to better 9. Kupfer, D., First, M. B. and Regier, D. E. (2002) agreement and produce data that are useful in gen- Introduction, in A ResearchAgendafor DSM-V erating hypotheses. However, there is no alternative (eds D.Kupfer,M.B.First and D.E.Regier), to a categorical description when decision making American Psychiatric Association, Washington, DC, has to follow on from the identification of a pp. xv–xxiii. syndrome or disorder, so all diagnostic continuums 10. AmericanPsychiatricAssociation(1980)Diagnosticand should have anchor points that allow categorical Statistical Manual of Mental Disorders, 3rd revision, separation, often a dichotomous one, so that they American Psychiatric Association, Washington, DC. can be used to advise decision makers. With this in 11. Kendell, R. and Jablensky, A. (2003) Distinguishing mind a somewhat nebulous concept such as ‘social between the validity and utility of psychiatric inclusion’ can now be placed on an investigative path diagnoses. American Journal of Psychiatry, 160, and explored in the same way as any other scientific 4–12. concept. 12. Papadopoulos, F. C., Ekbom, A., Brandt, L. and Ekselius, L. (2009) Excess mortality, causes of death and prognostic factors in anorexia nervosa. British REFERENCES Journal of Psychiatry, 194, 10–17. 13. Carroll, B. J. (1985) Dexamethasone suppression test: a 1. Likert, S. (1932) Technique for the measurement of review of contemporary confusion. Journal of Clinical attitudes. Archives of Psychology, 140, 1–55. Psychiatry, 46, 13–24. 2. Livesley, W. J., Schroeder, M. L., Jackson, D. N. and 14. World Health Organization (WHO) (1992) ICD-10: Jang, K. L. (1994) Categorical distinctions in the study Classification of Mental and Behavioural Disorders, of personality disorder: implications for classification. World Health Organization, Geneva. Journal of Abnormal Psychology, 103, 6–17. 15. Munro, A. (1978) Monosymptomatic hypochondriacal 3. Clark, L. A., Watson, D. and Reynolds, S. (1995) psychoses: a diagnostic entity which may respond to Diagnosis and classification of psychopathology: pimozide. Canadian Psychiatric Association Journal, challenges to the current system and future directions. 23, 497–500. Annual Review of Psychology, 46, 121–153. 16. Torgersen, S., Kringlen, E. and Cramer, V. (2001) 4. Bronisch, T. (1992) Diagnostic procedures of personality The prevalence of personality disorders in a disorders according to the criteria of present community sample. Archives of General Psychiatry, classification systems. Verhaltungstherapie, 2, 140–150. 58, 590–596. 5. Bronisch, T. and Mombour, W. (1994) Comparison 17. Coid, J., Yang, M., Tyrer, P. et al. (2006) Prevalence and of a diagnostic checklist with a structured interview correlates of personality disorder in Great Britain. for the assessment of DSM-III-R and ICD-10 British Journal of Psychiatry, 188, 423–431. personality disorders. Psychopathology, 27, 312–320. 18. Home Office/Department of Health (1999) Managing 6. Lewis, G. and Appleby, L. (1988) Personality disorder: Dangerous People with Severe Personality Disorder. the patients psychiatrists dislike. British Journal of Proposals for Policy Development, Home Office, Psychiatry, 153, 44–49. London.

CATEGORIES AND CONTINUUMS 37 19. Kass, F., Skodol, A. E., Charles, E. et al. (1985) Scaled 22. Tyrer, P. and Johnson, T. (1996) Establishing the ratings of DSM-III personality disorders. American severity of personality disorder. American Journal of Journal of Psychiatry, 143, 627–630. Psychiatry, 153, 1593–1597. 20. Oldham, J. M., Skodol, A. E., Kellman, H. D. et al. 23. Pulay, A. J., Dawson, D. A., Ruan, W. J. et al. (1992) Diagnosisof DSM-III-R personality disordersby (2008) The relationship of impairment to personality two semistructured interviews: patterns of comorbidity. disorder severity among individuals with specific American Journal of Psychiatry, 149, 213–220. axis I disorders: results from the National 21. Dolan, B., Evans, C. and Norton, K. (1995) Multiple Epidemiologic Survey on Alcohol and Related axis-II diagnoses of personality disorder. British Conditions. Journal of Personality Disorders, 22, Journal of Psychiatry, 166, 107–112. 405–417.



4 Social, biological and personal constructions of mental illness Derek Bolton Institute of Psychiatry, King's College London and South London and Maudsley NHS Foundation Trust, London, UK 4.1 CAUSES, DEFINITIONS AND BOUNDARIES There are various ways in which mental illnesses may truction of the representation of mental illness as interact with features of social organization and social opposed to health and the boundaries between them? processes at macro- and micro-levels. One set of What is the role of other factors, particularly the issues has to do with causal pathways affecting the biological and the personal? How does it come to be population prevalence of mental illness generally or of that this or that condition is termed an illness, as particular conditions, that is to say, involvement of opposed to, for example, distress or disability resulting social factors in aetiology. A different aspect of the from social exclusion, or a limitation of a person’s same issue is the influence of social factors on main- character, or a natural human condition, or life-style tenance of the condition, affecting duration and choice,andsoforth?Whyillness–withitsconnotations course. Relevant macro-features of social organiza- of pathological abnormality, a matter for doctors and tion and processes mainly have to do with exclusion medicaltreatment,tobereferredtoandmanagedbythe from access to resources of one kind or another, such great institutions of health care? Here features of social as education, employment, goods, social status and organization and power relations clearly play a major access to treatment, and they are realized in social role, with the views and values of government policy communities, families and individuals, interacting makers, popular cultural media, health care institutions there with micro-level risk and protective factors of and professions and manufacturers of treatment tech- many kinds. These questions of aetiology are only nologies (more or less coinciding with the opinions of mentioned here to set them aside as not the focus of the well-resourced and enfranchised key treatment this chapter – they are addressed elsewhere in the consumers) having dominant roles. In brief, it is current volume. plausible to suppose that the representation of mental The main questions for the present chapter are more illness as opposed to health is at least in large part like this: What is the role of social factors in the cons- created and maintained substantially by social factors. Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

40 PERSPECTIVES AND METHODS Here is Jaspers on this nearly a century ago ([1], pressures from a wide variety of stakeholders, and p. 652): vary across cultures and subcultures. To bring some order into this relativity, one needs What health and illness mean in general are matters something like a fact of nature, and the general which concern the physician least of all. He deals approach that follows this alternative tack is accord- scientifically with life processes and with particular ingly called naturalism. Naturalism seeks to underpin illnesses. What is ‘ill’ in general depends less on the the natural assumption often enough made by many judgement of the doctor than on the judgement of the physicians, psychiatrists, some other health care pro- patient and on the dominant views in any given fessionals, andevident in some laydiscourse, that there cultural circle. is some clear-cut distinction between illness and health, a fixed matter of fact independent of social On the other hand, while it may seem an easy slide opinions, a matter forthe biomedicalor biobehavioural from representations to social representations –we sciences, better recognized by trained health care can after all only think with concepts we learn in the experts, maybe hard to identify and clarify, but there culture – science stakes a claim to identifying phe- all the same, underpinning the discourse and the nomena that are not a matter of social opinion and institutions of health care. It is just a matter of getting values, but just are natural facts. Chemistry’s periodic clear about what this fact of the matter is. This turns table of elements, for example, would illustrate this out, however, to be very difficult. line of thought, as would the fact that the cardiovas- Naturalism in medicine has had a variety of forms cular system regulates blood pressure. So the question including as an unselfconscious presumption in the is: Does illness in general and mental illness in great paradigms in nineteenth century medicine and particular refer to some natural fact that is not a matter neurology: illness as disease or lesion. More recently, of social opinion and values, but is rather a matter for over the past few decades since the 1970s, the most the medical sciences? popular and influential kind of naturalism regarding Jaspers in the quote above apparently does not look illness and health has been based explicitly in ideas of to the sciences to work out what is illness and what is evolutionarytheory.This will beconsideredinthenext not, but refers instead to the patient’s opinion and section, but before that let us go back in time to when cultural views. Here is Kendell rounding on any such the whole problematic of the medical model in psy- idea ([2], pp. 307–308): chiatry first arose with gusto, in the 1960s. Two key themes in the 1960s critiques can be characterized The fact is that any definition of disease which boils briefly in the following way: first, that the medical down to ‘what people complain of’, or ‘what doctors model of mental illness rendered the conditions so treat’, or some combination of the two, is almost worse labelled and the people who have them meaningless, than no definition at all. It is free to expand or contract disqualified and socially excluded; second, that the with changes in social attitudes and therapeutic opti- medical model mistook deviance that was essentially mism and is at the mercy of idiosyncratic decisions defined by social norms as being a matter of medical by doctors or patients. If one wished to compare the fact. In Foucault’s archaeology [3,4], madness was incidence of disease in two different cultures, or in a single population at two different times, whose criteria defined in Western modernity as the mere absence of of suffering or therapeutic concern would one use? reason, coinciding more or less with moral absence, And if the incidence of disease turned out to be concluding that madness had nothing to say, no voice, different in the two, would this be because one was was silenced – to be socially excluded, at a micro- and healthier than the other, or simply because their atti- macro-social level. Broadly similar themes in contem- tudes to illness were different? porary context were explicated in Laing [5]. Mechan- isms of disqualification and exclusion were explored in Kendell identifies the problematic all but perfectly theory and in field studies by sociologists such as here. The consequences of a personal/social definition Goffman [6] and Rosenhan [7]. These historical and of health are very messy: boundaries are subject to sociological critiques identified social processes as

SOCIAL, BIOLOGICAL AND PERSONAL CONSTRUCTIONS OF MENTAL ILLNESS 41 creating the category and reality of madness, mean- It was this proposal that formed the basis of the ingless unreason, incomprehensible chaos much to be definition of mental disorder in the DSM-III, one that feared and avoided. In Foucault’s historical vision was carried forward, with some amendments, to the especially, these processes were not at all instigated DSM-IV. Thus the definition of mental disorder in the by medicine or psychiatry, but on the contrary, medical DSM-IV ([11], pp. xxi–xxii) and also the definition in psychiatry itself came into being to manage and the ICD-10 ([12], p. 5) both have distress and im- endorse them. In any case the upshot was that psy- pairment of functioning as fundamental. Both defini- chiatry as a branch of medicinewas in charge of mental tions state that mental disorder is not the same as social illness – as an object of medical science and with its deviance, that it has to involve ‘personal dysfunction’, legalized powers of confinement. Therefore psychia- this latter notion apparently resting on the generalized try, from this point ofview, stoodaccused – as inSzasz’ nature of impairment, not limited to a particular social influential in-house critique [8] – of misrepresenting context. conditions that were essentially defined by social and This approach to conceptualizing mental disorder is moral norms as instead matters of medical illness. appropriate for medicine and health care generally: These were key themes in the 1960s narrative that the fundamental reality is that people find themselves the category of mental illness – with its characteristic in much distress, or with disability, and they consult a features of meaninglessness, silence and disqualifica- health care professional for help with this; social tion – is socially created, and that psychiatry in deviance, violation of society’s rules and regulations, practice is a form of social control, masquerading as is a different matter, for other agencies and institu- a branch of medicine, pathologizing problems of tions. However, while this construction is correct, it living and social deviance. Here is Kendell again, soon becomes complicated under pressure. Here are identifying the problem ([9], p. 25): three broad kinds of complications: . Some conditions in the psychiatric manuals do The most fundamental issue, and also the most con- tentious one, is whether disease and illness are nor- seem to be a matter of social deviance, albeit mative concepts based on value judgements, or pervasive social deviance. This is explicit most whether they are value-free scientific terms; in other clearly in the diagnostic criteria for the antisocial words, whether they are biomedical terms or socio- conditions. For example, the primary criterion (A) political ones. in the DSM-IV diagnostic criteria for antisocial personality disorder ([11], pp. 649–50) opens with: A particular problem arose for the American Psy- ‘There is a pervasive pattern of disregard for and chiatric Association in the early 1970s, one that violation of the rights of others’, followed by caused the Association to set about the task of defining examples including ‘failure to conform to social ‘mental disorder’: gay activists lobbied to have homo- norms with respect to lawful behaviours’. The sexuality removed from the classification of mental primary problem here seems to be that other people disorders, then DSM-II [10]. The Association estab- or society at large are at risk of harm, rather than lished the Task Force on Nomenclature and Statistics, (as the primary problem) the individual himself. including Robert Spitzer, setting it the task of review- This of course is the kind of case that lends weight ing the controversy and proposing a solution. It was to the view that psychiatry is fundamentally in the concluded that ([10], p. 16): business of defining social deviance under the guise of medical disorder and acting as an agency of social control rather than health care. These The consequences of a condition, and not its etiology, determined whether the condition should issues are not explicitly discussed or explained in be considered a disorder. .. . It was proposed that the manuals. Meanwhile, however, outside of the the criterion for a mental disorder was either sub- calm of the diagnostic manuals, debates on these jective distress or generalized impairment in social issues have accelerated, concerning, for example, effectiveness. the status of severe antisocial personality disorder

42 PERSPECTIVES AND METHODS in law and mental health services, the role of mental disorders among the poor and in ethnic mental health services in the management of risk minority groups reflect social exclusion rather than to the public, and so on. (Literature on these issues disorder [19]. in the UK context include References [13] to [15].) In fact the DSM-IV definition makes what is ap- . Another kind of controversy refers to ‘disability’. parently an attempt to delineate normal distress from Advocacy groups point out that disability is not an pathological, which would solve some but not all of absolute category but is relative to expectations, the difficulties. It requires that mental disorder ‘must environments and task demands typically imposed not be merely anexpectable and culturally sanctioned by majority social groups. This line of thought can response to a particular event, for example, the death be applied to both physical and mental disabilities. of a loved one’ ([11], pp. xxi–xxii, italics added). This Such-and-such a condition would not give rise to italicized qualification appears to intend to rule out disability or distress, it can be argued, if social ‘normal’ or ‘normally caused’ distress and disability, expectations, environments and task demands were but the term ‘expectable’ is vague and uncertain and not incompatible with it, in which case inclusion in cannot bear much weight. Many medical and psychia- the psychiatric manuals as a mental disorder is tric conditions are ‘expectable’ in any readily avail- actually an illegitimate pathologizing and negative able sense, such as skeletal fractures following a evaluation of a difference, acover for discrimination (long) fall or stress reaction following profound psy- and social exclusion [16]. The 1970s controversy chological shock. What matters is apparently not about whether homosexuality was or was not a ‘expectability’ or otherwise, but consequences in mental disorder was one expression of this general terms of distress, disability and risk thereof. kind of issue. Contemporary examples include, for All these various controversies listed above interro- example, controversy about ADHD: whether this is gatethecategoryofmentalillnessormentaldisorderand real illness, to be treated by medication, or rather thelegitimatedomainofmentalhealthservices,echoing temperamentally based high-activity levels in problems identified in the 1960s. The definitions of (mainly) boys causing difficulties because of so- mental disorder in the psychiatric manuals were for- cially inappropriate expectations in, for example, mulated to clarify and legitimize, but, as is hardly school settings (e.g. see Reference [17]). surprising on reflection, definitions cannot possibly be expected to resolve all the complex and multifaceted . A third kind of problem with grounding the con- issues involved. The definitions serve as position state- ceptualization of mental disorder in distress and/or ments and are workable for most purposes, particularly disability has to do with distinguishing disorder in helping define perceived need to treat, allowing the distress from normal, nonpathological distress. It is manuals to get onwiththeir main task ofdescribingand obvious enough that distress may be unconnected classifying the conditions of interest as clearly as pos- to mental disorder. Other kindsorcausesofdistress sible,optimizingreliabilityofdiagnosisforthepurposes include, for example, distress due to normal life of research and communication generally. transitions, challenges and losses, in education The business of resolving the problems in concep- and work, family life, relationships, ageing, and tualizing mental illness, distinguishing this category so on; and distress associated with social depriva- from what is normal, on the one hand, and from social tion, exclusion or persecution. Given psychiatry’s deviance, on the other, has been taken up outside the understandable tendency to regard distress as fun- diagnostic manuals, mainly by psychiatrists and philo- damental to its domain, there is the inevitable risk sophers. I have reviewed this literature, covering at of its spreading into kinds of suffering that are greater length issues addressed in the present chapter, apparently not a matter of mental disorder. For in a recent book [20]. As indicated above, the main example, there is concern that normal life suffering conceptualalternativetoasocialdefinitionofillnessand is being pathologized as depressive disorder [18] health invokes naturalism, and since the 1970s this has and that typically higher rates of diagnosable been based mainly on ideas from evolutionary theory.

SOCIAL, BIOLOGICAL AND PERSONAL CONSTRUCTIONS OF MENTAL ILLNESS 43 4.2 THE INTERPLAY OF EVOLUTIONARY, SOCIAL AND PERSONAL NORMS In a series of highly influential papers beginning in the 1. Evolutionary natural selection early 1990s Wakefield has proposed that the proper definition of mental disorder is as harmful dysfunc- 2. Socialization processes (education, training, cul- tion. He argued that this is a plausible reading of what turalization) is intended in the DSM-IV [21] and that – this being the crucial claim – dysfunction is properly understood 3. Individual choice (signalled by individual differ- along the lines of some mental or behavioural me- ences, notwithstanding 1 and 2). chanism not functioning as ‘designed’, selected for, in evolution [22]. There has been a steady flow of The first of these is of course the one emphasized by literature produced over the past 15 years in which naturalism (and is presumably a matter for the ‘natural Wakefield’s proposal has been (usually) criticized in sciences’), the second is emphasized by the social many and various ways, and in which Wakefield has sciences and the third is emphasized by psychology. vigorously defended his position (e.g. see Refer- Relations between these three factors in human be- ences [23] to [37]). haviour are complicated and contentious. The point of Wakefield’s evolutionary theoretical approach to naturalism, however, is to construe mental and beha- defining mental disorder accepts that there is a socially vioural mechanisms only in the first way, excluding embedded evaluative component in the concept, in- the second, and then the third will not appear as very dicated by ‘harmful’, but insists that there is another relevant, since the main point is to distinguish the first component that is independent of social evaluations, from the second. which is a matter of fact, namely the reference to Here is Wakefield in one of his early papers ([22], mental and behavioural mechanism not functioning p. 381): ‘as designed’ (as selected for) in the evolutionary What, then, is a dysfunction? An obvious place to process. This appeal to facts of the matter independent begin is the supposition that a dysfunction implies an of socialevaluations is just what makes this approach a unfulfilled function, that is, a failure of some mechan- form of naturalism, intended in this case not so much ism in the organism to perform its function. However, as an alternative to a social construction of health and not all kinds of functions are relevant. For example, disease but at least as a crucial, principled limitation to one’s nose functions to hold up one’s glasses, and the it, based in an actual fact of the matter concerning sound of the heart performs a useful function in what is and what is not functioning as designed medical diagnosis. But a person whose nose is shaped (selected for) in evolution. in such a way that it does not properly support glasses So does this work? It is correct to suppose that does not thereby have a nasal disorder, and a person concepts of function and dysfunction refer to systemic whose heart does not make the usual sounds is not design and that evolutionary natural selection is a very thereby suffering from a cardiac disorder. A disorder particular theory about design, managing as it does is different from a failure to function in a socially preferred manner precisely because a dysfunction without a designer. There are, however, and of course, exists only when an organ cannot perform as it is other ways in which systems are designed, namely by naturally (i.e., independently of human intentions) individuals or groups making artefacts. By all means supposed to perform. Presumably, the functions that this has not much relevance to biological systems such are relevant are natural functions (italics added). as the cardiovascular system, which we may suppose has been designed, selected for, in evolution, without In this example of the function of the nosewe have a input from culture or individuals. However, in the case clear fix on the evolved function (a kind of sense of mental life and behaviour, who or what is respon- perception), a clear fix on an additional cultural sible for design and function? Here are three options: function that is entirely unrelated, and ‘design’ of the

44 PERSPECTIVES AND METHODS nose is pulled only one way. It is all much more isolation from social selection – accounts for the complex, however, when we move from glasses and class of psychological functions of interest. These noses to social behaviours that are highly evolved, invalidities compound in the conclusion, which pos- such as infant and child care, and to socially inter- tulates failure in hard-to-track-down natural, evolved preted evolved behaviours such as fear, whether in psychological functions, in which social factors social or natural environments. In these examples the play no role. clarity in identification of what is the evolved function Modification of the argument leading to Wakefield’s and what is the socially designed function, and the conclusion would replace the second premise. The sharp demarcation between them, are lost. In these origin and design of psychological functioning complex cases, which in the realm of the psycholo- typically include a complex mixture of genetic, gical are probably the great majority, the biological evolved factors and social factors, with individual evolved mechanism – though it does have content (e.g. differences running though them both. Depending anxiety is about threat perception) – is realized in a on which is dominant, or which is thought to be concrete social environment, and if the mechanism is dominant, we can attribute the origin – the design – failing, so is the concrete social norm. Conversely, if of the behaviour to human nature, to society, to no social norm is being broken, the mechanism is not subculture, to family (to family genes or behaviour broken either. or both) or to the individual’s constitution, character Wakefield’s analysis rests on a line of reasoning or personal values. In brief, according to current something like the following: behavioural science in an evolutionary/genetic fra- mework, three kinds of factor are implicated in 1. Concepts of systemic function and dysfunction and the design of human behaviour – evolutionary/ the distinction between the two essentially refer to genetic, environmental, including the cultural, and the design of the system: normal functioning is individual – and these three kinds of factor interact functioning as designed; abnormal functioning is in complex ways. To each kind of design there not functioning as designed. corresponds a type of norm: evolutionary/genetic, social and individual – again with no clear divisions 2. In the case of psychological functions, we are only and interplay between them. interested in those that are naturally designed (as It follows that a mental state or behavioural re- opposed to socially designed). sponse can be said to be dysfunctional – to deviate from design norms – in one or more of three ways: 3. In the case of natural psychological functions, the first, in that it fails to operate in the way designed by best scientific account of design that we have is evolution; second, in that it fails to operate in the way evolutionary selection. taught by and sanctioned by the culture; and, third, in that it fails to work in the way the person intended, 4. Hence, psychological dysfunction must involve according to his/her needs and values as he/she sees failure to function as designed by/selected for in them. However, these three kinds of dysfunction are evolution. not clearly separated, and they interact. The first kind belongs to an evolutionary theoretic framework and is The first premise is fine, or good enough, and so is relative to conditions in the Environment of Evolu- the inference to the conclusion from the three pre- tionary Adaptiveness. The second kind of dysfunction mises. The fault lies in the second premise, which is the one accessible to social theory; it is immersed in transfers to the third, and hence to the conclusion. The the present, in more or less diverse social realities. problem with the second premise is that it assumes There is also a third reading of dysfunctional psychic an invalid demarcation between natural and social life, the one at the individual level involving deviation psychological functioning. This invalid assumption from personal norms and values, evident to the person affects the third premise, because this now assumes involved. This meaning has been neglected, to do with what is also invalid, that evolutionary selection – in the fact that ‘madness’ was silenced – though it is

SOCIAL, BIOLOGICAL AND PERSONAL CONSTRUCTIONS OF MENTAL ILLNESS 45 increasingly apparent in discourse led by service tionary, the social and the individual – but are rather users. These are not, however, three meanings of three interwoven themes that run through all kinds psychological function and dysfunction – the evolu- of case. 4.3 IMPLICATIONS FOR THE PROPER DOMAIN OF HEALTH CARE Naturalism represents the thought that the proper eventually the new profession of psychiatry. In this domain of medicine and health care is defined by a story madness and psychiatry were constructed to- boundary in nature, in human nature, enclosing those gether, neither being the business of society at large. In states of the body/mind that are pathological, which brief, no one understands madness, nor psychiatry deviate from the natural way that things should be. either, and this is a particular reason why the reality of ‘Natural’ has to be used in this characterization, the clinical encounter, the distress and disability in- because otherwise it looks as though ‘the way things volved, while recognized as of high importance, has should be’ is going to have to have a social meaning, tended to remain unspoken and unheard. Of course, variable, arbitrary, fading into indelicacy in one direc- much water has flowed under these particular bridges tion and antisocial behaviour in another. Pathology – in the past few decades, including the linked processes so this line of thought runs – has to be something really of closing of the asylums, care in the community, wrong with the body/mind, regardless of social mores, steady growth of service user movements, and psy- that medical sciences aim to identify and doctors help chiatry and mental health care services openly tack- treat. In the sociological view of the matter, on the ling issues of what it can and cannot do, should and other hand, this appears as a working assumption of should not be asked to do. health care, more or less justifiable, somehow or other, The clearest position for mental health services to in any case necessarily self-serving, while in practice occupy at present seems to me to be this: they offer mental health services in particular continue miscon- treatment, based on the available science of body and struing social ills as pathology and managing social mind, to the person who comes to the clinic for help in deviance of one kind or another. The pieces have been distress, or better, distress they find unmanageable, stuck on the board like this for a while. and disability, in their terms. This is the core business There is, however, one clear reality in the middle of of mental health care, and, characterized in just this all this, which is the person in distress, or better, way, no one else’s. Social interventions at the national distress they find unmanageable, and disability, in political level or local community level may help their terms, who comes to the clinic for help. This communities as a whole and some individuals in reality is so obvious to the clinician that it is hardly them, typically over the medium to long term. They worth mentioning, except in exasperation that it has to are – obviously – not targeted on particular indivi- be. On the other hand, what comes into focus in the duals. Who is there for the individual? This is a sociological gaze is quite different: forms of social particular problem when social support structures organization, distribution of and access to resources, have done what they can, which is more or less, and and so on. What does not appear in the sociological the persons themselves have done what they can, and view is the individual’s inner life, this being however they are still in unmanageable distress and possibly the primary concern of the clinician. There are also unable to carry on with some aspect of their lives, specific issues here, not just to do with the general which they regard as crucial, child care or work for differences between socially and individually orien- example. Where then? Health care is one abiding tated approaches, issues specifically about mental option – it always has been and no doubt always will illness of the sort identified by Foucault. In Foucault’s be. Health care, dealing in constructs of illness and history, madness was constructed as incomprehensi- health, is defined by three essential features: one is ble danger, excluded and in need of control, gradually distress/disability of the person coming for help, the coming under the care of doctors and medical science, second is provision of help based on the available

46 PERSPECTIVES AND METHODS science of body/mind and the third is ethics, such as and clinical psychologists [44]. These controversies confidentiality, duty of care and cause no (more) harm. about the nature and proper domain of mental health The other profession that aims to help individuals in care are inevitable and good topics for ongoing ex- self-identified serious trouble is the pastoral priest- amination, just the upshot of there being no domain of hood, the contrast being that in this meeting the trouble (current mainstream biomedical) medicine carved out is interpreted as in the spiritual domain, suited to the of nature’s joints, still less for psychiatry, which is special knowledge and skills of the priest. Health care thoroughly involved in the social and moral and is an abiding option, so is spiritual guidance and so is personal world where norms and values are diverse. community care (care by the community). The extent However, notwithstanding and in the midst of these and boundaries and interplay between these various ongoing negotiations, the domain of mental health responses to distress vary and are always in flux, care is a permanent fixture, as one mode of response to influenced by available social representations, perso- individuals in unmanageable trouble. nal narratives, distribution and provision of resources Complicated though all this is, it is familiar enough and by more or less well-evidenced beliefs about what and manageable within a traditional view of medicine is the best way of helping what. and health care. Doctors treat patients in trouble using All of this is of course very messy, in the ways available medical science, and other groups advertise Kendell excellently identified in his 1975 paper [2], help too, nonmedical or alternative medical. A much quoted above. The boundaries of illness fluctuate with more difficult problem arises when the would-be the idiosyncrasies of would-be patients, medical en- patient does not identify him/herself as having a thusiasm and the advertising power of manufacturers problem, or not one that has anything to do with of treatment technologies; and they vary across cul- his/her state of body or mind, not one that has anything tures and subcultures. The hope of avoiding all this – to do with doctors, psychologists or any such thing. typically post- or late-modern – mess is of course the This is the ‘involuntary patient’, something of an attraction of naturalism, which is why Kendell in his oxymoron according to the traditional and comforta- 1975 paper [2] went on to turn to it, in its evolutionary ble conceptualization of health care. Interestingly, the theoretic form, later consolidated by Wakefield. It theories of medical naturalism, appealing to such as certainly would all be clearer if therewas some natural lesions, diseases or broken evolutionary mechanisms, fact about the body/mind which meant that it was can readily envisage that an individual is disordered or dysfunctional, disordered, quite apart from any social ill, quite regardless of whether they think they are so or theories about it, the legitimate domain for medical not. However, if naturalism is given up, because science and doctors. Evolutionary naturalism tried to unviable, and emphasis is put instead on harm as underpin this intuition, but does not work because experienced by the individual patient, then so far it psychological phenotypes do not cleave apart into is not obvious how someone who says they are OK, or natural/evolved functions, on the one hand, and so- in any case not in need of treatment, is any business of cialized functions, on the other. The influences on the health care professional. The reluctant patient of functioning are muddled up together. This leaves us course becomes the business of the health care profes- defining the primary feature of illness in terms of sional typically (though not invariably) because some- harm, but then medicine and especially psychiatry one else brings them or sends them to the clinic, not inevitably gets involved in debates with other experts because the person themselves is (necessarily) experi- about what are the causes of harm and what are the encing suffering or disability, but because of actual appropriate preventions or remedies, debates about harm or risk of harm to others. In these contexts the boundaries of health care with libertarians (e.g. see mental health services become involved in protec- References [38] and [39]), with social scientists [19], tion of the public from danger, in matters of public about the boundaries between illness and the spiritual safety, which is more obviously a function of state dimension [40,41], and in-house health care debates agencies for preserving public order. In its role in about the boundaries of medical approaches with compulsory detention and possibly treatment mental criticalandpost-psychiatrists[42],social workers[43] health services look like an agency of social control,

SOCIAL, BIOLOGICAL AND PERSONAL CONSTRUCTIONS OF MENTAL ILLNESS 47 a medicalized police force, and this is the role mechanisms – it is really no surprise at all, not an emphasized by the sociological and libertarian embarrassment, but a perspective to be welcomed, that critiques that charge psychiatry with being just that. distress and disability is not always just a matter for the A simple idealized solution would look something individual but involves the interests of families, like this. Health care professionals attend to willing friends, the community and, if and when they are patients in self-identified trouble who believe or who involved, the values of health care professionals. All are willing to try out the belief that the professional’s this just reflects our interrelatedness. I have spent expertise in matters of the body/mind can help them. most of my professional life as a clinician working The state and the police in particular deal with danger in child and adolescent services, where identification to the public. Perhaps the kind of case that best fits this of distress, disability, risk and need to treat are some- simple picture is that of the antisocial conditions, times matters of protracted investigation and negotia- which would be and mainly are directed (while they tioninvolvingmothers,fathers,grandparents,teachers, are the only problem) unambiguously to the judicial social workers, the child or adolescent brought along system, police and probation agencies, not to health and the peer group in absentia. The negotiation of care. However, in many kinds of case risk of harm to perspectives, stakeholding, values as to what is dis- others and risk of harm to self are co-present, e.g. in an tress, disability, risk and need to treat is all part of the antisocial personality with suicidal tendencies or in work. I assume adult mental health services involve some acute psychotic states, complicating the simple much the same, reflecting the socially and interper- picture. sonally embedded nature of illness and health care. Another kind of complication is that harm to self is judged not only by the identified patient but by others involved. There are the cases, for example, of psy- REFERENCES chotic episodes in which the person apparently pre- sents some risk to others, and often also to him/herself, 1. Jaspers, K. (1913) Allgemeine Pychopathologie, though he/she has no wish to be in hospital, is perhaps Springer Verlag, Berlin (3rd enlarged and revised terrified of it, terrified altogether. All cases of com- edition, 1923). English translation of the 7th edition pulsory admission involve by definition some over- by J. Hoenig and Marian W. Hamilton, General ruling of the identified patient’s own current will, but Psychopathology, Manchester University Press, the intention is that this is in the patient’s best interest, Manchester, 1963; Chicago University Press, possibly as well for the protection of others, and is thus Chicago, Illinois, 1963. Quotation with some revised translation from and page reference to this volume. intended to be consistent with, an extension of, the traditional model of health care. This of course is a 2. Kendell, R. E. (1975) The concept of disease and its implications for psychiatry. British Journal of contentious and well-trodden area, and there is just Psychiatry, 127, 305–315. one point to be made about it in the context of the 3. Foucault, M. (1965) Madness and Civilisation: A themes pursued in this chapter. The proper domain of History of Insanity in the Age of Reason, translated health care, if naturalism is givenup as unviable, has to by R. Howardof an abridged version of Folie et be understood in terms of distress that the person finds D eraison. Histoire de la Folie  a l’Age Classique ˆ unmanageable, a self-identified disability, and a doc- (Librairie Plon, Paris, 1961), Tavistock, London; tor, a mental health care professional, providing help reprinted Routledge, London, 1997. based on the available science of body and mind. This 4. Foucault, M. (2006) History of Madness, English is an entirely socially embedded understanding of translation of Histoire de la Folie  a l’Age Classique ˆ illness and health care, though there is, by the way, (Gallimard, London, 1972), Routledge, Paris. all the room and role needed for the science, inves- 5. Laing, R. D. (1960) The Divided Self, Penguin, tigating causes, prevention strategies and treatments. Harmondsworth. In this socially embedded view of illness and health 6. Goffman, E. (1961) Asylums, in Essays on the Social care – which is some conceptual distance away from Situation of Mental Patients and Other Inmates, hypothetical lesions and diseases and broken mental Penguin, Harmondsworth.

48 PERSPECTIVES AND METHODS 7. Rosenhan, R. (1973) On being sane in insane places. 20. Bolton, D. (2008) What is mental disorder? An Essay in Science, 179, 251–258. Philosophy, Science and Values, Oxford University 8. Szasz, T. (1961) The Myth of Mental Illness: Press, Oxford. Foundations of a Theory of Personal Conduct,Harper 21. Wakefield, J. C. (1992) Disorder as a harmful & Row, New York. dysfunction: a conceptual critique of DSM-III-R’s 9. Kendell, R. (1986) What are mental disorders? in Issues definition of mental disorder. Psychological Review, in Psychiatric Classification: Science, Practice, and 99, 232–247. Social Policy (eds A. M. Freedman, R. Brotman, I. 22. Wakefield, J. C. (1992) The concept of mental disorder: Silverman and D. Hutson), Human Sciences Press, on the boundary between biological facts and social New York, pp. 23–45. values. American Psychologist, 47, 373–388. 10. Spitzer, R. L. and Williams, J. B. (1982) The definition 23. Kirmayer, L. J. (1994) Rejoinder to Professor and diagnosis of mental disorder, in Deviance and Wakefield, Debate on ‘Is the concept of mental Mental Illness (ed. W. R. Grove), Sage, Beverly Hills, disorder culturally relative?’, with J. C. Wakefield, in California, pp. 15–31. Controversial Issues in Mental Health (eds S. A. Kirk 11. American Psychiatric Association (1994) Diagnostic and S. D. Einbinder), Allyn & Bacon, Boston and and Statistical Manual of Mental Disorders,4th London, pp. 17–20. edn, American Psychiatric Association, Washington, 24. Sadler, J.Z. and Agich,G. J. (1995)Diseases, functions, DC. values, and psychiatric classification. Philosophy, 12. World Health Organization (WHO) (1992) The ICD-10 Psychiatry, and Psychology, 2, 219–231. Classification of Mental and Behavioural Disorders: 25. Wakefield, J. (1995) Dysfunction as a value-free Clinical Descriptions and Diagnostic Guidelines, concept. Philosophy, Psychiatry, and Psychology, 2, World Health Organization, Division of Mental 233–246. Health, Geneva. 26. Fulford, K. W. M. (1999) Nine variations and a coda on 13. Kent County Council (2006) Report of the the theme of an evolutionary definition of dysfunction. independent inquiry into the care and treatment of Journal of Abnormal Psychology, 108, 412–420. Michael Stone, http://www.kent.gov.uk/publications/ 27. Cosmides, L. and Tooby, J. (1999) Toward an evolu- council-and-democracy/michael-stone.htm. tionary taxonomy of treatable conditions. Journal of 14. Thornicroft, G. and Szmukler, G. (2005) The draft Abnormal Psychology, 108, 453–464. Mental Health Bill in England: without principles. 28. Richters, J. E. and Hinshaw, S. P. (1999) The abduction Psychiatric Bulletin, 29, 244–247. of disorder in psychiatry. Journal of Abnormal 15. Lepping, P. (2007) Ethical analysis of the changes Psychology, 108, 438–445. proposed to mental health legislation in England and 29. Wakefield, J. C. (1999) Mental disorder as a black box Wales.Philosophy, Ethics,and Humanities in Medicine, essentialist concept. Journal of Abnormal Psychology, 2, 5, http://www.peh-med.com/content/2/1/5. Accessed 108, 465–472. 17 May 2007. 30. Murphy, D. and Woolfolk, R. L. (2000) The harmful 16. Oliver, M. (1990) The individual and social models of dysfunction analysis of mental disorder. Philosophy, disability, http://www.leeds.ac.uk/disability-studies/ Psychiatry, and Psychology, 7, 241–252. archiveuk/Oliver/in%20soc%20dis.pdf. 31. Wakefield, J. C. (2000) Spandrels, vestigial organs, and 17. Timimi, S. (2002) Pathological Child Psychiatry and such: reply to Murphy and Woolfolk’s ‘The harmful the Medicalization of Childhood, Brunner-Routledge, dysfunction analysis of mental disorder’. Philosophy, Hove. Psychiatry, and Psychology, 7, 253–269. 18. Horwitz, A. V. and Wakefield, J. C. (2007) The Loss of 32. Houts, A. C. (2001) The DSM’s new white coat and Sadness: How Psychiatry Transformed Normal Sorrow circularity of plausible dysfunctions: response to into Depressive Disorder, Oxford University Press, Wakefield, Part 1. Behaviour Research and Therapy, Oxford. 39, 315–345. 19. Kutchins, H. and Kirk, S. A. (1997) Making us crazy, in 33. Wakefield, J. C. (2003) Dysfunction as a factual DSM – The Psychiatric Bible and the Creation of component of disorder: reply to Houts, Part 2. Mental Disorders, Free Press, New York. Behavior Research and Therapy, 41 (8), 969–990.

SOCIAL, BIOLOGICAL AND PERSONAL CONSTRUCTIONS OF MENTAL ILLNESS 49 34. Wakefield, J. (2007) The concept of mental disorder: 40. Fulford, W. and Sadler, J. (in press) Delusion and diagnostic implications of the harmful dysfunction spiritual experience in the DSM: philosophical aspects analysis. World Psychiatry, 6, 149–156. of the concept of disorder, in Religious and Spiritual 35. Sartorius, N, (2007) A new way of reducing the Issues in Psychiatric Diagnosis: Considerations for a prevalence of mental disorders? World Psychiatry, 6, Revised DSM (eds J. R. Peteet, F. Lu and W. Narrow), 162–163. American Psychiatric Publishing, Inc. 36. Bolton, D. (2007) The usefulness of Wakefield’s 41. Bolton, D. (in press) Commentary on Fulford and definition for the diagnostic manuals. World Sadler’s ‘Delusion and spiritual experience in the Psychiatry, 6, 164–165. DSM’, in Religious and Spiritual Issues in Psychiatric Diagnosis: Considerations for a Revised 37. Gold, I. and Kirmayer, L. J. (2007) Cultural psychiatry DSM (eds J. R. Peteet, F. Lu and W. Narrow), American on Wakefield’s procrustean bed. World Psychiatry, 6, Psychiatric Publishing, Inc. 165–166. 42. Bracken, P. and Thomas, P. (2005) Post Psychiatry: 38. Szasz, T. (2006) Secular humanism and scientific psychiatry. Philosophy, Ethics, and Humanities Mental Health in the Post-Modern World, Oxford University Press, Oxford. in Medicine, 1, 5, http://www.peh-med.com/content/ 1/1/5. 43. Kirk, S. A., Wakefield, J., Hseih, D. and Pottick, K. (1999) Social context and social workers’ judgement of 39. Bolton, D. (2006) What’s the problem? A response to mental disorder. Social Service Review, 73, 82–104. ‘secular humanism and scientific psychiatry’. Philosophy, Ethics and Humanities in Medicine, 1,6, 44. Follette, W. C. (1996) Introduction to the special http://www.peh-med.com/content/1/1/6. Reprinted in section on the development of theoretically coherent Science and Ethics (ed. Paul Kurtz), Prometheus alternatives to the DSM system. Journal of Consulting Books, New York, 2007, pp. 204–207. and Clinical Psychology, 64, 1117–1119.



5 Social science perspectives: a failure of the sociological imagination 1 Craig Morgan and Arthur Kleinman 2 1 NIH Biomedical Research Centre and Institute of Psychiatry, King's College London, London, UK 2 Department of Anthropology, Harvard University, Cambridge, Massachusetts, USA The clustering of substance abuse, street violence, domestic violence, suicide, depression, post-traumatic stress disorder,.. . amongpeople living in disintegrating communities runs against theprofessional medical idea thatsufferers experience one ... major problem at a time. That grouping of human problems also defeats categorisation of such issues as principally psychological or medical and, therefore, individual (Kleinman et al., [1] p. ix) The primary purpose of this chapter is to introduce the perspective that mental disorder, in all its mani- and discuss select social science approaches to the festations, is a form of social suffering [2], a form of study of various aspects of mental disorder (primarily distress that is intimately entwined with other social drawn from sociology and, to a lesser extent, anthro- problems that both cause and result from experiences 1 pology ), in particular reflecting on what these have of emotional and mental distress. The emphasis on contributed to our understanding of mental disorder suffering as a collective, socially embedded experi- and on some key limitations. The broad areas we ence challenges the construction of mental disorder as discuss include: the social construction of disorder individual, as a set of discrete disorders requiring (especially labelling theory), the social determinants technical responses (be these psychological or phar- of disorder, and the social and cultural influences on macological) of ever-increasing precision. What is responses to disorder. More ambitiously, these more, it does this without negating or trivializing the approaches and their limitations are considered from distress that is experienced. 1 There is considerable overlap in the perspectives and methods of sociology and social anthropology. In this chapter, when we use the term ‘sociology’, we are indicating not only the discipline of sociology but also social anthropology. Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

52 PERSPECTIVES AND METHODS 5.1 SOCIAL SUFFERING As will be discussed below, the characteristic sub- Consider the following example. At the time of his jective phenomena of mental disorder (e.g. sadness, first contact with mental health services, James was anxiety, hallucinations) occur, at the very least in part, 33, unemployed and had lived alone for the past 14 3 as a consequence of troubling and traumatic social years in a small flat. He first began to hear voices experiences; they emerge in, and are shaped by, both when he was around 16 or 17 years old, or perhaps local social worlds and wider social and economic earlier. At that time, he was living with his father and contexts. In this sense, the experience of mental brother. He describes an extremely unhappy and disorder is fundamentally social [3]. It occurs and is abusive childhood living in a chronically unstable located within a particular social space, a complex household. He had a poor relationship with his father nexus of social relationships and social structures that and his brother, largely because of this, and he left mould and transform that experience. The experience school early with no qualifications. He never told his is not simply one of intrapsychic conflict or distress, father or brother what he was experiencing, despite of existential angst; it is frequently one of broken being distressed and angry, and says that he did not relationships, lost opportunities, social failures and discuss what was happening with anyone else for fear hostility; of stigma, grinding poverty and chronic they would think he was ‘crazy’. Only once did he isolation. Others (especially friends and family) are attempt to talk to someone about his experiences, a almost always involved (at least to begin with), and friend: ‘I did discuss it with my friend I did, but he much of the suffering occurs because the very things thought I was going crazy so I thought I’d leave it, you that come to matter most are threatened – jobs, know don’t discuss it to anyone.’ Since being forced to housing, social and family relationships, future possi- move out of his father’s home when he was 19, and bilities. These are threatened not just by the impact following a brief period of homelessness, he has lived of inner distress on the abilities of individuals to alone in a one-bed local authority flat, continued to assume and sustain these social roles but also by the hear voices and has had minimal contact with others. perceptions and responses of others in the wider Over time, James became increasingly isolated and society. In short, mental disorder is not experienced eventually would leave his flat only to get food and as a discrete problem, standing outside the flow of collect his welfare benefits. He would spend his days everyday life. It is part of an entangled set of circum- watching television and smoking cannabis. The con- stances. This is not to deny subjective experience; it dition of his flat deteriorated and his living circum- is certainly not to deny the reality of mental distress. It stances worsened. He had no furniture: ‘... there’s no is to recognize that experience is simultaneously carpet, no decorating or no chairs to eat your dinner subjective and collective. In other words, mental on; I used to eat my dinner on the fridge and the 2 disorder is a form of social suffering, intimately fridge’s broken, the door on the front of the fridge was entangled with the context within which it occurs, broken ....’ He describes at times being so hungry that both cause and consequence of interlinked problems he would shake, making it impossible for him to leave that may include poverty, social isolation, victimiza- his flat to get money or food. He eventually contacted tion and trauma. This clustering of problems, for his local housing officer to carry out repairs in the flat: example, is well illustrated in the World Mental ‘... I phoned because there was a hole in the sitting Health Report [4,5]. room and he come to have a look ... and he’d get 2 Kleinman et al. [1] provide the following broad definition: ‘Social suffering .. . brings into a single space an assemblage of human problems that have their origins and consequences in the devastating injuries that social force can inflict on human experience.’ (p. ix). 3 James’ story is from a study of narratives of the experience of mental disorder collected as part of a large study of psychosis conducted in the UK. The name has been changed, and details amended, to ensure anonymity.

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 53 someone to fix it, that happened ... because I had my (ICD), incorporates a set of assumptions about under- bed in the sitting room and when I sit on it the legs lying aetiologies and treatment approaches that pri- would go through the floor ....’ It was the housing vileges the individual, the biological and the technical. officer who, concerned about James’s welfare, con- This is part of a broader trend to strip all forms of tacted social services and arranged for a home visit. suffering from the social context within which it Initially reluctant to go into hospital, James eventually occurs, to individualize and medicalize, to provide agreed. When asked why, he replied: ‘I don’t know, I quick-fix cures in the form of pharmacological agents just felt like going out somewhere you know.’ What is and time-limited psychotherapy [6–8]. From a social striking about how James tells his story is that he science perspective, however, it is inadequate to see focuses primarily on his isolation, his lack of money, disease (be it mental or physical) or other forms of his hunger and the appalling state of his accommoda- distress as affecting a single person. There is a need to tion. The experience of hearing voices is one facet of a extend outwards into the social space – it is here that series of interconnected difficulties that James suffering occurs, is understood and managed. As the endures; fear of how others will respond compounds quotation at the head of this chapter suggests [1], his suffering and acts as a powerful barrier to confiding suffering rarely separates into discrete boxes, and its and help-seeking. causes and effects are rarely limited to individuals – as To emphasize the social, collective experience of much of the work reviewed below shows. This work, suffering – and its interconnectedness with other however, is theoretically disparate and just as the experiences in the flow of everyday life – presents a formulation of mental disorder as a form of social challenge to the predominant view in psychiatry of suffering provides a basis from which to critique mental disorders as discrete diseases. This perspec- psychiatry, so it provides for a critical reflection on tive, codified in the Diagnostic and Statistical Manual some of the more prominent social science theories of, (DSM) and International Classification of Diseases and approaches to, mental disorder. 5.2 THE SOCIOLOGICAL IMAGINATION In considering social science perspectives on mental norms. This, in Durkheim’s formulation, simulta- disorder, it is worth remembering that the original neously provides an explanation for a troubling social scope of sociology was broad and ambitious. The phenomenon and resonates more widely, revealing a founding sociological theorists (e.g. Comte [9], fundamental principle governing social life. Other Marx [10], Durkheim [11] and Weber [12]) were examples abound, albeit often based on different concerned with fundamental questions about the nat- ontologies and epistemologies. Parsons [14], for ex- ure of society, its failings and necessary modifications ample, used sickness as an exemplar to reveal how (or in Marx’s case, root and branch revolution). The societies maintain stability and order through tightly particular phenomena and problems that became the governed social roles, with rights, obligations focus of study (e.g. class, income distribution, health, and sanctions for breaches of these moral codes. suicide, social integration, religion, institutions) were Foucault [15,16] turned his gaze on psychiatry, criti- tools for revealing the workings of society and for quing it as a product of the enlightenment designed to understanding how social structures and social rela- police the boundary between normality and abnorm- tions impacted on individuals to shape and constrain ality – as a mechanism for silencing difference and life choices and opportunities. Perhaps the most per- unreason. In short, what drove the sociological tradi- tinent example is Durkheim’s classic study of sui- tion was a concern with significant questions of social cide [13], in which the unequal patterning of this order and change, and of how these became manifest deeply personal act among social groups was ac- in the lives of individuals. It is this coupling of the counted for by social structure, by the degree to which study of lived experience and broader social forces social groups were bound together by collective that Wright-Mills [17] famously christened ‘the

54 PERSPECTIVES AND METHODS sociological imagination’. Interest in mental disorder interested both (1) in how social structure and experi- and related institutions stemmed in part from an ence impacted on individuals to cause, exacerbate and understanding that these provide a lens through which prolong mental distress and (2) in the institutions elementary principles governing social life and orga- (psychiatry, asylums, diagnostic systems) that society nization may be revealed. From the outset sociologists constructed to manage the problems that came to be and other social scientists were consequently subsumed under the rubric of mental disorder. 5.3 A LEGACY OF MISTRUST The relationship between the social sciences and This was the underlying perspective that united a psychiatry has fluctuated over time, with some series of searing critiques of psychiatrythat emerged in periods of fruitful collaboration and others of the 1960s and 1970s, notably by Foucault [15,16], outright hostility. In the early classic studies of Szasz [24], Laing and Esterson [25] and Scheff [26] the social distribution and determinants of mental (not all social scientists). In short, mental illness was disorder conducted in the 1930s, 1940s and seen as a social construction, a myth built on false 1950s [18–22], sociologists and psychiatrists colla- analogy with physical illness; psychiatry was society’s borated with an ease that has perhaps not been seen stooge, an agent of social control, policing the socially since. At the same time, parallel perspectives devel- constructed dividing line between deviance and nor- oped within the social sciences and eventually con- mality, silencing difference and unreason. This simul- verged in a series of fundamental challenges to taneously provides a critique of psychiatry and its psychiatry and the understanding of mental disorder models of mental illness and illustrates more funda- it embodies. In the early 1930s (while Faris and mental issues relevant to how we create knowledge Dunham [18] were conducting their seminal studies and seek to understand and act on the social world. of the distribution of mental disorders across Psychiatry’s response was both a reassertion of the Chicago – see below), Ruth Benedict [23] published legitimacy of its approach to the understanding and her essay ‘Anthropology and the abnormal’, the treatment of mental disorder and a counterattack primary thesis of which centred around the observa- accusing its critics of being unscientific and engaging tion that in many cultures the behaviours and ex- in unfounded theorizing [27,28]. The charge of periences classified as abnormal by Western psy- unfounded theorizing also partly stemmed from chiatry – paranoia, seizures, trances – were consid- certain sociological theories of the origins of emo- ered normal. In other words, the boundary between tional distress, notably Bateson’s double-bind [29] and normal and abnormal was culturally relative, there Fromm-Reichman’s‘schizophrenogenicmother’[30]. were no absolute distinctions: as Benedict argued, Towards the end of the 1970s, Eisenberg [31] com- ‘all our local conventions of moral behaviour and of mented that the gap between psychiatry and the social immoral are without absolute validity’. This basic sciences was almost unbridgeable. The legacy of these premise, of the cultural relativity of abnormality, debates is an enduring undercurrent of mistrust, in was subsequently developed further and expanded, particular a mistrust (on the part of psychiatry) of as sociologists and others began to consider how, if sociological theory and in-depth methodologies (e.g. knowledge had no absolute basis, societies came to ethnography, qualitative interviews) and (on the part of accept certain versions of ‘truth’. sociology) of the power and social role of psychiatry. 5.4 SOCIAL CREATION OF MENTAL DISORDER The most enduring sociological expression of this individuals but a social construction – is Thomas critique – that mental disorder is not intrinsic to Scheff’s application of labelling theories of deviance

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 55 to mental disorder [26]. The basic premise, that de- fulfilling the stereotypical expectations of the label- viance is not intrinsic to particular behaviours and acts lers. Mental disorder is not ‘in’ the person, it is created but is created by social groups who label acts as by society ([35], p. 136). deviant [32], was extended by Scheff to mental illness. These broad perspectives have been extensively Symptoms were reframed as rule or norm violations. critiqued [36]. Perhaps the most salient problem is More specifically, Scheff viewed the symptoms of that, in challenging the illness model of mental dis- mental disorder as a kind of residual rule breaking, order, the very reality of the distress and suffering i.e. as norm-violating behaviour that cannot be readily experienced is either downplayed and trivialized (as ascribed to any other culturally recognized category. problems of daily living [24] for example) or roman- Once this ‘primary deviance’ is labelled, the argument ticized (as a sane response to an insane world [25] for proceeds, an individual is then treated on the basis of example). From this perspective, suffering is ironically the label and, in the process of being treated differ- reconstructed away. One consequence is that it be- ently, increasingly comes to take on the stereotypical comes difficult to investigate the origins and sources of characteristics of, in this case, a mentally disordered suffering, an essential task as part of any effort to person, the result being continued and amplified norm intervene. In a sense, there is no problem, no suffering violations, i.e. ‘secondary deviance’. It is the applica- to explain, or, perhaps, more precisely, what suffering tion of the label of mental disorder that traps indivi- there is,is a functionof psychiatryanditsimpositionof duals into the career of a ‘mental patient’. needless and damaging labels. At the risk of digres- A number of classic sociological studies conducted sing, this logic is at play particularly in the realm of during the 1960s and 1970s purportedly illustrated this culture and mental disorder, and more specifically can process (from slightly differing theoretical perspec- be seen in the particular phenomena of reported high tives). For example, in their study ‘On being sane in rates of serious mental disorder in migrant and min- insane places’ [33], Rosenhan and his colleagues ority ethnic populations [37]. Any response to this gained admission to psychiatric hospitals in the USA troubling social phenomena has been sidetracked by by claiming to hear voices saying a single word, such persistent debates about whether these high rates are as ‘empty’, ‘hollow’ and ‘thud’. After admission, all real or simply a function of psychiatry misdiagnosing ‘pseudo-patients’ then behaved normally. All but one psychosis in groups whose cultural background and was given a diagnosis of schizophrenia; most were modes of expressing distress are misinterpreted [38]. treated with powerful medication and kept in hospital Despite overwhelming evidence that this is a genuine for a number of weeks. Once applied, aspects of the public health problem, this debate rumbles on [39]. In ‘pseudo-patients’, behaviour and past were viewed by short, how is it possible to make a moral case for staff through the prism of the label, note taking as mental health care, for the promotion of mental health pathological ‘writing behaviour’ for example. in disadvantaged populations for example, or for the Goffman, in his seminal work Asylums [34], saw the expansion of global mental health, if the very nature of process of becoming a patient as one in which a series the problem remains contested – if, in the background, of actors, including those in positions of authority, e.g. there is a lingering suspicion that the problems are an police, and family and friends, convince the patient- invention of a peculiar Western tradition? to-be that his or her eccentricities and difficulties This important issue aside, the perspectives of relating to others are problematic and indicative of Scheff [26], Rosenhan [33], Goffman [34] and others mental disorder. Gradually, the person comes to ac- nonetheless contain within them important insights cept this view of themselves as mentally disordered that have continued to influence research. In a straight- and in need of treatment, and so embarks on what forward sense, all concepts are social constructs de- Goffman termed ‘the moral career of the mental signed to make sense of particular phenomena. These patient’. In short, it is society, through its labelling constructs, moreover, are not neutral; the ways in of certain behaviours as abnormal, that creates mental which phenomena are conceptualized can impact on disorder; the chronic course of a disorder is the and modify both related behaviours and experiences product of secondary deviance, of those labelled and the responses of others. There are a number of

56 PERSPECTIVES AND METHODS recent analyses of the seemingly endless emergence of outcomes through the negative reactions of others, ‘new’ mental disorders that illustrate this well (e.g. and it is in attempts to understand this process that multiple-personality disorder [8,40], post-traumatic interest in labelling theory persists [45,46]. This work stress disorder (PTSD) and attention deficit/hyperac- suggests that many of the difficulties sufferers experi- tivity disorder (ADHD)) [7]. A further particularly ence in terms of finding work, accessing decent ac- telling example is provided by Barrett [41,42] in his commodation and sustaining supportive social net- detailed analysis of the development of the concept of works are not simply due to the direct effects of the dementia praecox (the forerunner of schizophrenia). disorder, but also result from the stigmatizing and He notes how Kraepelin developed the concept, a core discriminatory reactions of others (see Chapter 26 for feature of which is expectation of gradual mental and a full discussion) [47]. In recent theoretical formula- functional decline or degeneration, on the basis of tions of stigma (which extend beyond Goffman’s observations of large numbers of patients housed in original definition of stigma as a spoiled identity) [48], long-stay asylums. As Barrett comments, rather than labelling (i.e. the process of signifying others as being a core feature of the illness or disorder, such a different) is conceived as one of a number of compo- trajectory may be as much a product of the impover- nents of stigmatization. Others include stereotyping ished environments of long-stay asylums – a proposi- (i.e. the linking of differences to undesirable charac- tion for which there is empirical support [43]. Barrett teristics ([47], p. 180)), separating (i.e. marking the has persuasively argued that an expectation of degen- labelled group as fundamentally different) and dis- eration and chronicity nonetheless remains at the core crimination (i.e. exclusion of the labelled group). of the concept of schizophrenia and that this continues This noted, in much recent work on stigma it is to contribute to therapeutic pessimism and an expec- possible to discern a narrowing of the focus of enquiry, tation of chronicity, despite evidence that the majority on to public attitudes towards the mentally disordered of those who meet criteria for a schizophrenia- and education campaigns, and a movement away from spectrum disorder do not have a continuous deterior- understanding stigma within the contexts in which it ating course [44]. This also draws our attention to the arises. Recent applications of moral theory indicate possibility that thevery institutions established to help just how embedded and powerful a part of the founda- may, unwittingly, become part of the problem and tions of society stigma is [49,50]. A necessary com- exacerbate suffering. ponent of developing strategies to reduce stigma is an Further, there is good evidence that the application understanding of what drives stigma (e.g. fear) in of diagnostic labels can impact on subsequent specific local moral worlds. 5.5 SOCIAL DETERMINANTS OF MENTAL DISORDER From the beginnings of the discipline, sociologists developed and applied to explain these associations. have been interested in the social patterning and Durkheim, for example, used the concept of anomie determinants of health and disorder (both physical (i.e.a state of normlessness andsocialdisorganization) and mental). It is possible to identify two broad strands to explain the association he observed between suicide within this, one concerned with social structure (struc- and periods of rapid economic and social change. This tural strain theory) and one concerned with individual- is further seen in Faris and Dunham’s use of theories of level social experiences (social stress theory). urban development proposed by Park and others [51] Structural strain theory, simply, locates the origins to frame their investigations, and use of the concept of of disorder in theorganizationand structuresof society social fragmentation to interpret findings that hospital and has its foundations in the work of Durkheim. The admissions for schizophrenia were highest in areas basis for this is the observation that the incidence and characterized by poverty, crime, poor and unstable prevalence of mental disorders vary by social group housing, and fractured social contacts [18]. The and context. In early work, sociological theory was seminal work of Faris and Dunham, in which core

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 57 facets of the social and geographical distribution of This recent work notwithstanding, the initial atten- serious mental disorders in urban areas were first tion to theory has gradually ebbed and much of the identified, laid the foundations for subsequent colla- recent social epidemiology of mental disorder has borations between sociologists and psychiatrists been largely atheoretical, seeking simply to document which further established clear links between social the rates of occurrence of specific mental disorders class and various aspects of mental disorder and and their sociodemographic correlates. This is mental health service use. Hollingshead (a sociologist) reflected, for example, in recent years in a series of and Redlitch (a psychiatrist), for example, conducted large-scale community surveys, including the US ground-breaking research in New Haven documenting Epidemiologic Catchment Area Survey [61], the US higher levels of mental disorder and more problematic National Comorbidity Survey [62] and the British access to care for those in lower socioeconomic Psychiatric Morbidity Survey [63]. The findings from groups [20]. Other studies followed, notably the Mid- many of these are summarized in Chapter 8. This town Manhattan study [21,22]. What is notable about approach is problematic from a sociological perspec- this body of research is the attention to detail and tive for at least two reasons. To begin with, as Farmer context, and the attempt to theorize the observed argues, the social variables employed (e.g. employ- relationships. In other words, there was a concern with ment, education, social class) are ‘often desocialized’, how structure becomes actualized in daily life – they are ‘decontextualised from larger social pro- an exercise of the sociological imagination. This cesses that are both historically rooted and linked to said, the actual mechanisms through which social persons and processes that are not visible to the survey structures become actualized or embodied to cause researcher’ (Castro and Farmer, 2005, p. 53). The the onset or exacerbation of mental distress have been sources of social inequalities, for example, that are largely unspecified, beyond somevague notion that the so strongly associated with many mental disorders do certain structures and contexts are a source of not feature, as if they are somehow a part of the natural ‘strain’ [35]. order and not a function of theways in which society is There are more recent examples of research organized. The ‘structural violence’, to use Farmer’s informed by sociological theory that seek to examine phrase, that exposes certain groups to the linked the relationship between aspects of the wider social miseries of poverty, poor housing, isolation, poor environment and mental disorders [52,53]. The health and health care, and so on, remains hidden – attempt to apply theories of social capital to the study the very social nature of suffering is obscured. The of mental disorder is one such example [54]. To date, data cease to be a window through which to under- most researchers have drawn specifically from stand how social structure and contexts become man- Putnam’s formulation of social capital as a collective ifest in the suffering of individuals, families and resource that inheres in the social ties and connections communities. They become simply a part of the of local communities [55]. There are a small number routinized book-keeping of modern societies, docu- of examples where other conceptualizations of social menting met and unmet need for technical and indi- capital have been invoked [56] (e.g. individual access vidualized interventions to reduce symptomatology. to stocks of social capital, following Bourdieu [57]). At times, researchers are left almost at a loss to However, as potentially fruitful as this may prove to explain, or even develop testable hypotheses for, be, inconsistencies in definitions and methods con- observed associations (e.g. the persistent link between tinue to undermine the development of a consistent urbanicity and psychosis [64] – a finding that is at body of knowledge in relation to mental disorder, and present largely uninterpretable [65]). Linked to this, findings are equivocal – some suggest a link, others do the problems and suffering faced become increasingly not [54,58–60]. This, moreover, hints at some of the splintered into discrete ‘disorders’, each separable key limitations of research that can be broadly from the other (as evidenced by the explosion of bracketed within a structural strain framework, disorders in DSM) and standing outside the context e.g. problems of consistent and accurate measurement within which they arise and persist. This is, in fact, of social structure and organization. belied by the very evidence from these surveys that

58 PERSPECTIVES AND METHODS suggests ‘comorbidity’, the technical terminology to population. These problems are interlinked, and a full account for the simultaneous occurrence of more than understanding of their origins and how to address one ‘discrete’ problem, is the norm [62]. More than them is unlikely to be achieved if each one is studied this, as suffering is decontextualized in this way and in isolation [39]. Instead, the sum of human suffering measured against a checklist of symptoms, so its very is reduced to quantifiable indices (as in, for example, meaning becomes subverted (sadness, for example, the Disability Adjusted Life Year). arising from job loss, relationship problems or finan- A further major strand in the social science of cial difficulties is transformed into depression, a dis- mental disorder is concerned with how individual- crete disorder primarily treatable with individualized level processes and experiences impact on the onset, therapy – be that psychological or pharmacological) course and outcome of disorder. Most of this, impli- (see also below) [6]. A notable exception to this trend citly or explicitly, can be grouped within the broad is Link and Phelan’s theory of fundamental social framework of social stress theory. Stress has become a causes [66], developed to explain why most disorders ubiquitous explanation for a multitude of ills in mod- tend to coalesce in disadvantaged populations, even as ern societies. Particularly within the social sciences, it the nature of these disorders changes (e.g. from in- is used as the thread to link adverse social experiences fectious disease to chronic physical illness to mental and mental distress. The concept was first introduced disorder) (see Chapter 14). into the sciences in the 1930s by the psychologist, In short, the use of crude ‘sociodemographic’ vari- Hans Selye [68,69]. Selye saw stress as the body’s ables in the absence of theoretically informed hypoth- physiological response to stressors, by which he eses about the processes underlying observed associa- meant anything that represented an insult or threat to tions means that the findings of large-scale surveys the body, such as extreme heat or cold (i.e. anything can tell us little about the origins of the social dis- provoking a stress response) [35]. His model of stress parities they document. Where more sophisticated was developed on the basis of experiments with concepts are employed (e.g. social fragmentation, laboratory animals and comprises four components: social capital), additional methodological challenges (1) stressors, (2) factors that mediate the impact of are faced, and it remains that the invisible hand of stressors on the body (e.g. personality, social support ‘strain’ or ‘stress’ has to be invoked to explain how networks), (3) the general adaptation syndrome (see social contexts impact on individuals. What does it below) and (4) responses, positive or negative ([70], mean on a daily basis to live within such contexts? p. 178). Selye conceptualized the three stages of How do individuals and families respond? In other physical response to stressors that animals pass words, what might mediate the impact and strain of through as the general adaptation syndrome: (1) alarm living in such conditions? It is this ‘black box’ of lived reaction, (2) resistance and (3) exhaustion. They experience and meaning that social epidemiology – constitute a process during which the body is physio- even that theoretically informed by sociology – is logically aroused and prepared to resist threat (i.e. generally ill-equipped to penetrate [67]. What is more, fight or flight), which, if continued for a sufficient the characteristic experiences of mental disorder do period, leads to exhaustion and illness. Conceptually, not occur in isolation; they tend to be interlinked with if not in detail, this is the basic framework for current other difficulties and these linked disadvantages tend views about how social stressors impact on individuals to coalesce in vulnerable groups, creating a pool of to increase risk of mental illness. In this, social problems that cannot be splintered and addressed experiences and contexts (e.g. life events, ongoing individually. For example, there has been a consider- difficulties, trauma, problematic interpersonal rela- able amount of research investigating the high rates of tionships) are reconfigured as stressors that threaten psychosis in the UK black Caribbean population [38]. the integrity of individuals. Almost all of this has been conducted without refer- A whole range of problematic contexts and occur- ence to the equally robust evidence of high levels of rences – i.e. social stressors (e.g. childhood abuse, school exclusions, substance use, involvement with bullying, trauma, unemployment, job loss, relation- criminal justice services and unemployment in this ship problems, family conflict) – have been linked

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 59 with a wide variety of physical and mental disorders context moderates the impact of stress arising from (including asthma, breast cancer, lupus, myocardial negative events (job loss, relationship break-up) and infarction, headaches, irritable bowel syndrome, de- chronic difficulties (financial hardship, interpersonal mentia, gastrointestinal disorders, diabetes, Crohn’s conflict), and how multiple problems interconnect and disease and, of course, the full range of mental health interact to cause and sustain mental distress (see problems, from insomnia to chronic schizophre- Chapter 16). This, moreover, can persist across gen- nia) [35]. Many of these are discussed at various erations. We know from other research, for example, points in chapters throughout this book. A number that the experience of mental distress in a parent of methodological approaches have also been used, extends outwards, affecting the whole family, espe- ranging from predefined brief checklists of events and cially children who are themselves consequently at traumas (in childhood and adulthood) [71] to more increased risk of subsequent disorder (see Chapter 9). detailed semi-structured interviews that attempt to However, for all this work contextualizes the contextualize events and difficulties as a basis for exposures (to use terminology from epidemiology), evaluating their meaning and likely degree of it leaves uncontested the nature of the outcomes (e.g. threat [72]. This latter approach, developed and depression, psychosis) and their conceptualization as applied by George Brown, Tirril Harris and collea- discrete disorders that stand apart from other experi- gues, has underpinned a programme of research that ences in the flow of everyday life. It is perhaps no has revealed much about the relationship between surprise that it is this strand of sociology that has been adverse events and difficulties and depression, leading most readily accepted and incorporated into psychia- to the development of a sophisticated model that posits try. For some, this constitutes a significant limitation. a primary role for adverse social experiences in the In so far as the diagnosis of depression, for example, aetiology of depression [73]. More specifically, this fails to take account of the context within which programme of research shows that: (1) the relationship feelings of sadness and distress arise, Horwirtz and is strongest for negative events, particularly those Wakefield [6] argue that the diagnosis is likely to involving loss, entrapment and humiliation; (2) the embrace all those whose distress is an understandable association remains when only events occurring and time-limited response to negative events (e.g. a before onset of disorder is considered; (3) the six- relationship break-up, job loss) – that is normal month period pre-onset is important; (4) social cir- sadness, not depressive disorder. Unwittingly, this cumstances and supportive resources mediate the research may then contribute to the professional risk, the risk being highest among women with three transformation of normal sorrow into mental disorder. children aged under 11 who lack supportive networks This returns us full circle to fundamental questions and relationships; and (5) positive, fresh-start events about how such distress and suffering should be can promote recovery [73]. Here we can see how conceptualized. 5.6 SOCIAL INFLUENCES ON RESPONSES TO MENTAL DISORDER Much social science research, in the time since the necessarily going so far as to propose the complete 1970s, has centred on understanding how social and cultural relativity of systems of knowledge and be- cultural contexts shape the manifestation of, and lief) [74]. For example, a classic study by Helman [75] responses to, mental disorder. This disparate body of illustrates how culturally based illness beliefs shape research draws from a number of theoretical threads. help-seeking and interactions with health care ser- Following on from Benedict [23], a significant vices. In his study of older residents in north London, interest in medical anthropology has been on how the Helman sought to explore what was meant by the experiences and behaviours characteristic of mental popular notion ‘feed a cold, starve a fever’ and how lay disorder and symptoms of physical illness are under- beliefs related to this notion influenced interactions stood and managed in diverse settings (without with health care professionals. The detailed accounts

60 PERSPECTIVES AND METHODS elicited by Helman revealed a widespread folk clas- role’ [14]). In brief, illness behaviour refers to the sificatory system of what medicine terms infectious ways in which individuals and significant others per- diseases, resting on the separation of illnesses into hot, ceive, evaluate and respond to symptoms of illness, cold, wet and dry. This system was based on per- and was formulated in the context of growing evi- ceived, and entirely subjective, notions of changes in dence that the nature and severity of symptoms alone body temperature associated with particular symp- did not determine when and what type of help was toms. Colds were caused by contact with the natural sought. On the basis of this early work, Mechanic [78] environment, e.g. damp or rain or wind, or by cold identified a small number of factors that influenced entering vulnerable areas of the body, such as the feet illness behaviour, including: the nature of symptoms; or the head. Treatment involved increasing a sufferer’s culture, particularly beliefs about illness which shape temperature with, for example, hot food, hot drinks the interpretation of symptoms; the impact of symp- and rest in a warm bed, and by generating personal toms on family and social functioning; the responses body heat using tonics or food – hence ‘feed a cold’. of significant others; and the range of treatment op- Fevers, in contrast, were thought to be caused by tions available (i.e. the structure of the local health germs, bugs or viruses. Germs and the like pass from care system). Much of the early research within this one person to another, as in the idea of ‘picking up a framework emphasized the processes of negotiation germ’, and as such are the result of social interaction. that occur within social networks in coming to an Remedies involve expelling the offending agent by, understanding of an episode of illness and deciding on for example, sweating it out or ‘coughing up the appropriate responses. Zola [79], for example, demon- muck’, or by attacking the germs with modern med- strated that illness behaviour and the decision to seek icine, i.e. antibiotics. Related to this, Helman noted medical help is, at least partly, a function of the family that doctors often couched their diagnoses and pre- and social context within which illness is experienced. scribed treatments in lay terms: ‘you’ve picked up a Resolving health problems, for example, is only one of bug’, ‘it’s a tummy bug – I’m afraid there’s one going a number of competing demands on individuals and it around’, thereby further reinforcing popular beliefs. may be that other needs, from the point of view of the This example illustrates how beliefs about causation sufferer, are more salient at any point. Caring for and the nature of illness can impact on how individuals children, work, study and other such activities may respond: for colds, a hot drink and warm bed; for take higher priority than seeking help for health fevers, getting it out of the system with the help of problems, particularly if consultation may lead to modern medicine. The study further suggests that time-consuming treatment or hospitalization. Further, where cultural idioms are shared ease of communica- the process of negotiating and weighing up the pros tion between doctor and patient is increased. and cons of seeking professional help will be influ- A number of the frameworks and concepts devel- enced by culturally defined roles, obligations and oped in ethnographic studies of healing systems in norms governing what is considered an appropriate diverse cultures have subsequently been widely em- response to specific symptoms and illness. For exam- ployed in studies of how illness is conceptualized and ple, a study of socioeconomic variations in responses managed cross-culturally (e.g. the notion of health to chest pain in Scotland [80] found that respondents care systems as cultural systems comprising at least from deprivedareas were less likely to seek help due to three overlapping sectors – the professional, the folk a tendency to normalize chest pains in the context of and the popular; explanatory models of illness, as a high general levels of ill-health and due to a constella- basis for understanding the ideas about episodes of tion of related beliefs about the appropriateness of illness that each individual – patient, relative and seeking help and fear of being blamed for ‘high-risk clinician – brings to the clinical encounter) [76]. behaviours’, such as smoking, drinking and poor In sociology, similar research has been conducted diet [80]. within the framework of illness behaviour, initially In the time since this early work there has been a developed by Mechanic [77] (and perhaps to a lesser vast amount of research investigating responses to extent using Parson’s formulation of the ‘sick illness and help-seeking, much of it conducted within

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 61 the theoretical frameworks briefly sketched here. It is illness are stripped from the local social worlds in beyond the scope of this chapter to even attempt a which they emerge and have meaning. There is, more- synopsis of this work. One observation (in a similar over, a related tendency to reconstruct the process of veinto ones made above) is worth noting. At the risk of managing an episode of illness as a series of rational overstatement (and a number of exceptions notwith- choices, in which individuals calculate the benefits standing – e.g. Pescolsolido et al. [81] and see Chapter and costs of seeking help against pre-existing ideas 24), there has been an increasing tendency to employ about what is wrong and options for putting it right. the concepts developed in early work (e.g. explanatory This may be the case for some, but it ignores the models) inflexibly, as fixed quantifiablevariables to be socially negotiated process that characterizes re- correlated with specified outcomes (e.g. admission to sponses to distress and suffering, a process recently hospital), as if these captured unchanging and readily well illustrated in a study by Pescosolido and collea- measurable characteristics of individuals, much like gues [81] detailing how people come to use mental taking blood pressure readings. This is particularly health services. In other words, this represents a true of studies of explanatory models, initially in- further area where initial theoretically-informed tended as a window into cultural context, or of beliefs research in the social sciences has been superseded about illness more broadly. The emphasis on context by increasingly atheoretical studies designed primar- and the interconnectedness of health systems and ily to address specific questions of perceivedrelevance behaviours is consequently lost – again, notions of to health services. 5.7 A FAILURE OF THE SOCIOLOGICAL IMAGINATION Within each of the areas discussed (the nature of contributed to a withdrawal of psychiatry from public mental disorder, the social role of psychiatry, the health, at least until recently. social distribution and determinants of mental disor- In other words, the social aspects of mental disorder der, the social and cultural influences on responses to are (again with some exceptions) frequently relegated disorder), there is a rich social science tradition (both to the periphery. Social science (for now) operates at theoretical and methodological) that has contributed these margins. Certainly, health service research and greatly to our understanding of the myriad connec- basic social epidemiology are important as a basis for tions between society and mental disorder. However, identifying need and developing appropriate interven- at the same time, social science perspectives within tions and treatment systems. However, our under- psychiatry have been in retreat, and it is possible to standing of what needs to be done would be greatly discern a general tendency to reify or ignore theore- enriched if social science research was once again tical concepts and decontextualize findings, with stu- broadened to embrace the wider social forces and dies of social aspects of mental disorder becoming contexts that impinge on and shape the lives of those increasingly narrow in their focus. This reflects, per- who mental health services are designed to help. It is haps uncharitably, a retreat of social science within this, for instance, that will allow a fuller understanding psychiatry into health services research and social of the interconnected problems that underpin the epidemiology and, with this, a decline of theory and suffering James described in the example provided interest in the major questions that initially drove above. What we have suggested throughout this re- early sociological studies of mental disorder – i.e. a view is that social suffering, with its emphasis on the failure of the sociological imagination. This has been intersubjective experience of distress and its origins in amplified by parallel developments in psychiatry, the interconnected adversities and traumas that indi- most notably an increasing dominance of biological viduals, their families and communities face, offers a theories of aetiology, categorical approaches to diag- unifying frame through which to consider the intrinsic nosis as codified in DSM and ICD, and an increase social nature of mental disorder and re-enrich the in pharmacological treatments. This has further sociology of mental disorder.

62 PERSPECTIVES AND METHODS REFERENCES 18. Faris, R. and Dunham, H. (1939) Mental Disorders in Urban Areas, University of Chicago Press, Chicago, 1. Kleinman, A., Das, V. and Lock, M. (1997) Illinois. Introduction, in Social Suffering (eds A. Kleinman, 19. Hare, E. (1956) Mental illness and social conditions V. Das and M. Lock), University of California Press, in Bristol. Journal of Mental Science, 102, 349–357. Los Angeles, California, pp. ix–xxvii. 20. Hollingshead, A. d. B. and Redlich, F. C. (1958) Social 2. Kleinman, A., Das, V. and Lock, M. (eds) (1997) Social Class and Mental Illness; A Community Study, John Suffering, University of California Press, Los Angeles, Wiley & Sons, Inc., New York. California. 21. Gruenberg, E. M. (1963) A review of mental health in 3. Morgan, C., Burns, T., Fitzpatrick, R. et al. (2007) the metropolis. The midtown Manhattan study. Social exclusion and mental health: conceptual and Milbank Memorial Fund Quarterly – Health and methodological review. British Journal of Psychiatry, Society, 41, 77–94. 191, 477–483. 22. Langner, T. S. (1960) Environmental stress, degree of 4. World Health Organization (WHO) (2001) The World psychiatric impairment and type of mental disturbance. Mental Health Report, World Health Organization, Psychoanalysis and the Psychoanalytic Review, Winter, Geneva. 47, 3–16. 5. Desjarlais, R., Eisenberg, L., Good, B. and Kleinman, 23. Benedict, R. (1934) Antropology and the abnormal. A. (eds) (1997) World Mental Health: Problems and Journal of General Psychology, 10, 59–80. Priorities in Low-Income Countries, Oxford University 24. Szasz, T. (1960) The myth of mental illness. American Press, Oxford. Psychologist, 15, 113–118. 6. Horwitz, A. and Wakefield, J. (2007) The Loss of 25. Laing, R. D. and Esterson, D. (1964) Sanity, Madness Sadness: How Psychiatry Transformed Normal and the Family, Basic Books, New York. Sorrow into Depressive Disorder, Oxford University 26. Scheff, T. (1966) Being Mentally Ill: A Sociological Press, Oxford. Theory, Aldine Publishing Company, Chicago, Illinois. 7. Horwitz, A. (2004) Creating Mental Illness, 2nd edn, 27. Bolton, D. (1997) Encoding of meaning: deconstructing Chicago University Press, Chicago, Illinois. the meaning/causality distinction. Philosophy, Psychiatry 8. Hacking, I. (1998) Rewriting the Soul: Multiple and Psychology, 4, 255–267. Personality and the Science of Memory, Princeton 28. Roth, M. and Kroll, P. (1986) The Reality of Mental University Press, Princeton, New Jersey. Illness, Cambridge University Press, Cambridge. 9. Comte, A. (1855) The Positive Philosophy, Harvard 29. Bateson, G., Jackson, D., Haley, J. and Weakland, J. University Press, Boston, Massachusetts. (1956) Toward a theory of schizophrenia. Behavioural 10. McLellan, D. (2000) Karl Marx: Selected Writings, Science, 1, 251–264. Oxford University Press, Oxford. 30. Fromm-Reichmann,F.(1948)Notesonthedevelopment 11. Durkheim, E. (1938) The Rules of Sociological Method, of treatment of schizophrenics by psycho-analytic The Free Press, New York. psychotherapy. Psychiatry, 11, 263–273. 12. Parkin, F. (2002) Max Weber (Routledge Key 31. Eisenberg, L. (1977) Psychiatry and society: a Sociologists), Routledge, London. sociobiologic synthesis. New England Journal of 13. Durkheim, E. (1952) Suicide, The Free Press, Glencoe, Medicine, 296, 903–910. Illinois. 32. Becker, H. (1963) Outsiders: Studies in the Sociology 14. Parsons, T. (1951) The Social System, Routledge & of Deviance, Free Press, New York. Kegan Paul Ltd, London. 33. Rosenhan, D. L. (1973) On being sane in insane places. 15. Foucault, M. (2001) Madness and Civilisation, Science, 179 (70), 250–258. Routledge, London (originally published 1967). 34. Goffman, E. (1961) Asylums, Anchor, New York. 16. Rabinow, P. (1991) The Foucault Reader: An 35. Thoits, P. A. (1999) Sociological approaches to mental Introduction to Foucault’s Thought, Penguin, London. illness, in A Handbook for the Study of Mental Health: 17. Wright-Mills, C. (2000) The Sociological Imagination, Social Contexts, Theories and Systems (eds A. V. Oxford University Press, Oxford (originally published Horwitz and T. L. Schied), Cambridge University 1959). Press, Cambridge.

SOCIAL SCIENCE PERSPECTIVES: A FAILURE OF THE SOCIOLOGICAL IMAGINATION 63 36. Morgan, C. (2008) Social science, psychiatry and 49. Yang, L. H. and Kleinman, A. (2008) ‘Face’ and the psychosis, in Society and Psychosis (eds C. Morgan, embodiment of stigma in China: the cases of K. McKenzie and P. Fearon), Cambridge University schizophrenia and AIDS. Social Science and Press, Cambridge, pp. 25–40. Medicine, 67 (3), 398–408. 37. Cantor-Graae, E. and Selten, J. P. (2005) Schizophrenia 50. Yang, L. H., Kleinman, A., Link, B. G. et al. (2007) and migration: a meta-analysis and review. American Culture and stigma: adding moral experience to stigma Journal of Psychiatry, 162, 12–24. theory. Social Science and Medicine, 64 (7), 38. Fearon, P. and Morgan, C. (2006) Environmental 1524–1535. factors in schizophrenia: the role of migrant studies. 51. Park, R., Burgess, E. and McKenzie, R. (1925) The Schizophrenia Bulletin, 32 (3), 405–408. City, University of Chicago Press, Chicago, Illinois. 39. Morgan, C. and Hutchinson, G. (2009) The social 52. Allardyce, J., Gilmour, H., Atkinson, J. et al. (2005) determinants of psychosis in migrant and ethnic minority Social fragmentation, deprivation and urbanicity: populations: a public health tragedy. Psychological relation to first-admission rates for psychoses. British Medicine, 1–5 (in press). Journal of Psychiatry, 187, 401–406. 40. Hacking, I. (1999) Mad Travellers: Reflections on the 53. Boydell, J., van Os, J., McKenzie, K. et al. (2001) Reality of Transient Mental Illness, Free Association Incidence of schizophrenia in ethnic minorities in Books, New York. London: ecological study into interactions with 41. Barrett, R. J. (1998) Conceptual foundations of environment. British Medical Journal, 323, schizophenia: I. Degeneration. Australian and New 1336–1338. Zealand Journal of Psychiatry, 32 (5), 617–626. 54. McKenzie, K. and Harpham, T. (eds) (2006) Social 42. Barrett, R. J. (1998) Conceptual foundations of Capital and Mental Health, Jessica Kingsley, schizophrenia: II. Disintegration and division. London. Australian and New Zealand Journal of Psychiatry, 55. Putnam, R. D. (2000) Bowling Alone: The Collapse and 32 (5), 627–634. Revival of American Community, Simon & Schuster, 43. Wing, J. and Brown, G. (1970) Institutionalism New York. and Scizophrenia, Cambridge University Press, 56. Webber, M. P. and Huxley, P. J. (2007) Measuring Cambridge. access to social capital: the validity and reliability of 44. Warner, R. (2003) Recovery from Schizophrenia: the Resource Generator-UK and its association with Psychiatry and Political Economy, 3rd edn, common mental disorder. Social Science and Medicine, Routledge, London. 65 (3), 481–492. 45. Link, B. and Phelan, J. (1999) The labelling theory of 57. Bourdieu, P. (1977) Outline of a Theory of Practice, mental disorder (II): the consequences of labelling, in A Cambridge University Press, Cambridge. Handbook for the Study of Mental Health: Social 58. De Silva, M. J., Huttly, S. R., Harpham, T. and Kenward, Contexts, Theories and Systems (eds A. V. Horwitz M. G. (2007) Social capital and mental health: a and T. L. Schied), Cambridge University Press, comparative analysis of four low income countries. Cambridge, pp. 361–376. Social Science and Medicine, 64 (1), 5–20. 46. Phelan, J. and Link, B. (1999) The labelling theory of 59. De Silva, M. J., McKenzie, K., Harpham, T. and Huttly, mental disorder (I): the role of social contingencies in S. R. (2005) Social capital and mental illness: a the application of psychiatric labels, in A Handbook for systematic review. Journal of Epidemiology and the Study of Mental Health: Social Contexts, Theories Community Health, 59 (8), 619–627. and Systems (eds A. V. Horwitz and T. L. Schied), 60. Stafford, M., De Silva, M., Stansfeld, S. and Cambridge University Press, Cambridge. Marmot, M. (2008) Neighbourhood social capital 47. Thornicroft, G.(2006) Shunned:DiscriminationAgainst and common mental disorder: testing the link in a People with Mental Illness, Oxford University Press, general population sample. Health Place, 14 (3), Oxford. 394–405. 48. Goffman, E. (1963) Stimga: Notes on the Management 61. Robins, L. E. and Regier, D. N. (1991) Psychiatric of a Spoiled Identity, Prentice Hall, Boston, Disorders in America: The Epidemiologic Catchment Massachusetts. Area Study, The Free Press, New York.

64 PERSPECTIVES AND METHODS 62. Kessler, R. C., Anthony, J. C., Blazer, D. G. et al. (1997) 71. Holmes, T. H. and Rahe, R. H. (1967) The social The US National Comorbidity Survey: overview and readjustment rating scale. Journal of Psychosomatic future directions. Epidemiologia e Psichiatria Sociale, Research, 11, 213–218. 6 (1), 4–16. 72. Brown, G. and Harris, T. (1978) The Social Origins of 63. Jenkins, R., Bebbington, P., Brugha, T. et al. (2003) Depression, Tavistock, London. British Psychiatric Morbidity Survey. International 73. Harris, T. (2001) Recent developments in understanding Review of Psychiatry, 15 (1), 14–18. the psychosocial aspects of depression. British Medical 64. Krabbendam, L. and van Os, J. (2005) Schizophrenia Bulletin, 57, 17–32. and urbanicity: a major environmental influence – 74. Helman, C. (2000) Culture, Health and Illness, 4th edn, conditional on genetic risk. Schizophrenia Bulletin, Butterworth Heinemann, Oxford. 31 (4), 795–799. 75. Helman, C. (1978) Feed a cold, starve a fever. Culture, 65. McGrath, J. and Scott, J. (2006) Urban birth and risk of Medicine and Psychiatry, 2, 107–137. schizophrenia: a worrying example of epidemiology 76. Kleinman, A. (1980) Patients and Healers in the Context where the data are stronger than the hypotheses. of Culture: An Exploration of the Borderland between Epidemiologia e Psichiatria Sociale, 15, 243–246. Antropology, Medicine and Psychiatry, University of 66. Link, B. and Phelan, J. (1995) Social conditions as California Press, Los Angeles, California. fundamental causes of disease. Journal of Health and 77. Mechanic, D. (1962) The concept of illness behaviour. Social Behaviour, 35, 80–94. Journal of Chronic Disease, 15, 189–194. 67. Berkman, L. F. and Kawachi, I. (eds) (2000) Social 78. Mechanic, D. (1968) Medical Sociology: A Selective Epidemiology, Oxford University Press, Oxford. View, Free Press, London. 68. Seyle, H. (1978) The Stresses of Life, McGraw-Hill, 79. Zola, I. (1973) Pathways to the doctor. Social Science Maidenhead. and Medicine, 7, 677–689. 69. Pearlin, L. (1999) Stress and mental health: a 80. Richards, H. M., Reid, M. E. and Watt, C. M. (2002) conceptual overview, in A Handbook for the Study of Socio-economic variations in responses to chest pains: Mental Health: Social Contexts, Theories and Systems a qualitative study. British Medical Journal, 324, (eds A. V. Horwitz and T. L. Schied), Cambridge 1308–1310. University Press, Cambridge, pp. 161–175. 81. Pescosolido, B. A., Gardner, C. B. and Lubell, K. M. 70. Wheaton, B. (1999) The nature of stressors, in A (1998) How people get into mental health services: Handbook for the Study of Mental Health: Social stories of choice, coercion and ‘muddling through’ Contexts, Theories and Systems (eds A. V. Horwitz from ‘first-timers’. Social Science and Medicine, 46 and T. L. Schied), Cambridge University Press, (2), 275–286. Cambridge, pp. 176–197.

6 Concepts and challenges in capturing dynamics of the wider social environment 1 Stephani L. Hatch and Dana March 2 1 Department of Psychological Medicine, Institute of Psychiatry, King's College London, London, UK 2 Department of Epidemiology and Center for Social Inequalities and Health, Mailman School of Public Health, Columbia University, New York, USA 6.1 INTRODUCTION As the social environment shapes individuals, indivi- How an individual engages with the social envir- duals shape the social environment. Yet examinations onment is heavily influenced by his or her acquired of the wider social environment in social psychiatry level of cultural competence and accumulated experi- have not fully embraced this dynamic. This chapter ence developed over the life course. We raise some of outlines conceptual considerations and highlights par- the challenges in capturing the social environment ticular challenges in understanding how the dynamics and, most importantly, address the primary processes and complexities of social structure and contextual that link individuals to the social environment. We factors across the life course ultimately bear on in- argue that emphasis should be placed on the relative dividuals’ mental health. Many assessments of the positioning of individuals within their social environ- wider social environment attempt to identify impor- ments, the salience of social context at different points tant exposures that confer either risk or protection across the life course and individuals’ embodiment of in terms of mental health outcomes, from social historically shaped social context. interactions (e.g. disorder, social capital) to physical The issues discussed in this chapter are by no means environs (e.g. builtenvironment, local services, access exhaustive. Rather, the concepts and challenges to to safe places, etc.) and mobility across settings (e.g. empirical examination of the wider social environ- residential stability, housing tenure, migration). We ment addressed below are meant to provide guidance discuss many of these environmental characteristics, in social psychiatric research considering social focusing specifically on social capital and mobility. context, underscore particular points that warrant Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

66 PERSPECTIVES AND METHODS further consideration and encourage critique of with special attention granted to depression and current thinking regarding the representation of the psychosis. An elaboration on social capital, which social world within social psychiatry and psychiatric illuminates the broader contours of a narrowly epidemiology research. We begin with the conceptua- applied sociological concept in social psychiatry lization of social structure and consider how social and psychiatric epidemiology, follows. Finally, we structure is embodied over the life course by sociali- underscore the importance of considering the life zation, drawing on an example from the study of course and consider the specific complexities in stress and mental health. We then proceed to a dis- empirical research concerned with social and geo- cussion of area-level factors and mental health, graphic mobility. 6.2 SOCIAL STRUCTURE ...[I]ndividuals contain within themselves their past and present position in the social structure ([1], p. 434). Social structure broadly refers to enduring pat- Time is a critical consideration in understanding terned behaviour and interactions between individuals how social structure becomes embodied. First, social or groups that represent hierarchical social arrange- structure itself is influenced by history, and perspec- ments within a social system [2,3]. The meaning and tives of social statuses shift over time. Second, social usage of the concept depends on the theoretical per- structure is embodied across the life course at critical spective being employed. Considerations of social times (e.g. childhood, adolescence and different structure in research are implicit in attempts to under- phases in adulthood). One of the main challenges is stand relations between social stratification and capturing changing social structure as it pertains to health, often categorized as health disparities or individual development and mental health over the life health inequalities research [4–6]. While the identifi- course. This has been most obviously addressed in cation of differences in the distribution of disorders researchers’ attempts to highlight the influence of is key in this arena, consideration of the interrelation- social structure on experience and behaviour, with an ships among hierarchically arranged social statuses emphasis on the contribution of social processes, such (e.g. socioeconomic status, gender, race/ethnicity, as socialization, described in detail below [12]. age) is imperative. Social interactions are defined, in part, by the posi- tion within a hierarchical social structure that a person 6.2.1 Socialization: linking the individual inhabits. That position, in turn, becomes a constituent to social structure of the person. One concept particularly pertinent to this sociological phenomenon is habitus. In his early work, Social processes have been overlooked in situating the Bourdieu [7] describes habitus as ‘embodied social individualwithin his or herenvironment.However, ina structures’ or schemes that reflect social status. Eco- convergence of thinking across disciplines, social pro- nomic and social conditions generate shared percep- cesses are receiving more attention, both directly and tions, observations, motivations and behaviours based indirectly. For example, research regarding ethnic on occupying certain social statuses and interacting identity development in relation to mental health and with others in the same status position [8]. Subsequent service use exemplifies the dynamics between social work by Bourdieu and others offering nuanced inter structure andtheindividual[13–15].Implicitintheprocess pretations has considered the impact of gender and of ethnic identity development is socialization [16]. racial/ethnic status positions on individuals’ interpreta- Defined as the manner in which individuals selec- tion and functioning within social environments [1,9–11]. tively acquire skills, knowledge, values, motives,

CONCEPTS AND CHALLENGES IN CAPTURING THE SOCIAL ENVIRONMENT 67 behaviours and roles [17,18], socialization is one of gender, class and racial or ethnic identities. The the common processes by which social structure is family provides the first instance of group member- embodied and maintained and by which individuals ship in which individuals are endowed with similar become embedded in their social environment. How- schemes, or habitus. For example, the development ever, it appears to be overlooked, particularly in of ethnic identity results from an intergenerational research that considers interactions between indivi- transmission of majority group views about race and duals and the social environment in relation to mental ethnicity from parents to children. Ethnic identity health. In this section, we discuss the main sources of represents a common heritage shared within a parti- and domains in which socialization can occur across cular group, and a sense of unification and cohesive- the individual life course. ness in conceptualizations of the self [22,23]. This Socialization processes occur within contexts that occurs through processes, both subtle and overt, shift in salience over the life course: the family, the which involve the parental transmission of attitudes school, among peers, the neighbourhood and in the and behaviours and the promotion of ethnic identity workplace. The explicit purpose and function of inter- development. actions within these contexts is to socialize and in- In such processes, parents’ practices are shaped and tegrate individuals at the group level [16]. Thus, influenced by experiences from their own families of embodiment of the social environment occurs partly origin and experiences in their immediate social en- through this process, particularly within social inter- vironment [24]. Children are prepared to adapt and actions and social participation. The complexities of assimilate to specific contextual demands influenced how this process unfolds over the life course are best by geographic location and historical period. Factors captured in longitudinal follow-up that allows re- such as behavioural distinctions, language differ- searchers to examine the impact of traditional life- ences, physical features that are expressed in stereo- phase transitions (e.g. education completion, work types and stigma add uniquely to the intricate nature of entry, marriage or household formation). Understand- ethnic identity formation, especially during childhood ing these life course transitions and thinking in terms and adolescence [25,26]. of life course trajectories are equally important in It is within the family context that both privilege determining how to promote social stability (e.g. and subordination are introduced, taught and commu- social integration, maintenance of social networks nicated to children. In doing so, social distance is and perceived social support) and to reduce social defined within specific social spaces (i.e. from the exclusion (e.g. instability in housing, debt, marital and neighbourhood to larger geographical areas). In the family breakdown), particularly as each pertains to example of race and ethnicity, parents from black and mental health. minority ethnic groups who place their children in a Bourdieu’s concept of habitus, as noted, is particu- majority white environment will no doubt transmit larly useful in a discussion of socialization. More fully, ways to move about these social spaces to either assert habitus is defined as the ‘mental or cognitive power among the privileged or reduce the likelihood structures’ or somewhat durable dispositions that com- of confrontation. Similar lessons apply to social class, bine into a series of schemes through which the social which are especially apparent in the absence of race world is perceived and evaluated [19–21]. These and ethnicity. As discussed in greater detail below, this schemes are utilized in social interactions, and provide constitutes a critical element in understanding the the guiding principles for making choices and replicat- intergenerational transfer of cognitive dispositions ingpastexperienceswithina givensocialenvironment. and appraisals, beliefs and behaviours, regardless of The family of origin is the initial source from the subtlety and inadvertence with which parents which meaning and purpose at the individual and communicate these social facts (in a Durkheimian group level arises, establishing patterns of power and sense, i.e. ways of acting, thinking, feeling that are structures of authority. Parents and primary care- external to the individual and made possible through givers take on a multitude of socialization goals. power of coercion by those within privileged status Many of these are central in the development of groups).

68 PERSPECTIVES AND METHODS Schools have more narrowly defined socialization wellbeing, especially when accompanied by a sense of goals, including formal instruction and cognitive mastery or self-efficacy. Work can also be a common development (i.e. the acquisition of knowledge, de- stressor and chronic strain [31]. velopment of analytic and verbal skills). Bourdieu identifies educational systems as among ‘the funda- mental institutions in the reproduction of class 6.2.2 An example of social structure inequality from generation to generation’ ([27], in mental health research p. 1463). Nash [28] offers an interpretation: schools have the capability to shape and impact habitus in a From a sociological perspective, social arrangements manner that extends beyond the role of the family. The and processes are fundamental to understanding both structural features of the classroom encourage stu- the causes and consequences of poor mental dents to formulate social comparisons and develop health [32]. These frameworks tend to focus on po- their responses to a system of costs and rewards. Most tential causal factors that are linked to occupying a important in terms of status attainment, schools func- certain social status position [33,34]. Sociological tion as a context in which perceptions about oppor- mental health research focuses on social variation in tunity structures and what is necessary to succeed are the incidence and prevalence of psychological symp- formed. toms related to social factors, such as social support, Peer groups are an avenue for socialization that are social capital, social integration or participation, and uniquely characterized by being voluntary social in- acute and chronic stressors. teractions in which boundaries are more fluid and The sociological study of stress and mental health is independence can be exercised. Friendships are, more an example of a research programme that heavily often than other relationships, associations between considers the influence of social structure (see Ane- status equals. They help to develop competence and shensel [35] for a review). Stress and related strains validation of the self and provide an opportunity to are referred to as consequences of social arrangements acquire knowledge not introduced in family and or as antecedents to a health-related disorder. Two school contexts. In the example of racial socialization, broad agendas dominate research in this area: the many studies limit their scope to a single sphere, the social distribution of stress and social variation in the family. However, it is at a relatively young age response to stress. For both, fundamental social sta- that children begin to engage in identity politics tuses (e.g. gender, social class, ethnicity – see Chapter with their peers. Peer socialization can heavily influ- 14) are crucial in understanding differential exposure ence the experience of two moments of reckoning: the to stressors [36–38]. moment children realize that being members of Sociologists assert that stressful events and chronic a minority status group represents difference and the stressors are entrenched in the structural contexts of moment children realize their position relative to people’s lives [34]. It is difficult to separate social and others, their label and the defined power dynamic. economic status from the likelihood of experiencing For minorities, coincident with the labelling are the stressors and strains. Individuals occupying disadvan- other distinct processes of stigmatization (i.e. the taged social positions experience specific experien- execution of disapproval, rejection, exclusion and tial, environmental and financial hardships. For discrimination) [29]. example, exposure to economic hardship, whether Finally, the work (occupational) context is often the somewhat temporary or chronic over the life course, most dominant setting in adulthood for socialization heightens exposure to life strains while, at the same beyond the family. It is organized around a bureau- time, limiting people’s ability to manage these cratic structure, and the socialization of adults in this strains. [39] context varies by the degrees of autonomy, confor- Stressors are evidently not randomly distributed mity, supervision, routinization and complexity ex- throughout the population [36,40,41]. Exposure to perienced [30]. Certain jobs confer access to status, stressors is an inevitable consequence of status prestige, knowledge and other resources that enhance arrangements and resulting inequalities that affect

CONCEPTS AND CHALLENGES IN CAPTURING THE SOCIAL ENVIRONMENT 69 access to power, resources and opportunities, and participation that fails to provide the expected re- status [35]. Two major pathways linking structure turns [35]. These are the routes through which with individual experiences of stressors involves ex- macro-structures are connected to micro-conditions clusion from full participation in the social system and in peoples’ lives [42]. 6.3 AREA-LEVEL FACTORS AND MENTAL HEALTH There has been an increase in attention to situating as gender, age and ethnicity [49,50]. Some evidence individuals in their social contexts by attempting to suggests that childhood psychosocial problems occur capture specific aspects of the environments in which more frequently in deprived areas, and that only a they live and work. Many of these studies focus on fraction of the differences between areas is explained neighbourhoods, communities and other areas in re- by individual and family characteristics [51]. How- lation to mental health [43,44]. In this section, we ever, as with studies of adult samples, it is unclear begin with a discussion of recent conceptualizations whether or not such associations reflect social drift of the wider social environment in mental health (e.g.parentswithpsychiatricproblemsmovingtocertain research, commonly focused on depression and psy- areas in the case of children) or suggest social causation chosis. Particular attention is paid to social capital as (the aetiological influence of social environment). an example of the narrow view that is taken regarding In addition, sociological theories of how the en- complex and dynamic social processes. vironment comes to influence the individual involve a Across disciplines, there is an interest in the cas- focus on how the effects of the community and cading effects of social context on individual mental neighbourhood influence communal care, which in health (see March et al. [43] and Kim [44] for recent turn impacts on the behaviour of children and ado- reviews). Known continuities and discontinuities in lescents. For example, social disorganization theory mental health over the life course [45] challenge the emphasizes the effect of weak social control net- notion that capturing current environment (i.e. at the works on adolescent behavioural outcomes through time outcomes are measured) is sufficient [46]. How- residential instability, segregation, isolation and low ever, with very few exceptions (e.g. see Wheaton and socioeconomic status (SES) [52–54]. (See Jencks and Clarke [47]), studies have failed to consider the in- Mayer [55] and Leventhal and Brooks-Gunn [56] for dividual life course and the history of the places where reviews of the theoretical frameworks and conceptual individuals live and work. For example, evidence models that consider the influence of place on child suggests that middle childhood may be a sensitive development.) Leventhal and Brooks-Gunn [56] period during which children are at the greatest risk of identified three broad factors that potentially shape being influenced by deleterious neighbourhoods (see children’s behaviour: lack of institutional resources Ingoldsby and Shaw [48] for a review). Each indivi- (e.g. health care and day care), lack of norms and dual has a history of exposure to a particular place, and collective efficacy in the neighbourhoods and com- often many places – the social context is embedded munities, and child–parent relationships. The latter is over time. Associations between an individual’s men- likely to be heavily influenced by the intergenera- tal health and the social environment is subject to the tional transfer of economic, social and health diffi- cautions applicable in any study that fails to consider culties that may result in psychosocial problems. For sequence and timing of exposure – i.e. whether mental example, there are likely to be gender differences in health is a cause or a consequence of exposure to the way parents socialize their children and adoles- particular environs. cents to engage in their social environment. Evidence From a developmental psychology perspective, suggests that girls tend to be more closely supervised Bronfenbrenner emphasizes multiple, interrelated in- by their parents, and low maternal monitoring of fluences – family functioning, school-related factors, behaviour is linked to the occurrence of externalizing peers and communities as well as social statuses such behaviour [57].

70 PERSPECTIVES AND METHODS Most sociological and psychological models pro- the previous six months. However, no association was pose that the effects of neighbourhoods on child and observed between reported cleanliness of sidewalks adolescent health and behaviour outcomes are indir- and streets and depression [59]. ect, operating through more proximal behaviours [56]. Research concerning theimpactofcontexton Potential direct effects are not generally ignored. disorders more prevalent in the general population, There is some consideration that the more embedded specifically depression, tends to focus on character- an individual becomes in a neighbourhood or com- istics of material and physical environments [44]. munity over the life course, the greater the possibility For example, the built environment includes the of direct effects of the neighbourhood (e.g. see Samp- state of buildings and housing, as well as the density son [53]). Moreover, recent research has shown that of supermarkets and green grocers, the availability of context is transmitted across generations [58]. Thus, basic amenities that facilitate social interactions, there may be generations of individuals that are ex- such as community centres and cafes, and areas that posed to direct deleterious exposures at the neighbour- make physical activity more viable, such as green hood level. spaces. Turning to adult mental health, a limited number Another rapidly strengthening strand of research of studies have explored the relation between the regarding the impact of the social environment on social environment and depression [44]. A recent mental health has focused on psychosis (see March systematic review conducted by Kim [44] focuses on et al. [43] for a review). There is an extensive history the most commonly studied neighbourhood charac- of research considering the role of the social envir- teristics, such as neighbourhood-level SES, built en- onment in the aetiology and development of psycho- vironment, internal and external physical features, tic disorders. However, most of this research has local services, amenities, social disorder and social relied on various concepts under the umbrella of capital. ‘wider social environment’. This generally refers to Neighbourhood SES is captured using an aggregate neighbourhood-level dimensions of deprivation, re- of household income, education, or occupational sta- sidential stability (e.g. housing tenure, migration), tus, usually at the level of the census tract. An assess- household/family structure and ethnic composition or ment of the collective economic deprivation facing density of an area [60]. However, it is unclear why or residents bound together is important. However, how these particular aspects of the social environ- neighbourhoods and communities where there is more ment are linked under this umbrella term. While integration across social class groups should attempt these aspects are important and in some way inter- to accurately document any benefits of economic related, there has been a paucity of theoretically desegregation. informed research in the past decade. Moreover, The built environment can refer to structural attri- other potentially relevant social and contextual fac- butes of a community, such as the state of housing or tors have received scant consideration. For example, buildings in a community or the presence and proxi- the influence of the physicality of place (e.g. the built mity of green spaces. The internal and external fea- environment and material resources) has been under- tures of the social environment can be useful for studied, despite this being a more direct way to thinking about how specific aspects of the immediate capture the most proximal area deprivation [43]. In physical environment may influence mental health addition, macro-level social processes (e.g. policies (e.g. see Galea et al. [59]). The features of an internal resulting in structural inequalities) and historical physical space can take into account the state of basic circumstances (e.g. racism resulting from particular amenities, such as nonshared bathroom and kitchen legacies of slavery, colonization, deindustrialization facilities or access to central heating. The immediate and economic downturns resulting in increased un- external space may also be important to consider. For employment rates and downward mobility) that example, in densely urbanized New York City the shape social context are seldom considered in think- condition of the building of residence and the sur- ing about how the wider social environment and rounding buildings was associated with depression in relevant exposures are formed. By neglecting theory

CONCEPTS AND CHALLENGES IN CAPTURING THE SOCIAL ENVIRONMENT 71 and history in construing what may be aetiologically may elide potentially beneficial community-level important about the wider social environment, we interventions as we move forward. 6.4 ONE APPROACH DOES NOT FIT ALL: THE EXAMPLE OF SOCIAL CAPITAL In most mental health-related research, social capital Coleman’s thinking stems from the rational choice generally refers to the features of social organization, school of thought. One of the main tenets of Rational specifically as it relates to social trust, civic participa- Choice Theory simply states that individuals choose tion and norms of reciprocity that facilitate coopera- actions that will maximize utility or satisfy specific tion for mutual benefit [61]. Briefly, social capital has needs and wants [64]. Here, social capital is defined by become a way of describing social processes that its function. The elements of social capital overlap shape interactions between individuals, groups or with some aspects of social structure, facilitate certain communities. Thus, variation in social environments actions within the structure and are embodied in may warrant consideration of different approaches to relationships between individuals and among indivi- social capital. duals. Finally, and possibly most important, social One striking aspect of the application of this con- capital that is valuable to one individual may be struct in research on psychiatric outcomes is the fail- useless to another [64]. ure to consider the differences in conceptualizations The most commonly used conceptualization of of social capital. While health research often utilizes social capital in research on psychiatric outcomes the aforementioned definition [61], there has been appears to be Putnam’s [65] definition, which focuses little application of other perspectives and approaches on the collective or the ‘civil society’ (briefly outlined rooted in different social theories and schools of above). From this perspective, social capital refers to thought. We describe each of the major theoretical the features of social life (e.g. social networks, norms perspectives on social capital below. This list is by no and trust) that enable participants in a social environ- means exhaustive, but is meant for critical considera- ment to act together more effectively to pursue shared tion. The question remains: Is it appropriate in every objectives. While this perspective is useful in general social milieu to consider the presence of social trust, terms when considering notions of how individuals civic participation and norms of reciprocity that facil- interact in a society, it has been criticized for not itate cooperation for mutual benefit? From whose considering the complex economic relationships in perspective is this being assessed and judged? Are societies that marginalize certain groups [66]. these mutually beneficial across groups? Finally, specifically in contrast to Putnam, Bourdieu, a social theorist known for his focus on Portes [66] places emphasis on distinguishing the agency (micro-level individual actors) and structure resources used to represent social capital from the (large-scale social structures, e.g. broadly referring ability to obtain them by membership within different to the systems within which individuals relate), social status groups. According to Portes [66], ‘social places emphasis on instrumental aspects of social capital stands for the ability of actors to secure capital and on benefits to the individual [62]. His benefits by virtue of membership in social networks definition of social capital comprises (1) the social or other social structures’ ([66], p. 6). Portes extends relationships that allow individuals to access re- Coleman’s thinking about the false assumptions con- sources and (2) the amount and quality of those cerning the universal benefit of social capital. He resources. From this perspective, these exchanges notes that the beneficial strong ties for some in a require deliberate investment of both economic and community or neighbourhood result in the exclusion cultural resources, may lead to access to economic of others. resources and are likely to increase cultural capital Portes is one of the few theorists to discuss the (institutional or embodied) [62]. Indeed, capital may negative consequences of social capital [66]. He be transferred [63]. argues that a focus on collective norms can restrict

72 PERSPECTIVES AND METHODS individual freedom and increase demands for confor- social and health inequalities have fragmented com- mity. As a result, there is a downward push of the munities and silenced members of disadvantaged norms of the majority, especially for social groups groups. With the growing emphasis on social exclu- characterized by marginalization and opposition to the sion – a somewhat nebulous term that generally is mainstream. For example, some researchers have defined as a lack of social, cultural and political considered voter turnout (or lack of) as an indicator participation [67] – this distinction deserves further of social capital in urban social environments, where consideration. 6.5 SOCIAL AND GEOGRAPHICAL MOBILITY In the final section of this chapter, we return to the term socialization is implied in this research, but not idea that consideration must be given to past and directly assessed. Changes in social distance and present positions in social structures. Mobility is social space often require newly formulated strategies broadly defined as movement from one distinct that result in adaptation and assimilation, particularly social environment to another, both directly through in cases of geographical mobility. Also, the question a change in physical environment (geographical mo- remains whether or not there are diminishing returns bility) and more indirectly through a change in from upward mobility, specifically as it relates to social status position (social mobility). With social intergenerational mobility. mobility may come geographic mobility, and vice One important aspect of mobility is the timing of a versa. In some studies, mobility of some type is shift from one social environment to another. Distinc- represented by residential instability, measured in- tions made about timing should refer to the intersec- directly through information on housing tenure and tion between biographical and historical transitions migration status. (e.g. during childhood versus adulthood, first genera- What happens to developed perceptions and eva- tion versus second generation), with emphasis on luations associated with one social environment when historical context and social change [70]. Individuals there is a notable shift in the type of social interaction are likely to differ in their responses to change in status and/or environment? Both social and geographical depending on the type and the timing of the mobility mobility require a certain level of adaptability. For during the life course. Is it more difficult for an adult some, these transitions are made with great ease; for than a child or adolescent to reconcile the loss in status others, mobility comes with a strong internal conflict and the shift in schema or habitus, or perhaps the that may not prove beneficial to health, particularly reverse? One recent meta-analysis of research regard- mental health. With respect to habitus, individuals can ing migration and schizophrenia [71] indicates that have an ‘inappropriate habitus’; i.e. a person may not second-generation migrants may have a higher risk of possess the appropriate habitus for their current situa- psychosis than first-generation migrants. One hypoth- tion in the social world [21]. esis generated by this evidence is that the duration of There is a paucity of research focused on associa- exposure to discrimination in the host contextis longer tions between social mobility trajectories and mental for second-generation migrants than for first-genera- health. What evidence there is suggests that upward tion migrants. Viewed through the lens of habitus, social mobility trajectories over the life course can be perhaps second-generation migrants repeatedly shift beneficial for mental health [68] and downward mo- schemas from their parents’ upbringing, since parents bility or loss of status over time may be detrimental for tend to socialize their children in the ways in which mental health [69]. they were socialized, to the majority norms of the Migration receives the most attention in recent context into which they were born. This could be research (e.g. see Reference [71]). The difficulty in stressful at the critical time when adolescents transi- managing the unfamiliarity and challenges presented tion to adulthood and enter the period of risk for by a new setting with old schemas resulting from long- psychosis.

CONCEPTS AND CHALLENGES IN CAPTURING THE SOCIAL ENVIRONMENT 73 In the case of geographical mobility, the migration residence in an extended network of other migrants of a particular group generally hinges on status-based (i.e. enclaves) may confer protective effects. For ex- resources and opportunities. For example, movement ample, ethnic density (the proportion of a particular is more common among individuals of a minority ethnic group in a given geographic location) has been group status that are either of a higher, more privi- shown to confer protection from psychosis across leged, status (possibly more educated), a more so- settings, even in the most deprived areas [72,73]. One cially integrated member of a larger supportive net- hypothesized mechanism is that of a buffering effect, work of other migrants, or both. In the case of the which mitigates the impact of discrimination from former, it is highly likely that they are moving from majority populations, though this remains to be tested. being a member of the majority group in their im- Moreover, the exact timing of this effect remains mediate social milieu to occupying a disadvantaged unknown, as the exposure has been measured around minority status position in the new society. However, illness onset. 6.6 CONCLUSION The wider social environment is dynamic, complex tapestry woven by social psychiatry and psychiatric and historically shaped. When considering how the epidemiology. Possible research questions include: social environment may impact an individual’s mental (1) Do intergenerational shifts from a schema of one health, the body and its state of health should be habitus to that of a different habitus impact mental considered as concrete representations of lived ex- health and (2) under which mobility conditions (e.g. perience situated in a sociohistorical and health-re- geographical, social) does a change in habitus affect lated context over the life course. With the guidance of mental health? Furthermore, does a disjuncture or social theory, as we have illustrated with the concept dissonance between one set of embedded cognitive of habitus,we may begin to understand how, across the dispositions and actions and the development of a new life course, social structures become psychologically habitus potentially result in the expression of patho- and biologically embedded in individuals. By high- logical symptoms? Do shifts in habitus potentially lighting the relative positioning of individuals within work against our achieving the benefits of upward their social environments, the shifting salience of mobility and exacerbate the impact of downward social context across the life course and ultimate mobility? In answering some of these questions, embodiment of social structures, we hope to stimulate perhaps we can narrow in on the wider social envir- strands of empirical research that may enrich the onment and its effects on mental health. ACKNOWLEDGEMENTS Stephani Hatch is supported by the Biomedical Re- 2. House, J. (1992) Social structure and personality, in search Centre for Mental Health at the Institute of Social Psychology: Sociological Perspectives, 2nd edn Psychiatry, Kings College London and The South (eds M. Rosenberg and R. H. Turner), Transaction London and Maudsley NHS Foundation Trust. Publishers, London, pp. 525–561. 3. Schooler, C. (1996) Cultural and social-structural explanations of cross-national psychological REFERENCES differences. Annual Review of Sociology, 22, 323–349. 1. Reay, D. (2004) ‘It’s all becoming a habitus’: beyond 4. Braverman, P. (2006) Health disparities and health the habitual use of habitus in educational research. equity: concepts and measurement. Annual Review of British Journal ofSociology of Education, 25,431–444. Public Health, 27, 167–194.

74 PERSPECTIVES AND METHODS 5. Berkman, L. F. and Kawachi, I. (2000) Social Sociological Perspectives, 2nd edn (eds M. Rosenberg Epidemiology: Rediscovering the Role of the and R. H. Turner), Transaction Publishers, London, pp. Physical and Social Environment, Oxford University 133–164. Press, Oxford. 18. Sewell, W. H. (1963) Some recent developments in 6. Marmot, M. and Wilkinson, R. G. (2005) Social socialisation theory and research. The Annals of the Determinants of Health, 2nd edn, Oxford University American Academy of Political and Social Science, Press, Oxford. 349, 163–181. 7. Bourdieu, P. (1989) Social space and symbolic power. 19. Bourdieu, P. (1977) Outline of a Theory of Practice, Sociological Theory, 7, 14–25. University Press, Cambridge. 8. Lindbladh, E., Lyttkens, C. H., Hanson, B. S. et al. 20. Bourdieu, P. (1990) Other Words: Essays Toward a (1996) An economic and sociological interpretation of Reflexive Sociology, Polity Press, Cambridge. social differences in health related behaviour: an encounter as a guide to social epidemiology. Social 21. Ritzer, G. (2000) Modern Sociological Theory, McGraw Hill, Boston, Massachusetts. Science and Medicine, 43, 1817–1827. 22. Helms, J. E. (1990) Black and White Racial Identity: 9. Bourdieu, P. and Wacquant, L. (1992) An Invitation Theory, Research, and Practice, Greenwood, New York. to Reflexive Sociology, University of Chicago Press, Chicago, Illinois. 23. Spencer, M. and Markstrom-Adams, C. (1990) Identity processes among racial and ethnic minority children in 10. Cicourel, A. V. (1993) Aspects of structural and processual theories of knowledge, in Bourdieu: America. Child Development, 61, 290–310. Critical Perspectives (eds C. Calhoun, E. Lipuma 24. Hughes, D. and Chen, L. (1997) What parents tell and M. Postone), Polity Press, Cambridge, pp. 89–115. children about race: an examination of race-related socialization among African American families. 11. Weininger, E. (2004) Foundations of Pierre Bourdieu’s class analysis, in Approaches to Class Analysis (ed. Applied Developmental Science, 1, 200–214. E. O. Wright), Cambridge University Press, Cambridge, 25. Hughes, D. (2003) Correlates of African American and pp. 82–118. Latino parents’ messages to children about ethnicity and race: a comparative study of racial socialization. 12. Corsaro, W. A. and Eder, D. (1995) Development and American Journal of Community Psychology, 31, socialization of children and adolescents, in 15–33. Sociological Perspectives on Social Psychology (eds K. S. Cook, G. A. Fine and J. S. House), Allyn and 26. Miller, D. B. and MacIntosh, R. (1999) Promoting Bacon, London, pp. 421–451. resilience in urban African American adolescents: racial socialization and identity as protective factors. 13. McLean, C., Campbell, C. and Cornish, F. (2003) Social Work Research, 23, 159–169. African-Caribbean interactions with mental health services in the UK: experiences and expectations of 27. DiMaggio, P. (1979) Review essay: on Pierre Bourdieu. exclusion as (re)productive of health inequalities. American Journal of Sociology, 84, 1460–1474. Social Science and Medicine, 56, 657–669. 28. Nash, R. (1990) Bourdieu on education and social and 14. Morgan, C., Mallett, R. M., Hutchinson, G. et al. (2004) cultural reproduction. British Journal of Sociology of Negative pathways to psychiatric care and ethnicity: the Education, 11, 431–447. bridge between social science and psychiatry. Social 29. Link, B. G. and Phelan, J. C. (2000) Evaluating the Science and Medicine, 58, 739–752. fundamental cause explanation for social disparities in 15. Richman, L., Kohn-Wood, L. and Williams, D. (2007) health, in Handbook of Medical Sociology, 5th edn (eds The role of discrimination and racial identity for mental C. E. Bird, P. Conrad and A. Fremont), Prentice Hall, health service utilization. Journal of Social and Clinical Englewood Cliffs, New Jersey, pp. 33–46. Psychology, 26, 960–981. 30. Kohn, M. (1980) Job complexity and adult personality, 16. Gecas, V. (1992) Contexts of socialisation, in Social in Themes of Love and Work in Adulthood (eds N. Psychology: Sociological Perspectives, 2nd edn (eds Smelser and E. Erikson), Harvard University Press, M. Rosenberg and R. H. Turner), Transaction Cambridge, Massachusetts, pp. 193–212. Publishers, London, pp. 165–199. 31. Pearlin, L. I., Menaghan, E., Lieberman, M. et al. 17. Bush, D. M. and Simmons, R. G. (1992) Socialisation (1981) The stress process. Journal of Health and processes over the life course, in Social Psychology: Social Behavior, 22, 337–356.

CONCEPTS AND CHALLENGES IN CAPTURING THE SOCIAL ENVIRONMENT 75 32. Aneshensel, C. and Phelan, J. (1999) The sociology of between childhood and adult life. Journal of Child mental health: surveying the field, in Handbook of the Psychology and Psychiatry, 47, 276–295. Sociology ofMental Health (eds C.S. Aneshensel and J. 46. March, D. and Susser, E. (2008) Social context and Phelan), Kluwer Academic/Plenum, New York, pp. developmental psychopathology, in Developmental 3–18. Psychopathology and Wellness: Genetic and 33. Link, B. G. and Phelan, J. (1995) Social conditions as Environmental Influences (ed. J. Hudziak), American fundamental causes of disease. Journal of Health and Psychiatric Publishing, Inc., Washington, DC, pp. Social Behavior, 35 (Extra Issue: Forty Years of 49–64. Medical Sociology: The State of the Art and 47. Wheaton, B. and Clarke, P. (2003) Space meets time: Directions for the Future), 80–94. integrating temporal and contextual influences on 34. Pearlin, L. I. (1989) The sociological study of stress. mental health in early adulthood. American Journal of Health and Social Behavior, 30, 241–256. Sociological Review, 68, 680–706. 35. Aneshensel, C. S. (1992) Social stress: theory and 48. Ingoldsby, E. M. and Shaw, D. S. (2002) Neighborhood research. Annual Review of Sociology, 18, 15–38. contextual factors and early-starting antisocial pathways. Clinical Child and Family Psychology 36. Turner, R. J., Wheaton, B. and Lloyd, D. A. (1995) The Review, 5, 21–55. epidemiology of stress. American Sociological Review, 49. Bronfenbrenner, U. (1986) Ecology of the family as a 60, 104–125. context for human development: research perspectives. 37. Kessler, R. C. and MacLeod, J. (1984) Sex differences Developmental Psychology, 22, 723–742. in vulnerability to undesirable life events. American Sociological Review, 49, 620–631. 50. Bronfenbrenner, U. (1992) Ecological systems theory, in Six Theories of Child Development. Revised 38. Turner, R. J. and Avison, W. (2003) Status variation in Formulations and Current Issues (ed. R. Vasta), stress exposure: implications for the interpretation of Jessica Kingsley Publishers, London. research on race, socioeconomic status, and gender. 51. Reijneveld, S. A., Brugman, E., Verhulst, F. C. et al. Journal of Health and Social Behavior, 44, 488–505. (2005) Area deprivation and child psychosocial 39. Pearlin, L. I. (1999) The stress process revisited, in problems:a national cross-sectional study among Handbook of the Sociology of Mental Health (eds C. S. school-aged children. Social Psychiatry and Aneshensel and J. Phelan), Kluwer Academic/Plenum, Psychiatric Epidemiology, 40, 18–23. New York, pp. 395–415. 52. Sampson, R. J. and Groves, W. B. (1989) Community 40. Kessler, R. C. (1979) A strategy for studying structure and crime: testing social-disorganization differential vulnerability to the psychological theory. American Journal of Sociology, 94, 774–802. consequences of stress. Journal of Health and Social 53. Sampson, R. J. (1993) Family and community-level Behavior, 20, 100–108. influences of crime: a contextual theory and strategies 41. MacLeod, J. D. and Kessler, R. C. (1990) for research testing, in Integrating Individual and Socioeconomic status differences in vulnerability to Ecological Aspects of Crime (eds D. P. Farrington, undesirable life events. Journal of Health and Social R. J. Sampson and P.-O. H. Wikstr€ om), National Behavior, 31, 162–172. Council for Crime Prevention, Stockholm, pp. 42. Hatch, S. L. (2007) Economic stressors, social 426–444. integration, and drug use among women in an 54. Sampson, R. J. and Laub, J. H. (1994) Urban poverty inner city community. Journal of Drug Issues, 37, and the family context of delinquency: a new look at 257–280. structure and process in a classic study. Child 43. March, D., Hatch, S. L., Morgan, C. et al. (2008) Development, 65, 523–540. Psychosis and place. Epidemiological Reviews, 30, 55. Jencks, C. and Mayer, S. (1990) The social 84–100. consequences of growing up in a poor neighborhood, 44. Kim, D. (2008) Blues from the neighborhood? in Inner-City Poverty in the United States (eds L. E. Neighborhood characteristics and depression. Lynn and M. F. H. McGeary), National Academy Press, Epidemiologic Reviews, 30, 101–117. Washington, DC, pp. 111–186. 45. Rutter, M., Kim-Cohen, J. and Maughan, B. (2006) 56. Leventhal, T. and Brooks-Gunn, J. (2000) The neighbor- Continuities and discontinuities in psychopathology hoods they live in: the effects of neighborhood residence

76 PERSPECTIVES AND METHODS on child and adolescent outcomes. Psychological 66. Portes, A. (1998) Social capital: its origins and Bulletin, 126, 309–337. applications in modern sociology. Annual Review of 57. Kim, J. E., Hetherington, E. M. and Reiss, D. (1999) Sociology, 24, 1–24. Associations among family relationships, antisocial 67. Morgan, C., Burns, T., Fitzpatrick, R. et al. (2007) peers, and adolescents’ externalizing behaviors: Social exclusion and mental health: conceptual and gender and family type differences. Child methodological review. British Journal of Development, 70, 1209–1230. Psychiatry, 191, 477–483. 58. Sharkey, P. (2008) Intergenerational transmission of 68. Poulton, R., Avshalom, C., Milne, B. J. et al. (2002) context. American Journal of Sociology, 113, 931–969. Association between children’s experience of 59. Galea, S., Ahern, J., Rudenstine, S. et al. (2005) Urban socioeconomic disadvantage and adult health: a life- built environment and depression: a multilevel analysis. course study. Lancet, 360, 1640–1645. Journal of Epidemiology of Community Health, 59, 69. Tiffin, P. A., Pearce, M. S. and Parker, L. (2005) Social 822–827. mobility over the life course and self-reported 60. Allardyce, J. and Boydell, J. (2006) Review: the wider mental health at age 50: prospective cohort study. social environment and schizophrenia. Schizophrenia Journal of Epidemiology and Community Health, Bulletin, 32, 592–598. 59, 870–872. 61. Kawachi, I., Kennedy, B. P. and Glass, R. (1999) Social 70. ElderJr,Glen H. (1994) Time,humanagency,and social capital and self-rated health: a contextual analysis. change: perspectives on the life course. Social American Journal of Public Health, 89, 1187–1193. Psychology Quarterly, 57, 4–15. 62. Bourdieu, P. (1986) Forms of Capital, Free Press, New 71. Cantor-Graae, E. and Selten, J. (2005) Schizophrenia York. and migration: meta-analysis and review. American Journal of Psychiatry, 162, 12–24. 63. Bourdieu, P. and Wacquant, L. (1992) An Invitation to Reflexive Sociology, The University of Chicago Press, 72. Boydell, J., van Os, J., McKenzie, K. et al. (2001) Chicago, Illinois. Incidence of schizophrenia in ethnic minorities in London: ecological study into interactions with 64. Coleman, J. S. (1988) Social capital in the creation of environment. British Medical Journal, 323, 1–4. human capital. American Journal of Sociology, (suppl.), S95–S120. 73. Veling, W., Susser, E., van Os, J. et al. (2008) Ethnic density of neighborhoods and incidence of psychotic 65. Putnam, R. (2000) Bowling Alone: The Collapse and disorders among immigrants. American Journal of Revival of American Community, Simon and Schuster, Psychiatry, 165, 66–73. New York.

7 Qualitative research methods Joanna Murray Institute of Psychiatry, King's College London, London, UK 7.1 INTRODUCTION The arrival of qualitative methods in the mainstream qualitative and quantitative research are based on of health care research was demonstrated by the empirical evidence (the data collected), it remains a publication in the British Medical Journal (BMJ) of fallacy in the health sciences that the two approaches a series of short methodological papers in the mid- are directly opposed and incompatible [14]. The 1990s [1–4]. These papers set out to de-mystify traditional methodological divide was bridged by approaches to the collection and analysis of nonnu- recognition of the contributions each can make to meric data and to explain the benefits of incorporating improving health care and also by the requirements qualitative methods into research on health and ill- of research funders, who increasingly view multi- ness. Efforts were made to demonstrate that metho- disciplinary and mixed methods research as the most dological rigour was perfectly achievable in qualita- productive in applied health sciences [15,16]. In tive research if systematic processes were applied at terms of the contribution qualitative methods can all stages of the study [5] and adequately described in make to epidemiology and treatment trials, research publications [6]. Further evidence for the integration protocols now commonly include the following qua- of qualitative methods in health research came from litative elements: the publication of six papers in the BMJ addressing theoretical approaches and critical appraisal [7–12]. . Identification of concepts of illness and develop- Much has been written about the fundamental ment of conceptual frameworks theoretical differences between qualitative and quan- titative research and it is not the purpose of this . Clarification of cultural variations in language and chapter to rehearse these arguments, nor to describe modes of expression the history and variety of theoretical approaches to qualitative enquiry [13]. The aim is to provide a . Development of theory prior to framing hypotheses practical introduction to qualitative research meth- ods in mental health and to illustrate their role in . Development, translation and validation of research understanding how events are shaped. Although both tools Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

78 PERSPECTIVES AND METHODS . Investigation of anomalies in the findings of surveys measurement of pain might seem to be a relatively and treatment trial. straightforward matter, requiring the use of valid and reliable self-report scales. However, the sub- The main difference between qualitative and quan- jective meaning and significance of the pain and the titative methodologies stems from the approach to attribution of causes are variable and subject to data analysis: while quantitative research relies upon a wide range of social and cultural influences. statistical analysis of numerical data, qualitative Decisions on seeking help may involve others who analysis is explicitly interpretative. Consequently the have their own perspective on the problem. Thus the sampling of data differs between the two methods, experience of symptoms and ways of responding are with random sampling in quantitative studies to not simply matters of medical concern but are achieve representative subgroups of the population personal and social phenomena involving individual contrasting with purposive sampling in qualitative beliefs and social norms. Qualitative research offers studies to identify research participants who possess an empirical approach to investigating the ‘lived the characteristics of relevance to the research experience’ of illness and, by interpretation of question. descriptive accounts, to generate theories concern- The two methods are not conflicting: they simply ing the relevant processes. In mental health this offer the appropriate means of answering different approach can contribute to the provision of more types of question. For example, while quantitative appropriate interventions and accessible services. methods are required to measure the prevalence The patient and/or carer is situated within his/her andriskfactors forlowerbackpain in a defined immediate social context, and by making the population, qualitative methods would be necessary patient’s (or carer’s) perspective central to investi- to investigate the experience of back pain and to gating mental health problems, the experience of develop theory on ways of coping with it. The treatment and the quality of care can be improved. 7.2 THE QUALITATIVE RESEARCH APPROACH One of the central tenets of qualitative research is the detailed interviewing and observation, the collection ‘bottom-up’ approach in which the perspective of of ‘thick’ or ‘rich’ data [17]. Although the divergence research participants is central to the process of data between the two approaches is not always clear-cut, it collection, analysis and theory development. It fol- stems from the centrality of deduction (theory testing) lows that the context in which to collect and under- in quantitative methods and induction (theory devel- stand the data should, ideally, be the natural setting in opment) in qualitative methods and the consequent which the phenomenon would be discussed or ob- differences in sampling, analysis and reporting of served (e.g. the participant’s own home rather than a findings. In general, qualitative methodology seeks medical setting). This contrasts with the ‘top-down’ to develop understanding of phenomena and/or to approach in quantitative studies where the research- construct theory by a process of induction. The focus ers predetermine the scale and content of the inves- is on analysing and making sense of rich data from tigation from a review of existing evidence. The relatively few selected cases, as a basis for developing setting for data collection is unlikely to be considered theory. In contrast, quantitative methodology is pre- influential. dominantly concerned with testing hypotheses derived Nonetheless, qualitative and quantitative research- from existing theory. Sampling of cases for quantita- ers share a common empirical approach, involving the tiveanalysis is based on probability, representativeness systematic collection and analysis of data. Both are and statistical significance. However, this apparent concerned with the individual’s point of view, but methodological divide is not always helpful and is qualitative researchers believe they can best under- often crossed for sound empirical reasons. Qualitative stand and represent the individual’s perspective by methods may be used in hypothesis testing, for

QUALITATIVE RESEARCH METHODS 79 instance, when nonnumerical data are collected develop theory to explain phenomena by analysing by incorporating open-ended questions into survey and interpreting individual accounts of experiences. questionnaires. In this instance, responses may be The main qualitative methods used in mental health systematically coded into categories by content ana- research, individual interviews, focus group discus- lysis, displayed as frequency distributions and simple sions and participant observation will be described. measures of association applied. The application of these different methods will be This chapter will focus on inductive methods, the illustrated with examples drawn from a variety of ‘bottom-up’ approach in which the researcher aims to mental health settings. 7.3 QUALITATIVE METHODS Data for qualitative analysis appear in many forms researcher’s role is collaborative in facilitating an other than the conventional transcribed interview be- account of the participant’s experiences and beliefs. tween researcher and participant. In some studies no It is important for the researcher to stress that the researcher is involved in the data recording; the ma- participant is the expert. Perceived social identities terial for analysis may be published accounts in news- are important and the researcher must establish good papers, narratives from fiction, biographical stories, rapport from the outset, explaining the context for the diaries, letters, video recordings and even televised study and their own perspective. This may concern soap operas. However, the typical method of data their professional background or personal experience collectionistheface-to-faceinterviewwhereresearch- of the research topic. The researcher should remain ers engage with their subject to produce a personal aware that their own characteristics may influence the account of the participant’s experiences and attitudes. course and content of the interview. Empathic listen- ing and tactful probes are essential to avoid excessive and unwanted intrusion. An explanation of the im- 7.3.1 Individual interviews portance of the research and the participant’s con- tribution and a full account of ethical procedures Interviews are a familiar aspect of modern life, often should be given. used to control access to resources and advancement. Interview procedures are flexible: pre-framed open- They are used to assess medical symptoms, eligibility ended questions may be used alongside prompts and for benefits, suitability for training, college entry or a probes for further development of the topic. Simple specific job. In these cases the interviewer is generally topic guides with alist ofitems fordiscussion can form in a position of authority and has the power to act on the basis of a conversational style of interview, though the information elicited from the interviewee. The this may be unsuited to the use of multiple inter- outcome may mean significant gain or loss for the viewers. Vignettes that tell a story relevant to the interviewee. In other circumstances the interviewee research topic are useful in generating discussion of may seek a more powerful role: celebrities and concepts that are sensitive or likely to be influenced authors who are taking part in an interview to promote by cultural variations (e.g. beliefs about depres- their latest book or film; politicians seizing the op- sion [18–20]). The interview should, if possible, be portunity to promote a policy. However, interviews audio-recorded so that a full and nuanced transcription may contain an element of coercion, where the inter- and the recording itself is available for analysis. viewee is reluctant to divulge information but may be Taking detailed notes during the interview is a sig- compelled to do so, as in a police interview of a nificant distraction to both parties and prevents close suspect. In contrast, the qualitative interviewer aims attention to the participant’s narrative. Key topics that to empower the participant by explicit focus on their arise during the interview that require further explora- expertise in the area of enquiry and by arranging the tion may be overlooked. There is no substitute for meeting in the most natural setting available. The verbatim quotes to support the interpretations.

80 PERSPECTIVES AND METHODS 7.3.2 Focus group discussions would endeavour to recruit a socially and culturally diverse sample. This would be a purposive or the- Qualitative researchers use focus group discussions to oretical sampling strategy in that the characteristics produce data and insights that would be less accessible of participants help to refine and delimit the general- without the interaction of the group. Data from the izability of the research findings. However, we also discussion is qualitatively different from that found in need to consider how much diversity of character- a series of individual narratives. The explicit aim of istics is conducive to intragroup interaction. convening the group is to encourage the type of interaction that would occur in everyday life but with . Should we aim for homogeneity or heterogeneity? greater focus, allowing researchers access to For many studies social norms of behaviour are the ‘ideologies, practices and desires amongst specific focus of enquiry. In this, bringing together a group groups of people’ [21]. For this reason focus group who share important key characteristics, such as discussions are particularly suited to exploring and age, gender, ethnicity, professional background, is understanding social norms and cultural differences in essential. However, if we were investigating the mental health research. Why, for example, do some experiences of users of a particular service, homo- people seek help for psychological difficulties while geneity would arise from their shared contact with others are extremely reluctant to accept professional that service and any diversity in demographic char- intervention? The following study by Prior et al. [22] acteristics of group participants would be less im- illustrates the advantages of focus groups in investi- portant. Including participants from different pro- gating concepts of mental health. fessional backgrounds within a single focus group Against a background of substantial research lit- would tend to militate against free discussion of erature on the widespread reluctance to disclose attitudes and social norms. symptoms of emotional distress to health profes- sionals, Prior and colleagues set out to investigate . Should participants be unknown to each other? lay concepts of common mental disorders. They There is no reason why participants should not be chose focus groups as especially appropriate for members of a pre-existing group of work collea- studying public attitudes to disclosure as it would gues, neighbours or social group, as this allows for enable them to judge the willingness of people to the discussion to take place in a natural setting. disclose a psychiatric history and the reactions of However, the sensitivity of the topic will have a others to the disclosures. In other words, they aimed bearing on the appropriateness of tapping into a pre- to explore public responses to sensitive issues rather existing group. than individual private experiences that would be more appropriately explored through individual in- . How many participants and where should they depth interviews. Their findings were surprising: meet? Five to ten participants is generally regarded stigma was not an important factor in disclosure; as the most effective for exploration of the topic. more important was the way that lay people con- Recruitment problems are a major source of failure structed common mental disorders as problems of of focus groups so it is advisable to recruit more living for which they considered a medical consulta- people than required and assume that some will not tion to be inappropriate. attend. The venue will depend on the study but The composition and venue for focus groups re- should be a natural setting for that group to meet quire careful consideration: since people’s accounts are dependent on the con- text in which they take place. The workplace, . Who should participate in the groups? This will school, club or health care setting may be appro- depend upon the range of perspectives of relevance priate or a neutral social venue. Expenses should be to the research question. If we were investigating paid for attendance plus a small gratuity where the influence of beliefs on response to symptoms we appropriate.

QUALITATIVE RESEARCH METHODS 81 7.3.3 Participant observation . What is the role of the researcher? The identity of the facilitator for each focus group should be care- fully considered in relation to the characteristics of Participant observation (or ethnography) is the most the participants. The facilitator can never be a traditional method of fieldwork, emerging directly neutral presence. Their gender, age, ethnicity and from the work of anthropologists. The ethnographic background are all influential, but to what extent do method of data collection is most faithful to the they need to share the characteristics of group principles of carrying out the study in the natural members? What impact might their appearance and setting in which the phenomena of interest can be vocabulary have on the discussion? Will their pre- explored. It is an appropriate method for gaining sence unduly influence disclosure, language and insights in an unfamiliar setting. The researcher is idiom? Service user-led research in mental health present in the setting for longer periods of time and has enabled the development of qualitative research data are recorded from observations of, and parti- into mental health care from the insider perspective. cipation in, the setting. The researcher’s contem- Focus groups facilitated by service user researchers poraneous fieldnotes contain detailed accounts may well produce more naturalistic discussion than and impressions of events and conversations for those led by a psychiatrist, whose presence might subsequent analysis. Fieldnotes do not include full encourage participants to frame their accounts with- verbatim transcriptions of speech; they are a dis- in a more clinical framework. tillation of the researcher’s experiences of an un- familiar setting. The content of individual focus groups is flexible, to In the field of mental health care there are examples suit the needs of participants, with a variety of ways to of participant observation studies in which the re- engage members of the group. The session should searcher spends time in an unfamiliar treatment set- begin by making participants feel welcome, offering ting to understand the culture(s) in which staff and refreshments and making introductions. The purpose patients live and work (e.g. see Reference [23]). Quirk of the research should be fully explained along with attended three acute psychiatric wards on an average ethical procedures and ground rules for protecting of two days a week over a period of three to four confidentiality outside the group. The discussion pro- months, recording his impressions in detailed field- gresses by use of a topic guide, which starts with notes and interviewing patients and staff in order to general opening questions and increasingly focuses on provide an ‘insider’ account of life on thewards. Visits more specific key areas. A set of prompts for each were made at varied times of day and night and ward topic ensures that these are explored in sufficient rounds and meetings were attended, but most of the detail. The presence of a co-facilitator helps to ensure researcher’s time was spent in informal interaction that key points are pursued and that all participants are with patients. Over time the recording of fieldnotes enabled to engage in the discussion. Activities can became more focused and structured around emerging be introduced to stimulate discussion: vignettes are themes such as patients’ strategies for managing risk useful, as are rating scales, and ranking a series of and how power is exercised and resisted on the ward. statements according to importance to the group. The distillation of events and social interaction Focusgroupsareoftenusedtoidentifylanguageand through the fieldnotes of a participant observer en- concepts in particular groups, to generate theory and abled a critical examination of the day-to-day func- hypotheses in the pilot stage of larger studies and as a tioning of an institution and identified a range of component in multimethod studies where the focus service implications to improve the experiences of groups help to clarify findings from other methods. in-patients. The validity of the researcher’s interpretations of Participant observation is not for the faint-hearted. the data generated by a focus group may be tested by It requires considerable preparation, high levels of reconvening the group at a later stage. Participants can detailed record keeping, constant attention to the also assist in the analysis of the data by this method. influence of the researcher on the process and

82 PERSPECTIVES AND METHODS interpretation (reflexivity) and considerable invest- the events and interactions observed? The researcher ment of time. Ethical considerations are most acute will need to address these issues before applying for in observational studies. From whom will informed ethical approval and should prepare the ground by consent be sought? Should all those present sign seeking advice and cooperation from key individuals consent forms or only those whose discourse is in- involved in the setting. Accounts of ethnographic cluded in the fieldnotes? What impact will the refusal studies in mental health can be found in Lee of consent from some potential participants have on et al. [24], Skultans [25] and Quirk et al. [23]. 7.4 RESEARCH PROCEDURES AND METHODOLOGICAL APPROACHES 7.4.1 Designing the study cultural idioms of participants. The vignette described an older person with symptoms typical of depression The process of designing and conducting a qualitative and was adapted from a version used in a cross- study conforms to the principles of all scientific in- cultural study [27]. As there was little empirical quiry: review of existing evidence, identification of a evidence to inform the study, an iterative approach gap in current knowledge, formulation of a research to data collection, analysis and further literature re- question to advance knowledge, selection of appro- view was most appropriate. Purposive selection of priate methods, collection and analysis of data, report- participants was guided by an initial sampling frame ing of findings. However, in qualitative studies the to ensure that data were included from participants procedures will vary according to the particular theo- with key characteristics. Interviewing and analysis retical approach. In grounded theory studies [26], for proceeded concurrently until no new conceptual in- instance,cyclesofdatacollectionandanalysisproceed sights were generated from the analysis. This final concurrently, with each stage of analysis informing stage of analysis/data collection is often referred to as further data collection and further literature review. theoretical saturation, particularly in grounded theory. The following is an example of a study design that In practical terms, theoretical saturation occurs when set out to identify ways of improving access to care for the researcher finds similar instances repeatedly. older people with depression [18,19]. Current evi- In the example above, there were a number of dence shows that a very small proportion of older considerations in selecting the methods of inquiry: people with depression receive a diagnosis and an even smaller proportion receives active treatment. . Depression is a sensitive issue. Would older people People from black and minority ethnic groups are be willing to join focus groups to discuss their considerably less likely to receive treatment. Against experiences and needs? How would focus groups this background, the initial aim was to identify barriers be constituted? What would be a natural setting? To and facilitators to accessing services. A qualitative test feasibility we conducted one focus group in a approach was required to explore the experiences, local day centre for older people. Although most beliefs and needs of older people with depression participants were ready to talk in general terms from different ethnic groups, in order to understand about local health services and the most appropriate how older people conceptualize depression and any treatment for depression, some expressed a wish to cultural variations that might be implicated in help- take part in an individual interview; the data gen- seeking. Topics for the initial interview guide were erated in the interviews were richer and more re- generated from a review of the literature and the levant to the aims of the study. advice of health care professionals from the relevant ethnic minority communities in the area. Aware of the . The stigma of depression is greater in black and sensitivity of the subject and the potential impact of minority ethnic groups: the literature suggests this stigma on participants, a vignette was used to stimu- affects disclosure and access to services. How late discussion and to capture the descriptors and would we recruit participants with depression if