Illness Narratives Revisited: The Failure of Narrative Reductionism

by Paul Atkinson
Cardiff University

Sociological Research Online 14(5)16
<http://www.socresonline.org.uk/14/5/16.html>
doi:10.5153/sro.2030

Received: 22 Jun 2009     Accepted: 1 Oct 2009    Published: 30 Nov 2009

Abstract

The argument uses the proliferating research literature on 'illness narratives' to make a more general analytic point about the proper treatment of narratives and life-stories by social scientists. It is suggested that, notwithstanding earlier commentary and criticism, and despite the sophistication of authors such as Mishler, too many narrative-based studies fall far short of a thoroughly analytic approach to such spoken actions. Too often narratives are celebrated as the means for analysts to gain access to personal experience, to the subjective or private aspects of illness. It is argued that we still need analytic strategies that treat illness (or any) narratives as speech acts, based on socially shared resources.

Keywords: Narrative Analysis, Illness Narratives, Accounts, Speech Acts

Introduction

1.1 Charles Tilly is by no means the first or the only social scientist to propose the analysis of systemic relations between local phenomena and ?big? questions. Indeed, he is not the only one to have focused on speech-acts of various sorts as a possible key to the analysis of major social processes: Judith Butler?s use of speech acts is a key case in point (Butler 1999). Tilly?s project seeks to connect recurrent preoccupations of sociological analysis with the fine-grained inspection of everyday life, as exemplified by the work of Erving Goffman and others in the interactionist tradition. His extended essay Why? is an especially striking example of the genre (Tilly 2006). He develops his arguments from such sociological classics as C. Wright Mills?s treatment of ?vocabularies of motive? (Mills 1940) and Goffman?s analysis of social encounters (e.g. Goffman 1981). (All have common roots in the work of Kenneth Burke: see e.g. Burke 1989)

1.2 Tilly?s is, however, one of the recurrent, necessary correctives to preoccupations with ?narratives? and other speech-events or speech-acts that overlook the work of sociological analysis, while celebrating ?narratives? as privileged forms of event in their own right, devoid of cultural conventions, and devoid of social life. Indeed, it is one of the intriguing paradoxes of contemporary social science that essentially non- or even anti-sociological explanatory frameworks are treated as explanatory resources. Sociological interests are too often subsumed into accounts of the subconscious, of subjectivity, or of identities, all divorced from their social and cultural matrix. In a parallel vein, some sociological perspectives have become individualised, so that the work of social inquiry is represented as the inspection of personal, even private, experience in the search for an interior biographical life. Any sense of shared, socially distributed knowledge or competence, any attention to culturally-determined categories ? these are not explored systematically. As a consequence, potentially rich traditions of analysis are routinely lost to view or are reduced to a jejune version of their possibilities.

1.3 The analysis of narratives is one extremely fashionable area that exemplifies these issues. ?Narratives? have been the topics of inquiry from various standpoints, but the treatment of narrative in general remains problematic. In the hands of too many authors, narratives are treated as proxies for the direct apprehension of subjective, personal experience. Narratives have become one of the main vehicles for a form of reductionist social thought. That is, a mode of analysis (or rather a failure of analysis) through which ?the social? is reduced to the ?personal?, social action is (mis)represented as ?experience? and social identities are reduced to ?voice?. This is nowhere more apparent than in the literature concerning ?illness narratives?.

Illness narratives

2.1 In recent years there has been a remarkable growth of interest in ?illness narratives?. The number of published studies of patients? and others? accounts of illness, disability and other bodily phenomena has grown rapidly. Path-breaking studies by authors such as Kleinman (1988), Mishler (1984) and Frank (1995) helped to establish the study of illness narratives, and to construct it as a specialist domain of inquiry in its own right. There are, however, some possible problems with this emergent specialism, not least an apparent ambiguity concerning the appropriate analytic approach to narratives. Moreover, it is apparent that the implicit celebration of illness narratives, as opposed to their systematic, comparative analysis, may create unhelpful expectations on the part of social researchers concerning the existence and nature of narrative accounts of illness. Since the early and influential publications on illness narratives, the field has grown exponentially, and includes several more major texts, as well as numerous research papers. There is, moreover, the movement towards ?narrative medicine? in which attention to narratives of illness is advocated as a core interest of medical practice itself (cf Charon 2002; Clark and Mishler 1992; Elwyn and Gwyn 1998; Greenhalg and Hurwitz 1998). The problems I shall allude to are not confined to narratives of health and illness, but the extensive literature exemplifies generic issues beyond the specific genre.[1]

2.2 In 1997 I argued that the major works based on illness narratives display at best an unresolved ambiguity concerning the status of such narratives themselves (Atkinson 1997). I suggested that while authors like Kleinman, Frank and Mishler promote the analysis of narratives in the interests of social-science research, they implicitly endow illness narratives and their production with a special significance. That particular argument fed directly into the paper I wrote with David Silverman (1997) at the same time. Those publications were intentionally polemical and challenging, voicing as they did a thoroughly sceptical view on a fashionably emergent interest. Responses and subsequent publications have done relatively little to dispel my own feeling that the study of illness narratives is sometimes based on inappropriate assumptions, frequently lacking in analytic rigour, and in need of clarification. Some authors (e.g.Miczo 2003) have managed to interpret the critical comments by Silverman and myself in completely the wrong way, contriving to read our observations as if we were uncritically endorsing the view of interviews and narratives as windows onto private experience (an example of how one can be quoted verbatim but out of context). My views are congruent with those offered by Sandelowski (2002) who decries the methodological naiveté of much ?qualitative health research? for an uncritical reliance on the interview as a window on ?authentic experience? (p. 105). Despite these reservations, a great deal of research continues to treat the interview and the narratives it generates with uncritical approval.

2.3 The state of the field merits a renewed evaluation. In the rest of this paper, therefore, I shall examine whether my original strictures still hold true, whether there is a coherent theory of illness narratives to be excavated from the current literature, and how we can best approach the study of narratives of illness, the body and similar phenomena. Before doing so, I need to recapitulate briefly some of my earlier contentions, if only to set the scene for this current discussion.

2.4 The tendency to promote illness narratives is also prominent in the movement ? of relatively recent date ? to encourage a ?narrative? approach to medical practice and medical understanding. In keeping with Kleinman?s original formulation, the collection and interpretation of narratives is seen as an especially productive way of ensuring that the ?voice? of the patient is audible in the clinical encounter, and that ?lay? understandings of illness are grasped adequately. These perspectives reflect in turn the critique of everyday medical practice formulated by Mishler (1984). Mishler?s discussion draws on his distinction between the ?voice of medicine? and the ?voice of the lifeworld?. The two ?voices? mutually interrupt one another in the clinical encounter. The voice of ordinary medical practice is described in terms of de-contextualized and impersonal discourse, while the voice of the lifeworld is grounded in the mundane, the personal and the experiential. Mishler?s analysis is not intended to imply that the single physician embodies only the voice of medicine, nor that the voice of the lifeworld is articulated only by the lay patient. Nonetheless the moral asymmetry implied by Mishler?s dualism is powerful. In the process, the value of a narrated lifeworld is asserted. The voice of medicine seems to be marked precisely by the absence of personal narrative. The more that the lifeworld is celebrated in terms of complex personal experience and interpretation, the more the world of medicine is emptied of human, personal content.

2.5 Likewise, despite Frank?s avowed analytic purpose, his own treatment of illness narratives has a strong moral aspect, in that a narrative mode of understanding and expression is celebrated. It is part of a representational device on Frank?s part that portrays illness sufferers as (potential) heroes in their own narrated dramas of sense-making and survival. Kleinman?s lengthy collection and exegesis of illness narratives also implies that such accounts embody an especially important source of insight into the narrator?s unique understandings of her or his condition. Kleinman advocates close attention on the part of physicians to their patients? narratives, suggesting that such a perspective would encourage the equivalent of an ethnographic understanding of the patient and her or his lifeworld of illness experience. In recent years, we have witnessed an extension of the narrative turn that advocates not merely analytic attention to illness narratives, but a form of narrative medicine that places narrative work at the heart of medical professional practice and competence.

2.6 In some ways, the promotion of a narrative perspective can be seen as the further development of critiques of conventional medical consultations that advocate close attention to the words and experiences of the patient, and ascribe at least some degree of therapeutic benefit to a mutually attentive consultation. It also reflects a prevalent style of thought among sociologists, anthropologists and health psychologists that accords an enhanced status to the knowledge and experiences of ?lay? persons ? patients and their family-members, or lay activists and interest groups. (The latter have also been linked to the newly fashionable idea of biological citizenship.) The production of illness narratives can thus be portrayed as a feature of cultural competence through which lay patients can display their practical, biographically-grounded comprehension of medical phenomena. Likewise, their self-knowledge is celebrated in comparison with the impersonal knowledge of the medical practitioner.

2.7 This suggests that what we might call the ?classic? literature on illness narratives was founded on an implied dualism linked to a therapeutic commitment. Within such a framework, it is also clear that an appeal to the biographical is also strongly allied to an appeal to ?experience?, where the patient?s experience of illness is contrasted with the biomedical knowledge of the medical practitioner. It is perfectly clear that such dualism is to be found deeply embedded in recent and current literature, as is the emphasis on illness experience. While I do not want to suggest that such interests are entirely spurious, I shall re-affirm that they can be misleading.

2.8 The original inspiration of work such as Mishler?s was clearly an attempt to engage with some ?big? issues of contemporary medical practice. His analysis of discourse was intended to capture social phenomena, and not merely the idiosyncrasies of individual experience. His treatment of the medical encounter went well beyond the sort of ?culture of complaint? that has marred too much discussion of medical encounters. Mishler?s analysis is not just an account of how medical practitioners do not communicate adequately with their patients, or do not listen to them with sufficient attention, or do not deal with them with proper civility (see Clark and Mishler 1992). (Although all of those may be true.) Mishler?s analysis is, on the contrary, essentially a structural one. His close analysis of transcribed consultations is not a de-contextualised account of medical encounters alone; his account of the ?voice of medicine? and the ?voice of the lifeworld? does not simply endorse the latter as the source of authentic personal experience. His are ideal types, and they are deployed in order to address major features of medical knowledge. They are not used simply in order to generate ironic contrasts between the medical and the personal. Unfortunately, far too many accounts of ?illness narratives? have done just that, and have been all too ready to endorse the narrative voice of the lifeworld as a mirror of personal subjectivity, and as a means of exploring the interiority of patients? personal ?experience?.

2.9 Consider the following formulation in the editorial introduction to an influential collection of papers. The editors observe that papers in their volume illustrate a narrative perspective that respects:

?the requirement that human subjectivity should no longer be seen as the devalued opposite of scientific objectivity, linked in some assumed zero-sum relationship whereby more of the one must necessarily mean less of the other. Rather, objective assessment (for example, medical diagnosis) and objective intervention (for example, medical treatment or palliation) provide but one important dimension of knowing. However complete the objective dimension, if we exclude subjectivity and its narrative expression through dialogue, we remove diversity of viewpoint and impoverish the knowledge we can gain about human suffering and the impact of our efforts to care. (Hurwitz, Greenhalgh and Skultans 2004: 3-4)

2.10 The equation of ?narrative? with ?subjective? experience and the identification of an ?objective? medical culture is a recurrent device in the narrative literature and the justifications for illness-narrative research. It is sometimes accompanied by the now familiar characterisation of medicine or biomedicine as a monolithic system. The person is thus counterposed to an impersonal system of knowledge, practice and institutions. The narrative analyst thus aligns her- or himself alongside the individual, potentially oppressed patient, contra the impersonal system of medicine and healthcare. The insertion of the uniquely biographical ? and Mishler?s lifeworld ? remains one of the recurrent, primary justifications for narrative research.

2.11 Mishler?s analysis is couched in terms of ?voices?, and the repeated narrative-analytic emphasis on experience is paralleled by a preoccupation with ?voice?. In Mishler?s original treatment, voice is essentially a discursive construction. But it is too often treated as a reflection of personal standpoint. The voices of narrative are, therefore, too often treated as sources of authenticity, grounded in the biographical particularities of speaking subjects. The notion of ?voice; in contemporary social research is a potentially ambiguous one. On the one hand, it can be used to warrant a form of identity politics, through which particular categories of actor are celebrated. On the other hand, it can refer to a particular idiolect of discursive practices through which distinctive constructions of self and others are enacted. The former reflects a view of narrative-as-testimony, the latter a view of narrative-as-account. It is this tension that informs the remainder of this paper.

2.12 Moreover, the function of narratives can be represented as providing a meaningfulness to illness episodes and biographical sequences that are otherwise meaningless or chaotic. The narrative furnishes interpretative frameworks within which some degree of coherence is achieved by patients and sufferers. Such perspectives have a long and distinguished history in the sociology of health and illness. (For a recent example of analysis that focuses on ?meaning? in this vein, see Lee and Poole 2005.) Narratives may, of course, be devices whereby biographical coherence is enacted, but too often this performance is elided by celebrants of narrative with an existential state of wholeness or reconciliation. The narrative is itself portrayed as an act of self-creation, and represented as an act of resistance. A number of authors in the collection edited by Hurwitz, Greenhalgh and Skultans are grouped together under the rubric of ?counter narratives?. This the editors gloss in terms of ?an alternative voice from that offered in the standard biomedical account? (p. 9). Jones and Bunton (2004) recount two stories of deafness: the ?wounded? story and the ?warrior? story. They are themselves quite right to identify these as genres of deaf narrative (and I shall return to this below). They characterise the narratives in terms closely related to Frank?s treatment: ?The wounded storyteller?s narrative progresses from order to chaos and back to order again? (p. 201), whereas ?The warrior narrative seems to have no such simple progression from chaos back to order. It promises only struggle compensated for by a sense of belonging, to an increasingly small group? (ibid.). Likewise Baldwin (2004) presents the narratives of actors accused of Munchhausen Syndrome by Proxy, Le Fanu (2004) discusses parents? narratives surrounding shaken-baby syndrome, and Richardson (2004) discusses parents? stories from the UK organ-retention scandals. These latter three papers do rather little (or nothing) critically or analytically to examine narrative acts themselves, or to examine the rhetorical devices used by narrators in order to produce their undeniable narrative force. The treatment of these stories is clearly coloured by the fact that the academic authors are in favour of a sympathetic affiliation with the story-tellers and invite an identification with their victimhood on the part of the reader. (I do not thereby mean to imply that sympathy with such individuals should be withheld in general: merely that it does not substitute for social science.)

2.13 When patients cannot or do not produce adequately developed narratives, then authors can help them to do so. For instance, a commonplace story can be embellished by translating it into an overtly poetic form, as do Kendall and Murray (2004) They justify their approach by pointing out that narrative accounts have rhythms, while the rhythms of poetry can reflect the cadences of everyday speech. They also invoke the claims of other authors who have advocated transcription and analysis of narratives in terms of stanzas. Now there is nothing inherently wrong in transcribing and analysing accounts in terms of their ?poetic? qualities. Distinguished authors like Dell Hymes have demonstrated the potential power of such an approach (e.g. Hymes 1981, 1996). But it needs to be treated as an analytic issue ? linked to the prosody of spoken language: we need narratives and accounts analysed in terms of their rhythms, cadences, and so on, if such ?poetics? are to be anything other than whimsical impositions upon the data. Likewise, a satisfying narrative can be pieced together from fragments, so ?restoring? a sense of narrative unity that would otherwise by missing. This is an approach exemplified by Mattingly (2004) for example. Again, there may be nothing absolutely unacceptable about that, but we need to recognise that such acts of narrative reconstruction and restitution can clearly act to enhance the narrative or performative quality of spoken accounts. In contrast, will unvarnished, restricted, penny-plain accounts be dismissed by narrative researchers, on the grounds that they are insufficiently ?poetic?, or not adequately revelatory of an inner experience on the part of the informant? Indeed, must we implicitly ask of every sick or disabled person that they produce for us floridly constructed and highly personalised accounts of their troubles? Shall we not treat them seriously unless they display such verbal arts to a sufficiently high degree? There is a clear danger here of merging essentially analytic issues with aesthetic judgments.

2.14 We need to ask ourselves why narrative research in health and medicine has achieved the kind of popularity that it enjoys. The first answer might be a purely empirical one. It is clear that when interviewed ? either in the clinic or in a research encounter ? some patients and family members undoubtedly produce stories and accounts. It is undeniably true that they locate their illness, diagnosis and patient career within a biographical framework. But I do not think that this quite goes far enough to explain the popularity. As I have already suggested, the sustained interest in illness narratives seems to reflect a particular set of assumptions about patienthood and personhood. In the face of an impersonal world of biomedical knowledge and practice, the patient is reconstituted through narrative reconstruction and narrative restitution. This embodies a particular ? if sometimes unspoken ? model of the actor. It is in many cases an under-socialised individual. The stress on narrative frequently strips the inquiry of any sociological or anthropological thrust. Individuals narrators are portrayed as just that: they speak alone, about themselves and for themselves, in a social vacuum. In general there is little in the canon of published narrative literature to dispel my earlier misgivings ? to the effect that the narrative turn embodies a Romantic view of the speaking subject. Individualised and personalised, but devoid of social identity or cultural resources, the narrating speaker is celebrated as an atomised subject. This is a reflection of identity politics, and the equation of the social with the personal. In other words, a great deal of so-called narrative analysis involves a de-socialisation of the patient. It also reflects the remarkable popularity of the extended interview as a research method, often under the aegis of a more general appeal to ?qualitative research?.

2.15 Riessman has acknowledged the force of this argument from within the narrative tradition. (Indeed, her distinguished contribution to narrative analysis escapes many of the strictures I have outlined up to this point.) She herself recognises an intellectual discomfort with what she calls the ?tyranny? of the narrative (Riessman 1997, 2002). Her views converge with mine and David Silverman?s (Atkinson and Silverman 1997) : we continue to witness an almost compulsive obsession with the extended interview among social scientists. This tendency is in turn linked to professional (therapeutic) and commercial (celebrity) interests in talking about ?experience?. What Silverman and I called ?the interview society? continues to permeate not only everyday life but also the professional activities of researchers in the social sciences (cf. Gubrium and Holstein 2002). I want to suggest that this ?tyranny of the narrative? is based on some tacit assumptions that deserve reflection.

2.16 There is a barely acknowledged assumption that patients, sufferers or family members ?ought? to produce extended and elaborated accounts of personal experience. Equally, there may be the accompanying expectation that researchers ?ought? to generate extended narratives of a particular sort ? heroic, resistant, sense-making and so on. When narratives of the right sort are not forthcoming, there is the danger ? again implicit ? that the research informant, the researcher or both will be regarded (or regard themselves) as ?lame?. The absence of a satisfying illness narrative could therefore be seen as indicating failed illness behaviour, rather than a research finding or topic of investigation. There is also a danger of universalising the performance of narratives. By that I do not necessarily mean the specific content of narratives, but the social fact of narrative itself.

2.17 In other words, when ?narrative? is uncritically proposed as an essentially human form of conduct, and as a special form of sense-making, then there is little analytic impetus towards the investigation of the conditions of narrative production, the social distribution of narrative types, or the association of narrative with social position and status. In other words, the elevation of narrative to a form of life in itself implicitly robs the field of analytic edge. The universal narrative subject, in other words, seems to be an articulate, reflective actor, who displays the confessional aptitude of a member fully socialised into the conventions of the interview society itself. Most publications on ?narrative? take little account of the voluminous research literature on the ethnography of talk and the anthropology of communication. Yet all that we know from such research literature suggests that talk is by no means uniform. The rights to talk, the permissible topics of talk, the conventions of participation in speech events are all constitute of social position. We cannot assume a universal narrating subject.

2.18 There is, in the repeated emphases on narrative and medicine, the danger of assuming that patients, or other informants, somehow ?ought? to be prolific narrators, whose biographical performances are replete with florid stories. Ideally, in these implicit assumptions, the stories should be heroic, or oppositional, or should otherwise be expressive of a humane response to illness, and to be expressive of personal experience. But these forms of narrative seems to be a distinctively Anglo, middle-class mode of self-expression. Much of the current literature seems implicitly to endorse a particular type of narrative, and universalising it. Moreover, such an apparently uncritical perspective on narrative implicitly endorses a naïve view of narrative itself.

2.19 What is needed is not merely a celebration of illness narratives, nor a celebration of narrative medicine as a form of humanistic practice (admirable though the latter might be in practice). We have need to take stock of what analysis of illness narratives might amount to. We need, in other words, to develop more systematic attention to the social-science strand of work in this vein, while not being seduced by the celebratory tendency.

Analytic perspectives

3.1 My critical remarks to this point should not blind us to the fact that there has been sustained serious research that has been informed by analysis rather than celebration, and by the recognition of illness narratives as speech acts, as performative, and as cultural phenomena that express socially shared resources. We need, therefore, to piece together the outlines of a general analytic framework, and to outline what is known analytically.

3.2 It is already clear that we can identify a number of possible genres among the illness narratives that have been collected and discussed in the existing literature. We can start to map out some relevant analytic issues, even though we shall find that overall, the presence of analytic models ? even at an implicit level ? remains patchy at best.

3.3 In one way, Frank?s original work is a useful starting-point, notwithstanding my earlier criticisms. Frank outlines a number of key narrative types. He identifies the restitution narrative, which is the professionally preferred account of acute illness and the sick-role. This narrative form he contrasts with chaos ? when ?no narrative is available to express what is being experienced? (Frank 2004: 140). Thirdly, when no restitution story is available or satisfactory, then, Frank suggests, patients move to a quest story. Likewise, Jones and Bunton (2004) describe two ideal types ? the ?wounded narrative? and the ?warrior? narrative. Wounded narratives describe damage, exclusion and integration, while the warrior narratives describe fight (for rights), belonging (to a culture or a community) and separatism (being part of a linguistic minority). Now these, and analyses like them, are clearly of value in examining story-types. But it is my contention that they are far too global, too general to provide serious analytic resources. Such ideal types focus on too restricted an array of narrative functions. They pass far too rapidly from narratives themselves to sweeping characterisations of ?grand narratives? of personal meaning, resistance and the like. Further, the potential proliferation of genres implied in such an approach will do little to produce any analytic order in the field as a whole.

3.4 Jorden, Little, Paul and Sayers (2001) provide one useful analysis from a different perspective. They identify five different types of illness narrative: First, there are the ?narratives of personal experience?, that conform to the type and structure identified by Labov (e.g. Labov and Waletsky 1967). (It is instructive that they treat this ?Labov? structure as one particular genre rather than having generic relevance.) Second, there is the genre of ?recount?, which ? as the title implies ? recount ?personal experience unproblematically?. Then there are three additional story-like types: the Anecdote, the Exemplum and the Observation. The point of the analysis provided by Jorden et al is not simply to list these genres or types, but to use them to construct an analysis of narrative complexity. Narrative complexity here refers to the number and variety of story-like elements within a single account. They then use this measure of narrative complexity to index the degree of ?life disruption? experienced by the narrator. This is a useful way forward, although I am not convinced by their use of narrative form to measure life disruption (when the latter seems to be a narrative function itself). It does, however, have the merit of trying to identify some structural constituents of narratives themselves.

3.5 We have somewhat isolated accounts of specific tropes ? such as Skott?s account of metaphors in cancer narratives (Skott 2002) as well as the generic tropes identified by Jones and Bunton (2004) ? and they readily find links with broader cultural and anthropological accounts of illness. Metaphors of cancer and AIDS are among the most prominent of genres in this vein. But illness narratives are poorly integrated with anthropological or sociological analyses of the semantics of ill-health, suffering and embodiment. While these have been discussed in broad, cultural terms the close analysis of such tropes has not featured prominently in the current literature on illness narratives, and have not been fully integrated into the close reading of patients? (or indeed professionals?) accounts.

3.6 Clearly multiple functions of illness narrative have been identified. They include various versions of biographical reconstitution, identity-work, political resistance to organised medicine. But too often these kinds of ?analysis? are poorly grounded in the formal properties of narratives themselves. They are not tied to specific accounting devices or narrative tropes. Equally, we find extracts of accounts of illness causation. But we seem to have progressed little, if at all, since Gareth Williams?s pioneering paper (Williams 1984) in which he attempted to locate beliefs about the aetiology of arthritis within the discursive function of narrative reconstruction. Williams, that is, suggests that ?explanations advanced by afflicted individuals have both causal and purposive or functional components. They represent not only explanations for the onset of a given disease, but also acts of interpretation, narrative reconstructions of profound discontinuities in the social processes of their daily lives? (p. 179). In the intervening quarter of a century we still do not see a well-developed, cumulative series of analyses of causal accounts. We do have equivalent discussions, focused on specific conditions and ?lay? explanations for their causes and trajectories. We lack, however, anything resembling even the most elementary of general analyses of the formal properties of such accounts. For if we need series of accounts based in particular medical conditions and problems, we also need to look beyond the possibilities of narrative ?butterfly collecting?. We need to develop systematic syntheses and meta-analyses of how, in general, such accounts are constructed, using what rhetorical devices, with what functions, and with what effects (see e.g. Gwyn 1999; Hydén 1997; Katz and Shotter 1996).

3.7 Equally, narratives of self and identity are represented in the research literature. Yet there is a remarkable lack of systematic and sustained attention to the forms of biographical work that are sustained in illness narratives. The fact that the ?self? is a narrative accomplishment is important, but it is not the whole story. We need to develop much more explicit models of how that biographical work is accomplished, and what resources of narrative (and other forms of spoken action) are mobilised in its enactment. Zinn (2004) is one notable example of recent work that seems to point in the right direction. The empirical research that informed his account was not designed originally to address issues in health and illness. Indeed, it is noticeable that Zinn makes no reference to the extensive literature on illness narratives and accounts. Nonetheless, the typology of biographical strategies is a useful exemplar of how a more formal, sociologically-informed mode of analysis can relate issues of health and illness to wider analytic perspectives. (I am not advocating the particular sociological preoccupations of Zinn, based as they are on notions of late modernity, and the work of authors like Beck and Bauman.) The work is rooted in an analysis of modes of biographical certainty, autonomy and control, that are in turn derived from informants? autobiographical accounts. This is not the whole answer, but I think it is a fruitful example of what can and should be done. (Because of its origins, the paper makes no reference to any prior work on biographical disruption and illness narratives ? which would have strengthened the literature on both fronts.)

3.8 We await a more developed array of systematic analyses of attributions of responsibility, or of moral agency, or stigma. There has been little or no advance since the original publication of Margaret Voysey?s pioneering analysis of accounts of parents of children with handicaps (Voysey Paun 2006). Refusing to assume that such parents self-evidently had ?problems? of living with handicap, Voysey analysed their personal accounts in order to demonstrate how family members enacted ?normal parenthood?. She also examined how parents deployed vocabularies of legitimation that reflected discursive genres that mirrored the discursive resources furnished by professional agents. We need, therefore, to recognise that narrative genres and devices ? such as those of illness narratives ? incorporate accounting methods that themselves reflect the interests and knowledge-forms of professionals and other sources of authority. A sustained analysis of narrative forms is not restricted to an appreciative understanding of ?storied experience? in a social and cultural vacuum. Plummer?s account of ?sexual stories? is an important exemplar, suggesting how sociologists and others might take forward a sustained analysis of the public forms of personal experience (Plummer 1995). In the same vein, we need to recognise that ?memory? and indeed ?time? are enacted through narrative and discursive devices.

3.9 In positive terms, therefore, the analysis of accounts ? including illness narratives ? needs to pay close attention to the following analytic features.

3.10 Of course, these injunctions apply to the analysis of narratives in general. A further, more specific, injunction therefore arises: There is nothing special or privileged about the genre of narrative (such as illness or sexual narratives): we ought to avoid assumptions concerning their existential status. I recognise that in raising my objections to some current work, and in making these more positive recommendations, I am not alone. There are many complementary approaches to narrative analysis that recognise the value of systematic, disciplined analysis (see Stanley 1992 and Riessman 2008 for statements that span nearly two decades and represent complementary perspectives, and for a series of programmatic statements see e.g. Josselson 2006; Taylor 2006; Kraus 2006; see also Labov 1997; Maines 1993; Young 1997; Wortham 2000). It seems, however, that these strictures and caveats still need reiteration in the face of appeals to qualitative research that under-value systematic analysis.

Discussion

4.1 As Brett Smith (2007) has reminded us in his reflections on the state of the art, narrative analysis ? and related fields of inquiry ? are at best in a state of creative variety. At worst, the field is incoherent and muddled. I make no apologies for echoing the remarks that Sara Delamont and I made in a special issue of Narrative Inquiry (Atkinson and Delamont 2006): the field of narrative analysis, and work on illness narratives in particular, remains in desperate need of a disciplined renewal. It is simply not enough to conduct some extended interviews, to identify some general themes, to reproduce extracts from those interviews, and to celebrate the fact that patients, sufferers and family members express themselves through stories. It is not enough to identify a small number of narrative functions. These are all, admittedly, significant. But in the absence of serious and sustained analysis of major corpora of narrative materials, then the field will remain at a stunted level of development.

4.2 Furthermore, we need to free ourselves collectively from the state of mind that celebrates narratives as special kinds of phenomena. Analysis needs to be grounded in a thoroughgoing social and cultural analytic framework. Narrative needs to be brought within the disciplinary traditions of sociology, anthropology or discursive psychology. We must reunite narratives with classic work on accounts and accounting devices (see Antaki 1988). We also need to free the field from the implicit assumption that there is something intrinsically ?good? about illness narratives ? that they capture experience, that they embody resistance to medical authority, or that they furnish access to an enhanced and more humane form of medicine.

4.3 There is a clear need to draw on the now extensive literature of illness narratives, in order to construct more analytic work. Such analyses also need to be more firmly reintegrated with formal discourse analysis, with sociological and anthropological interests in accounting devices, and with the systematic analysis of spoken action generally. We cannot, collectively speaking, afford to let ?illness narratives? be privileged as a special form of social action; we need to rescue them from the methodological vacuum of ?qualitative health research?.

4.4 As I have suggested, we need ? collectively ? to remember to analyse accounts, descriptions, narratives, claims and the other speech acts that constitute the characteristic data of contemporary ?qualitative? social inquiry. Failure to do so constitutes a betrayal of any claims to well-warranted knowledge on our part. Furthermore, such analyses clearly need to reengage with the core, recurrent preoccupations of disciplinary interests in social science. In the spirit of Charles Tilly, we need simultaneously to pay serious attention to the forms and function of narratives themselves, while excavating the interests of the interactionist and interpretative traditions that have fostered ?micro-sociological? analyses of social processes. Equally, we need to recognise that Tilly himself is recapitulating a longstanding tradition of sociological work ? largely associated with the interactionism ? that is too often a submerged tradition (Rock 1979; Atkinson and Housley 2003). David Maines (2001) has demonstrated, for instance, the power of narratives in an interactionist framework. Narratives are among the mechanisms and resources through which ?mind, self and society? are constructed. Sociologists and anthropologists forget at their peril that there are robustly analytic ways of accounting for shared phenomena, and that there is no need to appeal to private, interior ?experience? in order to make sense of the narrative mechanisms of social action and reproduction. This is not a knee-jerk affirmation of traditional disciplinary boundaries and prerogatives, although it does sound a warning note against treating ?narrative? as if it furnished a sub-discipline in its own right. Indeed, it does not really matter what disciplinary label is applied: what is important is the disciplined application of an analytic apparatus that is not dependent merely on an empathetic appreciation of personal accounts.

Notes

1 The extensive literature on illness narratives, on which this paper draws includes the following: Arrington (2003); Bailey (2004)Brown and Addington-Hall (2008); Bülow (2003); Carricaburu and Pierret (1995); Carter (2004); Ezzy (2000); Faircloth et al. (2004); Grace (2003); Hörnsten, Sandström and Lundman (2004); Kelly and Dickinson (1997); Lillrank (2003); Nettleton et al. (2004); Poindexter (2004); Roberts (2004); Sakalys (2003); Stephens, Carryer and Budge (2004); Thomsen and Jensen (2007).

References

ANTAKI, C. (1988) Analyzing Everyday Explanation: A Case Book of Methods. London: Sage.

ARRINGTON, M.I. (2003) ?I don?t want to be an artificial man?: narrative reconstruction of sexuality among prostate cancer survivors, Sexuality and Culture, 7 (2): 30-58. [doi:10.1007/s12119-003-1011-9]

ATKINSON, P. (1997) Narrative turn or blind alley?, Qualitative Health Research, 7 (3): 325-344. [doi:10.1177/104973239700700302]

ATKINSON, P. and Delamont, S. (2006) Rescuing narrative from qualitative research, Narrative Inquiry, 16 (1): 164-172 [doi:10.1075/ni.16.1.21atk]

ATKINSON, P. and Housley, W. (2003) Interactionism: An Essay in Sociological Amnesia. London: Sage

ATKINSON, P. and Silverman, D. (1997) Kundera's Immortality: The interview society and the invention of the self, Qualitative Inquiry, 3: 304-325. [doi:10.1177/107780049700300304]

BAILEY, P.H. (2004) The dyspnea-anxiety-dyspnea cycle ? COPD patients? stories of breathlessness: ?It?s scary when you can?t breathe?, Qualitative Health Research, 14 (6): 760-78. [doi:10.1177/1049732304265973]

BALDWIN, C. (2004) Narrative analysis and contested allegations of Munchausen syndrome by proxy, in B. Hurwitz, T. Greenhalgh and V. Skultans (eds) Narrative Research in Health and Illness, London: Blackwell/BMJ, pp. 205-222. [doi:10.1002/9780470755167.ch13]

BROWN, J. and Addington-Hall, J. (2008) How people with motor neurone disease talk about living with their illness: A narrative study, Journal of Advanced Nursing, 62 (2): 200-208. [doi:10.1111/j.1365-2648.2007.04588.x]

BÜLOW, P.H. (2003) In dialogue with time: Identity and illness in narratives about chronic fatigue, Narrative Inquiry, 13 (1): 71-97.

BURKE, K. (1989) On Symbols and Society. Chicago: University of Chicago Press.

BUTLER, J. (1999) Gender Trouble, 2nd edition. London: Routledge.

CARRICABURU, D. and Pierret, J. (1995) From biographical disruption to biographical reinforcement: The case of HIV-positive men, Sociology of Health and Illness, 17 (1): 65-88. [doi:10.1111/1467-9566.ep10934486]

CARTER, B. (2004) Pain narratives and narrative practitioners: A way of working ?in-relation? with children experiencing pain, Journal of Nursing Management, 12(3): 210-16.

CHARON , R. (2001) Narrative medicine: Form, function, and ethics, Annals of Internal Medicine 134 (1): 83-87.

CLARK, J.A. and Mishler, E.G. (1992) Attending to patients? stories: Reframing the clinical task, Sociology of Health and Illness 14 (3): 344-70. [doi:10.1111/1467-9566.ep11357498]

ELWYN, G.J. and Gwyn, R. (1998) Stories we hear and stories we tell: Analysing talk in clinical practice, in P. Greenhalgh and B. Hurwitz (eds) Narrative Based Medicine. London: British medical Association.

EZZY, D. (2000) Illness narratives: time, hope and HIV, Social Science and Medicine, 50 (5): 605-17. [doi:10.1016/S0277-9536(99)00306-8]

FAIRCLOTH, C.A., Boylstein, C., Rittman, M., Young, M.E., Gubrium, J. (2004) Sudden illness and biographical flow in narratives of stroke recovery, Sociology of Health and Illness, 26 (2): 242-61. [doi:10.1111/j.1467-9566.2004.00388.x]

FRANK, A. (1995) The Wounded Storyteller: Body, Illness, and Ethics. Chicago: University of Chicago Press.

GOFFMAN, E. (1981) Forms of Talk. Oxford: Blackwell.

GRACE, V. (2003) Embodiment and meaning; Understanding chronic pelvic pain, Journal of Consciousness Studies, 10 (11): 41-60.

GREENHALGh, T. and Hurwitz, B. (eds) (1998) Narrative Based Medicine: Dialogue and Discourse in General Practice. London: BMJ.

GUBRIUM, J.F. and Holstein, J.A. (2004) From the individual interview to the interview society, in J.F. Gubrium and J.A. Holstein (eds) Handbook of Interview Research, Thousand Oaks CA: Sage, pp. 3-32.

GWYN, R. (1999) ?Captain of my own ship?: Metaphor and the discourse of chronic illness, in L. Cameron and G. Low (eds) Researching and Applying Metaphor. Cambridge: Cambridge University Press.

GWYN, R. (2000) ?Really unreal?: Narrative evaluation and the objectification of experience, Narrative Inquiry, 10 (2): 313-40. [doi:10.1075/ni.10.2.03gwy]

HÖRNSTEN, Å., Sandström, Lundman, B. (2004) Personal understandings of illness among people with type 2 diabetes, Journal of Advanced Nursing, 47 (2): 174-82.

HURWITZ, B., Greenhalgh, T. and Skultans, V. (eds) (2004) Narrative Research in Health and Illness. Oxford: Blackwell/BMJ

HYDÉN, L.C. (1997) Illness and narrative, Sociology of Health and Illness, 19 (1): 48-69.

HYMES, Dell (1981) In Vain I Tried to Tell You: Essays in Native American Ethnopoetics. Philadelphia: University of Pennsylvania Press.

HYMES, Dell (1996) Ethnography, Linguistics, Narrative Inequality: Towards an Understanding of Voice. London: Taylor and Francis.

JONES, L. and Bunton, R. (2004) Wounded or warrior? Stories of being or becoming deaf, in B. Hurwitz, T. Greenhalgh and V. Skultans (eds) Narrative Research in Health and Illness, London: Blackwell/BMJ, pp. 189-204. [doi:10.1002/9780470755167.ch12]

JORDENS, C.F.C., Little, M., Paul, K. and Sayers, E-J. (2001) Life disruption and generic complexity: A social linguistic analysis of narratives of cancer illness, Social Science and Medicine, 53(9): 1227-36. [doi:10.1016/S0277-9536(00)00422-6]

JOSSELSON, R. (2006) Narrative research and th4e challenge of accumulating knowledge, Narrative Inquiry, 16, 1: 3-10. [doi:10.1075/ni.16.1.03jos]

KATZ, A. and Shotter, J. (1996) Hearing the patient?s voice: Towards a social poetics in diagnostic interviews, Social Science and Medicine, 43: 919-931. [doi:10.1016/0277-9536(95)00442-4]

KELLY, M.P., Dickinson, H. (1997) The narrative self in autobiographical accounts of illness, The Sociological Review, 45(2): 254-78. [doi:10.1111/1467-954X.00064]

KENDALL, M. and Murray, S. (2004) Poems from the heart: Living with heart failure, in B. Hurwitz, T. Greenhalgh and V. Skultans (eds) Narrative Research in Health and Illness, Oxford: Blackwell/BMJ, pp. 52-72. [doi:10.1002/9780470755167.ch4]

KLEINMAN, A. (1988) The Illness Narratives: Suffering, Healing and the Human condition. New York: Basic Books.

KRAUS, W. (2006) The narrative negotiation of identity and belonging, Narrative Inquiry, 16, 1: 103-11. [doi:10.1075/ni.16.1.14kra]

LABOV, W. (1997) Some further steps in narrative analysis, Journal of Narrative and Life History, 7, 1-4: 395-415.

LABOV, W. and Waletzky, J. (1967) Narrative analysis: Oral versions of personal experience, in J. Hel, (ed.) Essays on the Verbal and Visual Arts. Seattle: University of Washington Press.

LE FANU, J. (2004) Confounding the experts: the vindication of parental testimony in shaken baby syndrome, in B. Hurwitz, T. Greenhalgh and V. Skultans (eds) Narrative Research in Health and Illness, London: Blackwell/BMJ, pp. 223-38. [doi:10.1002/9780470755167.ch14]

LEE, A.M. and Poole, G. (2005) An application of the transactional model to the analysis of chronic illness narratives, Qualitative Health Research, 15 (3): 3346-64 [doi:10.1177/1049732304272914]

LILLRANK, A. (2003) Back pain and the resolution of diagnostic uncertainty in illness narratives, Social Science and Medicine, 57 (6) 1045-54. [doi:10.1016/S0277-9536(02)00479-3]

MAINES, D.R. (1993) Narrative?s moment and sociology?s phenomena: Toward a narrative sociology, Sociological Quarterly, 34, 1: 17-38. [doi:10.1111/j.1533-8525.1993.tb00128.x]

MAINES, D.R. (2001) The Faultline of Consciousness: A View of Interactionism in Sociology. New York: Aldine de Gruyter.

MATTINGLY, C. (1998) Healing Dramas and Clinical Plots: The Narrative Structure of Experience. New York: Cambridge University Press.

MATTINGLY, C. (2004) Performance narratives in the clinical world, in B. Hurwitz, T. Greenhalgh and V. Skulktans (eds) Narrative Research in Health and Illness, Oxford: Blackwell/BMJ, pp. 73-94. [doi:10.1002/9780470755167.ch5]

MICZO, N. (2003) Beyond the ?fetichism of words?: Considerations on the use of the interview to gather chronic illness narratives, Qualitative Health Research 13 (4) 469-90. [doi:10.1177/1049732302250756]

MILLS, C.W. (1940) Situated actions and vocabularies of motive, American Sociological Review, 5: 439-52. [doi:10.2307/2084524]

MISHLER, E.G. (1984) The Discourse of Medicine: Dialectics of Medical Interviews. Norwood NJ: Ablex.

NETTLETON, S., O?Malley, L., Watt, I. and Duffey, P. (2004) Enigmatic illness: Narratives of patients who live with medical unexplained symptoms, Social Theory and Health, 2 (1): 47-66. [doi:10.1057/palgrave.sth.8700013]

PLUMMER, K. (1995) Telling Sexual Stories: Power, Change and Social worlds. London: Routledge. [doi:10.4324/9780203425268]

POINDEXTER, C.C. (2004) Medical profiling: narratives of privilege, prejudice, and HIV stigma, Qualitative Health Research, 14 (4): 496-512. [doi:10.1177/1049732303262450]

RICHARDSON, R. (2004) Narratives of compound loss: Parents? stories from the organ retention scandal, in B. Hurwitz, T. Greenhalgh and V. Skultans (eds) Narrative Research in Health and Illness, London: Blackwell/BMJ, pp. 239-56. [doi:10.1002/9780470755167.ch15]

RIESSMAN, C.K. (1990) Strategic uses of narrative in the presentation of self and illness, Social Science and Medicine 30 (11): 1195-1200. [doi:10.1016/0277-9536(90)90259-U]

RIESSMAN, C.K. (1993) Narrative Analysis. Newbury Park CA: Sage.

RIESSMAN, C.K. (2003) Performing identities in illness narratives: Masculinity and multiple sclerosis, Qualitative Research 3(1): 5-33.

RIESSMAN, C.K. (2008) Narrative Methods for the Human Sciences. Thousand Oaks CA: Sage.

ROBERTS, B. (2004) Health narratives, time perspectives and self-images, Social Theory and Health, 2 (2):170-83. [doi:10.1057/palgrave.sth.8700027]

ROCK, P. (1979) The Making of Symbolic Interactionism. London: Macmillan.

SAKALYS, J.A. (2003) Restoring the patient?s voice: The therapeutics of illness narratives, Journal of Holistic Nursing, 21 (3): 228-41. [doi:10.1177/0898010103256204]

SANDELOWSKI, M. (2002) Reembodying qualitative inquiry, Qualitative Health Research, 12 (): 104-15.

SKOTT, C. (2002) Expressive metaphors in cancer narratives, Cancer Nursing 25 (3): 230-35. [doi:10.1097/00002820-200206000-00011]

SMITH, B. (2007) The state of the art in narrative inquiry: some reflections, Narrative Inquiry, 17 (2): 391-98.

STANLEY, L. (1992) The Auto/Biographical I. Manchester: Manchester University Press.

STEPHENS, C., Carryer, J., Budge, C. (2004) To have or to take: discourse, positioning, and narrative identity in women?s accounts of HRT, Health, 8(3): 329-50. [doi:10.1177/1363459304043473]

TAYLOR, S. (2006) Narrative as construction and discursive resource,Narrative Inquiry, 16, 1: 94-102. [doi:10.1075/ni.16.1.13tay]

THOMSEN, D.K. and Jensen, A.B. (2007) Memories and narratives about breast cancer: Exploring associations between turning points, distress and meaning, Narrative Inquiry, 17 (2): 349-70.

TILLY, C. (2006) Why? Princeton NJ: Princeton University Press.

VOYSEY PAUN, M. (2006) A Constant Burden: The Reconstitution of Family Life. 2nd edition, Aldershot: Ashgate.

WILLIAMS, G. (1984) The genesis of chronic illness: Narrative re-construction, Sociology of Health and Illness, 6 (2): 175-200. [doi:10.1111/1467-9566.ep10778250]

WORTHAM, S. (2000) Interactional positioning and narrative self-construction, Narrative Inquiry, 10, 1: 157-84. [doi:10.1075/ni.10.1.11wor]

YOUNG, R.L. (1997) Account sequences, Symbolic Interaction, 20, 3: 291-305. [doi:10.1525/si.1997.20.3.291]

ZINN, J.O. (2004) Health, risk and uncertainty in the life course: A typology of biographical certainty constructions, Social Theory and Health, 2: 199-221. [doi:10.1057/palgrave.sth.8700033]