Methodological introduction and overview
In this article, I consider the turn to narrative theory that swept through the medical humanities in the 1980s and 1990s. My primary goal is to offer what may be the first historical overview of the field in its decisive second phase (the first running roughly from 1960 to 1980; see Vickers, ffrench, and Lefève in this issue). My focus will be on the efforts of scholars to mobilize ‘narrative’ and related terms to develop the medical humanities as a distinctive field of study. My approach will be genealogical, aiming to identify and delineate the various narrative theories that were put to use, and the aims they were intended to serve.
I believe that the centrality of narrative thinking that took hold of the field in this phase enabled it to make unprecedented intellectual advances. Nothing that preceded or followed it bears comparison with those achievements. Specifically, narrative thinking enabled the medical humanities to make a powerful and wide-ranging critique of biomedical rationality. I will summarize this critique in the second half of this article, devoting the first half to its elaboration. For a time at least, narrative also gave the field the concentration and stability of a determinate discipline. Indeed, I will suggest that in this period the medical humanities acquired something approaching a unifying purpose.
The medical humanities are best understood as practices of description. The scholars who pioneered the field did so by reframing aspects of health and healthcare-related experience the better to understand hidden aspects of both (see e.g. Cassell, 1969; Coles, 1967; Zaner, 1964). It was a ‘critical’ task in the sense in which the Frankfurt School used that term.
1
I shall argue that the family of projects that came together under the sign of ‘narrative’ in the 1980s and 1990s has origins that go back almost a century earlier, to Emile Durkheim, one of the founding fathers of sociology.
It is sometimes said that the medical humanities arose as an adjunct to medical care. Clinicians, alive to the dizzying complexity of their task, turned to the humanities and the social sciences for guidance. This view is often offered as an origin story for the medical humanities in the UK (usually referring in the first instance to the USA). On this view, they grew out of a concern to make medical practice more humane (Woods, 2012). Without denying the existence of this strand in the field's history, it is important to note that it is only one strand. The medical humanities are also rooted in distinctly academic ways of looking at the world. Sometimes these ways of looking sprang from university disciplines or quasi-disciplines: the meaning-oriented (hermeneutic) tradition in social anthropology and depth psychology were always central, as was phenomenology. Sometimes they issued from movements with points of anchorage in medicine. Anti-psychiatry would be a case in point, as would popular writing about death (e.g. by Geoffrey Gorer [1955] and Elisabeth Kübler Ross [1969]). Sometimes they grew from shared values. Many of the pioneers in the field in the USA were committed Christians (E. A. Vastyan, Edmund D. Pellegrino, H. Tristram Engelhardt, Robert Coles, Ronald Carson, Samuel Banks). The Society for Healthcare and Human Values, founded in 1969 – the world's first professional association of medical humanities scholars – was controlled by the Committee on Healthcare and Human Values, which was explicitly Christian in orientation. These preoccupations were more foundational, in my view, than the clinical focus. The medical humanities were fundamentally about revealing the human values surrounding health. The Hippocratic notion of the ‘whole person’ was often invoked in many different forms (see e.g. Canguilhem, 1991[1966]; Cassell, 1976, 1984; Coles, 1989; Pellegrino, 1980; Zaner, 1981). Some founders of the narrative-based medical humanities, like Elliot G. Mishler, did talk about making medical consultations more humane, but Mishler still thought the lifeworld was more important than the consulting room. Most if not all of the works I will consider are focused on the lifeworld.
The medical humanities were and are a collective endeavour encompassing many different aims. Different disciplines had, and continue to have, their own stakes in the field. It is not surprising, therefore, that the term narrative was taken up in many distinct senses, or that it was imbued with a variety of purposes. Disentangling these influences and purposes will be at the heart of my genealogical approach.
In his idiosyncratic masterpiece Keywords (originally published 1976), Raymond Williams recalled the effect of returning to university after the Second World War and hearing the word culture being used much more frequently. He noted that it seemed to be acquiring two new primary meanings: ‘In the study of literature … some central formation of values … [and] in more general discussion, a use which made it almost equivalent to society: a particular way of life – “American culture”, “Japanese culture”’ (Williams, 1983: 12). With a 30-year retrospect, it was clear to Williams that these two new meanings reflected the ascendancy of English and sociology in British post-war university life, and at Cambridge in particular. These disciplines rested, respectively, on the literary critical tradition beginning with Matthew Arnold (for whom criticism meant most fundamentally the criticism of national culture) and Karl Mannheim's work on the sociology of knowledge, which gave a central place to cultural figures (writers and intellectuals for the most part). As Williams began to reckon with these new meanings, he realized they ran in parallel with changes in the frequency of certain other words: ‘The words I linked it with, because of the problem its uses raised in my mind, were class and art, and then industry and democracy. I could feel these five words as a kind of structure’ (ibid.: 13).
Narrative did not rate its own entry in Williams' original Keywords, but few would dispute its claim to keyword status today.
2
In the period with which I am concerned, many distinct clusters of words constellated around narrative. In almost every case, narrative was invoked in order to expand the analytic power of some other term with its own disciplinary history and its own satellite structure of related words. I shall be concerned with four main discipline-based accounts of narrative: social anthropology, cognitive anthropology, social psychology, and bioethics. They shared some assumptions with one another, but at the outset, at least, there was no single overarching concept of narrative linking them all.
The rise of the narrative-based medical humanities coincided with, and drew strength from, two enormously significant developments in the health sciences. The first was the rise of the biopsychosocial model of health and illness (Bolton and Gillett, 2019; Engel, 1977). A fundamental tenet of the biopsychosocial model is that meaningful generalizations can be made about psychosocial causation across the whole human species. Since at least the ancient Greeks, Western medicine has always recognized psychosocial causation as a possibility, especially in connection with mental illnesses; what has been in doubt is how far disease processes in general can be ascribed to a psychosocial cause. Over the 1980s and 1990s, the scientific basis for such generalizations were made possible by large-scale epidemiological studies. Almost all the major theorists of the narrative-based medical humanities were committed to developing the biopsychosocial model and saw narrative as a way of doing so. Arthur Frank was perhaps the only major theorist never to make reference to the model. The critique of biomedical rationality that I will describe in the second half of this article is essentially biopsychosocial.
The second major development was the refounding of medical anthropology along Geertzian lines by a group of social anthropologists based at Harvard. Byron Good has remarked that for much of the 20th century medical anthropology was regarded by other anthropologists as not quite respectable. Part of the specialty was seen, with some justification, as crypto-racist because it looked for ‘primitive’ equivalents for the institutions of Western medicine. The most interesting work of the mid-century, allied to Talcott Parsons’ notion of the sick role, came from sociology, and addressed the health experiences of Westerners (Parsons, 1951); from an anthropological point of view, Parsons’ theory was perhaps less satisfying because it did not defamiliarize Western medicine very much or make it accountable to other medical systems. The fundamental idea of the new medical anthropology was that consultations with healers (in whatever system) are, in Leon Eisenberg's words ‘requests for interpretation. Patient and doctor together reconstruct the meanings of events in a shared mythopoesis’ (Eisenberg, 1981: 245). Jens Brockmeier has shrewdly observed that the narrative-based medical humanities owed far more to the interpretative turn of the 1970s than they did to the narrative turn of the 1980s (Jones and Tansey, 2015: 41). I think this is fundamentally a reflection of the immense influence of the Harvard group on the medical humanities more generally.
For the Harvard anthropologists and those seeking to build on their work, narrative overshadowed and eventually subsumed ‘explanatory models of health and illness’ (Arthur Kleinman, Leon Eisenberg, Linda Garro) and ‘semantic illness networks’ (Byron Good). The satellite structures to the concept of narrative included illness meanings, psychosocial experience, and the pervasive notion that culture itself performed authoritative acts of interpretation. For anthropologically focused psychological contributors, narrative belonged to a complex of terms designed to highlight the workings of intentionality in social settings (Cheryl Mattingly, Jerome Bruner). Their structure of related terms included folk psychology and social conflict. Social psychologists such as Elliot Mishler whose commitments were to anti-psychiatry, ethnomethodology, and the Frankfurt School, turned to ‘narrative’ as a way of highlighting the ideological dominance of the managerial-technocratic elite. Again, ‘conflict’ was central, but so were ‘social norms’ and ‘rational action’. Finally, bioethically oriented scholars used narrative as a bridging concept linking selfhood and personal identity with social and historical context.
Medical humanities programmes in universities represented important channels of influence for these disciplinary innovations. The thinkers I am concerned with here were trying to establish research programmes that would deepen thinking about medical practice and health experiences. There was nothing amateurish about them.
In all of the theories I have identified, narrative was used as a term to capture a dynamic, often very rapid, interaction or set of interactions. This is one reason why, contrary perhaps to what one might expect, literary theories of narrative had a comparatively small effect. Most literary theories of narrative go to work on texts whose shapes are relatively fixed. (Kenneth Burke's theory of ‘dramatism’ is an exception, and it is no coincidence that that his is the literary theory that figures most prominently in the narrative-based medical humanities; Burke, 1945). Speaking at the Wellcome Trust Witness Seminar on the rise of narrative methods in health care, Kathryn Montgomery Hunter, who has a PhD in English, recalled,
As a group we were, I’d say, dissenters from formalism; that is we’d been trained in New Criticism but for the most part – given the fields we were interested in – you can’t live a New Critical life in eighteenth-century studies, you’ve got to look up some footnotes, you’ve got to know what things mean. And comparative literature is the same.… By and large, though, we were not taken with French deconstruction as the 1980s and 1990s went on – we knew about it but it didn’t really apply to teaching in a medical school. (Jones and Tansey, 2015: 6)
I suggest that insofar as literature and literary study made a contribution to the narrative-based medical humanities, they did so by mobilizing psychological and anthropological theories to expand the notion of the biographical. This was the special mission of the journal Literature and Medicine for the first 15 years of its existence at least. The work of Balint, Luria, and Sacks was especially important to that journal. In many respects, its contributors anticipated the turn to life writing made by many contemporary health humanities writers (e.g. Barry, Murray, Rakoczi, Salisbury, Vickers, Whitehead, and Zimmermann). It is also the case, as I shall demonstrate in relation to the rise of narrative bioethics in this period, that canonical literature and literary theory played an important part in modelling what was at stake in ethical inquiry (e.g. by highlighting the constructedness of case narratives and of medical knowledge more generally); nonetheless, they were accorded no primacy of their own.
I will examine around 15 years of scholarship, beginning with Arthur W. Kleinman's Patients and Healers in the Context of Culture (1980); through Elliot G. Mishler's The Discourse of Medicine (1984); Linda C. Garro's unpublished 1983 doctoral thesis (‘Variation and Consistency in a Mexican Folk Illness Belief System’), as well as her papers from the 1980s and early 1990s and the special number of Social Science and Medicine that she co-edited with Cheryl Mattingly in 1994; Jerome S. Bruner's Actual Minds, Possible Worlds (1986) and Acts of Meaning (1990); Howard Brody's Stories of Sickness (1987); Arthur Kleinman's The Illness Narratives (1988) and Writing at the Margins (1994); Mattingly's (1989) MIT doctoral thesis (‘Thinking With Stories: Story and Experience in a Clinical Practice’); Kathryn Montgomery Hunter's Doctor's Stories: The Narrative Structure of Medical Knowledge (1991); Ann Hunsaker Hawkins’ Reconstructing Illness: Studies in Pathography (1993); Byron Good's Medicine, Rationality and Experience: An Anthropological Perspective (1994); and Arthur Frank's The Wounded Storyteller (1995).
The remainder of this article unfolds as follows. In the first half I briefly consider a period in the history of the medical humanities – roughly 1960–80 – before the so-called ‘narrative turn’ in order to demonstrate that narrative had enjoyed a special methodological status in the medical humanities more or less from their inception (and quite distinct from the one it acquired after 1980). I offer detailed characterizations of the four aforementioned accounts of narrative (social anthropological, cognitive anthropological, psychological, and bioethical), focusing on why the scholars associated with each discipline turned to narrative and what they did with the concept. In the second half of the article, I summarize the critique of biomedical rationality that the narrative-based medical humanities achieved and ask why they eventually ran out of steam.
I have not engaged at length with Angela Woods’ two papers on the use of narrative in the medical humanities (2011, 2012). Woods mostly addresses a later phase in the field's evolution and is largely concerned with the bioethical strand. Within these parameters I agree with much of what she says. But, as will become apparent, most of the writers I consider here did not think that humans were ‘naturally narrative’, still less that patients ought to have a rich autobiographical sense.
Finally, for the avoidance of misunderstanding, let me emphasize that I am not advocating a return to narrativity. The development of the field since the early 2000s has enabled it to consider aspects of health that had not been conceptualized in much detail before then (neuroscience, the so-called ‘omics’ disciplines, and AI, to name but three). The emphases on race, class, gender, sexuality, and disability have added real depth. I would not turn the clock back. But I do believe that the health humanities today cannot be understood without reference to this extraordinarily fertile period in their evolution and that it deserves greater acknowledgement.
Narrative in the medical humanities before the narrative turn
Even before the narrative turn, narrative had a special methodological status in the medical humanities. The medical humanities have often gathered strength by placing themselves in the slipstream of developments in other fields. One that has proved particularly productive – I would say the most productive in the field's history – is the peculiar link between psychiatry and ethnography. Psychiatry is a medical specialty without a provable medical basis. Devotees of psychological accounts of mental illness have often tried to test their theories ethnographically. The work of Erik Erikson proved particularly inspirational in these endeavours. Erikson had moved to the United States from Vienna in 1933 and, despite holding no university degrees of any kind, was recruited by Harvard University to take part in a study of education among Sioux children in South Dakota. The experience, which had a lasting impact on him, is described in two chapters of his most important book, Childhood and Society (1950). The children Erikson studied had recently been compelled to take part in state education, without any account having been taken of what it meant for the communities they came from. For them, it was a moment of cultural disintegration. Erikson's work in South Dakota shaped his famous account of eight stages of psychosocial development. His encounter with the Sioux people is described entirely in narrative, ethnographic terms.
In similar fashion, in his three widely read pathographies of Hitler (1942), Luther (1958), and Gandhi (1969), Erikson turned to narrative history in order to subject his own psychoanalytic theories of child development to a powerful transcultural challenge. Luther had undergone a crisis of identity over his decision to become a priest in defiance of his father, who berated him for his ‘demonic’ nature (the son hallucinated regularly). Hans Luther consented to his son's wish only after two of his other sons had been struck down by lightning, which he took to be a sign from God. The whole of the great reformer's struggle, as Erikson recounts it, centred on an effort to find in the church an alternative to his father's view, only to find that the church itself was diabolical. Like the Sioux children in mid-1930s South Dakota, Luther was in the grip of what Erikson would famously describe as an ‘identity crisis’. Psychoanalytic theory and historical narrative work in tandem in Erikson's oeuvre, as contrary elements holding one another in tension.
Erikson took pains in his pathographies to show how the very notion of mental illness was saturated with cultural assumptions. A reader of the first edition in 1958 might have wondered if Luther was mad; but, in an environment where demonic possession and signs from God were regarded as legitimate facets of everyday life, such a designation would miss the larger point that the human personality is ‘a combination of capacities created in the distant past and of opportunities divined in the present; a combination of totally unconscious preconditions developed in individual growth and of social conditions created and recreated in the precarious interplay of generations’ (Erikson, 1958: 14). It is in doing something with this admixture that mental health and illness emerge.
In 1960, Erikson became a professor at Harvard, where he mentored two psychiatrists whose work was to prove decisive for the medical humanities. The first, Robert Coles, began teaching medical students at Harvard in the 1960s and became its first professor of the medical humanities in 1977. The second was Robert Jay Lifton. Both Coles and Lifton had a wide reach. They were both regular contributors to the New York Review of Books and the New York Times. Coles won two Pulitzer prizes in the early 1970s, and Lifton won the National Book Award in 1967 for his Death in Life, published that year.
In 1958, Coles went to live in Mississippi, to complete his military service as the chief of an army neurological unit. Slowly, he became aware of the scale of discrimination black people faced not only in Mississippi but all across the South. Coles' training had been in child psychiatry, not neurology, and, as he makes clear in the first volume of his five-volume study Children of Crisis (1967), his notions about children's psychological development were derived from Erikson and from Erikson's own analyst and mentor, Anna Freud. With the introduction of desegregation in New Orleans, and with Erikson's South Dakota experience in mind, he decided to study the families of the first children to attend desegregated schools. This was not easy to do, as many white parents tried to stop desegregation by not allowing their children to go to school at all. Coles acted as a participant-observer in most of the major conflicts involving children on behalf of the civil rights movement. He told his subjects that he was ‘a physician interested in how people managed under the kinds of difficulties then in progress’ or ‘a research physician specializing in problems of human adjustment under stress’ (ibid.: 27, 299). To that end, he carried out long ethnographic interviews with hundreds of children and their parents and teachers. Most of his black interviewees were extremely impoverished.
A point Erikson had made repeatedly was that psychoanalysis had overemphasized the pull of regression during childhood and had disregarded the forward momentum in pre-adult development. Coles was ready to find that his research subjects had been traumatized by the experiences they went through. Some were, but on the whole the children were surprisingly robust. Coles found himself puzzling over this resilience. In a way, he had stumbled upon a population that would enable him to study the causes of mental health and not just mental illness. He was discovering for himself a perspective that had antecedents in Erikson's work and reverberated with the optimism of socially oriented medicine in the 1960s and 1970s; later, Aaron Antonovsky (1979) would call it salutogenesis (i.e. the causes of health). Because of the primacy of ethnography in Coles’ work, his books became increasingly tilted against theory and towards narrative, though the psychoanalytic account of child development remained in the background.
Robert Jay Lifton's Death in Life (2012[1967]) was the other monumental work of Eriksonian research. This book is renowned for popularizing the notion of ‘survivor guilt’. Lifton's chosen subjects were not children but the adult survivors of the bomb dropped on Hiroshima by the US in August 1945. Like Coles, he wanted to understand the psychological consequences of exposure to a highly abnormal reality (abnormal, that is, by the standards his readers would have taken for granted). The adults Lifton interviewed regarded themselves as being in some sense responsible for the bomb and for the ‘cosmic disruption’ it had brought into the world. They seemed psychically numb. Lifton introduced the notion of ‘formulation’ to consider the task confronting his subjects after 1945.
Formulation includes efforts to re-establish three essential elements of psychic function: the sense of connection, of organic relationship to the people as well as non-human elements in one's life space, whether immediate or distant and imagined; the sense of symbolic integrity, of the cohesion and significance of one's life, here including some form of transcendence of the A-bomb experience; and the sense of movement, of development and change, in the continuous struggle between fixed identity and individuation. (Lifton, 2012[1967]: 367)
The atom bomb had laid upon his subjects a sense of disturbance that was almost religious in its reach. As with Coles, a psychoanalytic picture, largely derived from Erikson, informed his account of his subjects’ experiences.
Lifton's use of the term formulation is known to many in the medical humanities thanks to Anne Hunsaker Hawkins' use of it in Studies in Pathography (1993). But in fact Eric Cassell (1972) had spotted its significance long before her as a potential metaphor for the transformations attendant upon a major diagnosis.
No richer picture of the destructive effects on living that result from an injured relationship between self and body comes to mind than that so beautifully written by Lifton in his study of the survivors of Hiroshima. There, in those unable to transcend it, the self remains shackled to a body seen as somehow already dead or permanently tainted in some inexplicable way. The interrelationships of the two independent but inseparable parts of being – the symbiotic halves of existence – humble us by their mystery and complexity. (Cassell, 1972: 537)
Both these authors’ work had an immediate impact on medical humanities theories concerning the boundaries of illness as a concept and its dependence on cultural context (see e.g. Cassell, 1969, 1972). Helene Scott-Fordsmand (2023) has recently called for a ‘reversal’ of the medical humanities, characterized by a focus ‘on the cultural and humanistic productivity of medicine’. The entire Eriksonian tradition within the medical humanities was based on such a focus. It attempted to articulate a socially progressive vision of life using the healer's point of view as its starting point. It is not unlike what John Berger and Jean Mohr attempted in their classic book A Fortunate Man: The Story of A Country Doctor (1967).
Social anthropology rediscovers sickness
For Erikson and those writing in his wake, there was no better way of refining a psychological theory than that of putting it to work in a culturally unfamiliar setting. Behind this assumption lay an aspiration derived from the classical world and which was still commonplace until the final decades of the 20th century, namely, that knowledge should be universalizable. Erikson wanted his theory of child development to shine a light on psychology and culture in general. It could only be validated for use in the contemporary world by something like the ethnographic interviews carried out by Coles and Lifton. Conventional psychoanalytic case histories of, say, children and adolescents from affluent East Coast backgrounds would not have enabled the role of culture in shaping mental health to stand out to the extent that Coles’ interviews with very poor children had done.
In fact, a much more professional and systematic set of efforts to make meaningful transcultural comparisons in relation to health and illness got underway not long after Coles completed his first monumental sequence. Towards the end of the 1970s, a new generation of anthropologists tried to bring the analysis of health and sickness into dialogue with feminism, critical race theory, and Marxism. But the absence of a credible overarching framework for medical anthropology was acutely felt, and in the late 1970s and early 1980s a powerful effort was made by phenomenological social anthropologists to provide one.
Clifford Geertz (1973) had presented religion, aesthetics, and common sense as cultural systems enabling humans to constitute their experiential worlds symbolically. The Harvard group initially tried to examine healing systems as particularly impactful cultural systems in the Geertzian sense. The group came together in the early 1980s at the invitation of Leon Eisenberg, a psychiatrist with a strong interest in transcultural psychiatry. They set up what became the Department of Global Health and Social Medicine.
Eisenberg's first hire was Arthur Kleinman, who, as well as being a liaison psychiatrist, had carried out extensive studies of Chinese medical and health experience with his wife Joan Kleinman, a Sinologist. In Patients and Healers in the Context of Culture (1980), Kleinman initially conjectured that at the heart of every healing system were ‘explanatory models’ of health and disease. The framework was designed to facilitate cross-cultural comparisons. In this early work, the principal contrast was between biomedicine as an explanatory model and traditional Chinese medicine organized around the concept of qi. It was necessary to study healers, Kleinman contended, because they embodied the cultural prestige of explanatory models. Healers didn’t have to practise anything akin to medicine. They just had to embody the theory and practices underpinning a particular approach. In the West, faced with a physical indisposition, people might call a doctor or a priest. Both would be healers in Kleinman's sense. Kleinman identified five characteristics of healers. First, healers ‘construct illness as a psychosocial experience’. Second, they provide ‘general criteria to guide the health care seeking process and to evaluate treatment approaches that exist prior to and independent of individual episodes of sickness’. Third, they manage ‘particular illness episodes through communicative operations such as labeling and explaining’. Fourth, they carry out ‘healing activities per se, which include all types of therapeutic interventions, from drugs and surgery to psychotherapy, supportive care, and healing rituals’. Finally, healers take responsibility for therapeutic outcomes (Kleinman, 1980: 70–1).
In The Illness Narratives (1988), Kleinman pushed healers into the background to focus on the lived experience of chronic illness. Using a variety of case histories from his practice, he was now concerned with the range of actors involved in shaping and sustaining explanatory models. The book's central argument was that ‘the meanings of chronic illness are created by the sick person and his or her circle to make over a wild, disordered natural occurrence into a more or less domesticated, mythologized, ritually controlled, therefore cultural experience’ (ibid.: 48). The task he set himself in The Illness Narratives was to show how culture held chronic illness experience in place. The ‘sick person and his circle’ became the central actors.
In Patients and Healers in the Context of Culture, the curative claims implicit in explanatory models were paramount. In The Illness Narratives, however, the term was expanded to encompass ‘the notions that patients, families, and practitioners have about a specific illness episode’ (Kleinman, 1988: 121). These notions could be extremely idiosyncratic and irrational. The central focus now was on the innumerable sources of illness meanings at the disposal of the sick and their entourage. Healers and their skills were just one such source. It was for this reason that Kleinman first had recourse to the concept of narrative. What he called ‘illness narratives’ were his own attempts to capture the way in which a variety of explanatory models meshed dynamically within a lifeworld.
Explanatory models are most often tacit, or at least partially so. Not infrequently, they contain contradictions and shifts in content. They are our representations of the cultural flow of life experience; consequently, as the epigraph to this chapter suggests, they congeal and unravel as that flow and our understanding of it firms up in one situation only to dissolve in another. (Kleinman, 1988: 121–2)
Kleinman's illness narratives were knowingly fuzzy around the edges. There were, nonetheless, three guiding principles bringing order to the whole enterprise. The first of these was the primacy of the body. Westerners think of their bodies in Cartesian terms as a ‘thing’ attached to a mind. In non-Western cultures, the body was often seen as ‘an open system linking social relations to the self, a vital balance between interrelated elements in a holistic cosmos’. Pathognomonic signs could not be understood without reference to the prevailing ‘normative conceptions of the body's status within a given culture’. This focus on ‘how bodily experience itself is influenced by meanings, relationships and institutions’ was a core commitment of the original Harvard group (DelVecchio Good, Good, and Fischer, 1988: 7). The Illness Narratives offers a multi-level theory of illness meanings, beginning with normative accounts of the body, through normative conceptions of illness or ill-being, culminating in the construction of what Kleinman called ‘clinical reality’.
The second point of reference was the biopsychosocial model. ‘In the biomedical model’, Kleinman observed, ‘the disease is an occluded coronary artery; in the biopsychosocial model it is a dynamic dialectic between cardiovascular processes (hypertension or coronary artery insufficiency), psychological states (panic or demoralization), and environmental situations (a midlife crisis, a failing marriage, the death of a parent from the same disorder)’. Each of Kleinman's illness narratives contains a penetrating account of this dynamic dialectic.
Finally, Kleinman had a quasi-Bionian commitment to the idea that a physician's even silent grasp of the cultural underpinnings of chronic illness could itself be therapeutic (Bion, 1963). ‘Meanings can heal’, as he puts it in his recent memoir (Kleinman, 2019: 30). He hoped that his meanings-oriented approach to treatment would be taught in medical schools by teachers of ‘the medical social sciences and humanities’. Chapter 15 of The Illness Narratives sets out a method for applying it. It's rather wooden, in my view, but the core proposal is radical: physicians should become ethnographers of their patients’ lifeworlds, empathically witnessing their struggles and helping them to cope practically with the psychosocial crises characteristic of chronic illness experience.
In The Illness Narratives, narrative was a hermeneutic tool for use by specialists (academics and clinicians) seeking to understand long sequences of complex interactions. There was a tacit assumption that a grasp of medical anthropology was necessary to write one. That assumption fell away in Kleinman's next book, Writing at the Margin (1994). Under the influence of his colleagues Byron Good and Mary-Jo DelVecchio Good and of Pierre Bourdieu and Luc Boltanski, he now began to interest himself in the idea that sickness was just biomedically accredited suffering. From this point on, his central preoccupation became the violence of everyday life – social suffering, to use Bourdieu's coinage – and the role of the institutions of medicine in amplifying or mitigating it. Following the model in Bourdieu's massive La misère du monde (1999[1993]), he now published patient interviews in which the patient voice was allowed to appear unredacted, followed by his own commentaries.
Byron Good's trajectory overlaps with Kleinman's to a large extent. He began by trying to differentiate healing systems by describing distinct ‘semantic illness networks’. These were ‘the network[s] of words, situations, symptoms and feelings which are associated with an illness and give it meaning for the sufferer’ (Good, 1977: 40). The overlap with explanatory models was commented on by Young (1982). Healers, however, never had the prestige in Good's thinking that they had in Kleinman's. Semantic illness networks had more in common with Kleinman's (1988) account of explanatory models, but they were fixed linguistic entities. With Mary-Jo DelVecchio Good, Good was one of the first medical anthropologists to recognize the importance of epidemiological research for the kind of anthropology the Harvard group were creating. And the Goods too were committed to the biopsychosocial model. As early as 1981, they recognized that the pattern of disease prevailing in any culture would reflect the differential make-up of psychosocial causes peculiar to each (Good and DelVecchio Good, 1981). Perhaps as a result of Mary-Jo DelVecchio Good's training as a sociologist, the Goods were more concerned than Kleinman was to use patterns of ill-being to understand the social worlds in which they arose.
They began using the term narrative around the same time as Kleinman as one of several ‘naturally-occurring discursive forms through which emotion and illness are articulated, communicated, and experienced’ (DelVecchio Good, Good, and Fischer, 1988: 3). Over the next few years, they sought to theorize narrative as a medium of social action in the world.
The Goods concurred with Kleinman that the health sphere might be described as a set of explanatory models operating within a wider field of explanatory models, curative and non-curative. They were also profoundly influenced by his call in 1991 (with Joan Kleinman) for the ‘ethnography of experience’ to be at the heart of medical anthropology. But narrative was a curiously undertheorized entity in Kleinman's work during the 1980s. It was merely the resultant of his account of how explanatory models collide and transform one another in the process. Kleinman cites Erikson and Balint and Mishler and Cassell, but nowhere does he place his conception of narrative in dialogue with any other narrative theory. Here, the Goods were able to extend the Harvard paradigm considerably. They took their bearings on narrative largely from the child psychologist Jerome Bruner but also borrowed from Ricoeur's Time and Narrative trilogy (1990) and Wolfgang Iser's reader-response theory (1993).
Bruner had been interested in the role of narrative in children's learning processes since the early 1970s. Mutually dependent communal life requires us to be able to explain our point of view to others. We have to say ‘how things are’ or ‘how they were’ and, because of that imperative, we find ourselves in a world of stories. So pressing is this demand that children learn to organize their experience into narrative structures from a very early age. Indeed, Bruner hypothesized that babies and toddlers spend as much time grasping narrative structures as they do learning language. They often tell stories that don’t make sense – evidence, he believed, of the primacy of narrative structures over language acquisition. (A similar phenomenon can be seen in people with dementia in whom the storytelling capacity often remains intact, even when the raw materials of a story are missing.) A point Bruner never tired of making was that narratives are ‘subjunctive’ in nature. What he meant was that narratives enable us to try out experiences in advance, without necessarily committing to them in real life. Narrative capacity is part of the predictive equipment of the brain in general and of the child's brain in particular. Nowadays we would probably place this idea in the context of the Bayesian brain (Friston, 2012). The Goods also highlighted this predictive dimension of narrative and called attention – in a way Kleinman had not – to individual patient narratives.
Although neither of them was medically qualified, the Goods were as committed as Kleinman to narrative's curative power. Narrative, for them, was a way of reclaiming agency. As such it is a crucial resource at the disposal of the sufferer. ‘Narrativization’, Byron Good wrote, in a claim with which Bruner would have concurred,
is a process of locating suffering in history, of placing events in a meaningful order in time. It also has the object of opening the future to a positive ending, of enabling the sufferer to imagine a means of overcoming adversity and the kinds of activities that would allow life experience to mirror the projected story. (B. Good, 1994: 128)
When he considered narrative ethnography Good turned to Paul Ricoeur and Wolfgang Iser. Listening to accounts of ‘epilepsy’ in Turkey and Iran, he found himself trying to decipher a world-system in much the way the reader of a novel must do.
The reader is engaged imaginatively in constructing a ‘virtual plot’, in attempting to extract configuration from what has been heard, in determining the nature of relationships among events and characters, separating the related from the irrelevant, conceiving potential outcomes in the world of the text. The activity of ‘emplotting’ thus has a special affinity to the experience of persons with debilitating chronic illness, and the literature on this concept is especially useful in understanding their stories. (B. Good, 1994: 145)
Emplotment or mise-en-intrigue was a Russian formalist term Ricoeur had brought to wide notice. The sick, Good observed, often find themselves with competing narratives or plots. They live with a variety of mimetic models of themselves. And that is why Bruner's insistence on narrative's subjunctive quality and on openness to change remained so important in the illness case.
The importation of these additional considerations altered the emphasis in the Harvard group's interest in narrative. Kleinman had introduced narrative as a method of capturing the contribution of culture in the unfolding of disease processes. Culture itself was depicted as a kind of massive confrontation of multiple narratives in the group's work during the 1990s. And explanatory models and semantic illness networks remained in the picture, even if they had receded far into the background by the mid 1990s.
Cognitive anthropology and narrative
As their names imply, explanatory models and semantic illness networks are cognitive constructs. It is not surprising, therefore, that cognitive anthropologists also took a strong interest in them. Linda Garro holds two PhDs, one in social psychology (Duke University, 1982), the other in cognitive anthropology (University of California at Irvine, 1983). Her work in the 1980s tracked Kleinman's closely, largely because she was interested in subjecting his theories to rigorous empirical tests. She wanted to know how individuals handle explanatory models: the extent to which these require conscious formulation; how far individuals need to take them to heart in order to be influenced by them. Her dual training in cognitive anthropology and social psychology made her uniquely placed to consider these questions. Most of us are aware of a range of models in the culture to explain symptoms and signs. Although the range is stable, our allegiance to any of them may be more mobile. How can we tell which models are in play for a given person at a given moment in time? Does not the early Harvard approach imply a degree of invariability seldom found in real life? In her unpublished anthropology PhD (1983), Garro compared the folk medical beliefs of ‘folk-curers’ and ‘non-curers’ in rural western Mexico and, as Kleinman might have predicted in 1980, she found that the curers had a more stable set of beliefs and were more inclined to agree with one another than non-curers. What was interesting about the Mexican folk curers was that their knowledge was not seen by the communities they served as particularly esoteric (a very different situation from biomedicine!), though they were assumed to have a better command of it than non-curers. Garro went on to hypothesize that folk curers exemplified ‘cultural learning’ as the repositories of a folk tradition (Garro, 1983, 1986).
As her work developed in the 1980s, she began to study the Ojibway native American community in Manitoba, Canada, a community that had access to biomedicine and an indigenous healing tradition. She chose to look at hypertension because it was a biomedical construct with no obvious equivalent in Ojibway folk medicine. The questions she was interested in were simple:
The presence of individual variation raises questions about the nature of shared knowledge. What is the basis for determining whether a cultural model exists? Further, in situations where support is found for a cultural model, what variation can be explained within this model and what is truly idiosyncratic? (Garro, 1988: 99)
Garro used a statistically based consensus model to measure the hegemony of a variety of models and the degree to which individual informants approach the shared standard. She found that hypertension was imagined in culturally distinctive terms by her Ojibway interviewees, with a high level of consensus. Those terms were radically at odds with the biomedical account of hypertension. Cultural learning seemed to lie behind explanatory models.
Although Garro didn’t use the term narrative in her writing until 1990, she turned to it for the same reasons the Harvard group had: explanatory models and semantic illness networks placed too much emphasis on cognitive deliberation; what was needed was a conceptual apparatus that highlighted the dynamism of cultural context; narrative was her preferred means of achieving that end. Like Kleinman, at first she saw narrative as a hermeneutic device for scholars and clinicians but as the 1990s progressed, she gave primacy to the sick person's own implication in the narrative process, reasoning that culture's influence could be seen most clearly in what mattered to the patient.
It was in the 1990s that Garro began her collaboration with Cheryl Mattingly. Mattingly had written a thesis under Jean E. Jackson at MIT on clinical reasoning by occupational therapists at an unnamed acute care hospital in Boston. Mattingly was interested in stories as an action-oriented organizing framework for making sense of experience. She called particular attention to the way occupational therapists used imaginary scenarios to think about their patients’ futures (she explicitly referenced Bruner's account of the subjunctive character of narrative in this connection). When they had to imagine their disabled patients’ lives, they thought in largely phenomenological terms, dispensing with the medical model. Illness experiences, Mattingly concluded, could be expressed only in narrative terms. Following Ricoeur, she stressed narrative's mimetic character. ‘The imitation of action in a story’, she wrote, ‘provides implicit explanations and evaluations’ (Mattingly, 1989: 373). As imitations of actions, stories offered a different record of experience from the more cognitively oriented picture that the Harvard group had originally promoted.
Following MacIntyre and Bruner, Mattingly saw intentional states as the primary building blocks in narratives, a key idea I will explore later on. The therapists’ attempt to imitate their patient's intentional states ‘often gave the patients described in the story a consciousness, a suffering of their own, which “spilled over” a conventionalized reading of them as diagnostic or cultural types’ (Mattingly, 1989: 373).
When the occupational therapists were training others, they told stories that were organised around a conflict between anticipation and desire, on the one hand, and actual events on the other. Therapists told stories about times when they or their patients had expectations which clashed with the clinical events that transpired. Therapists chose situations of conflict and reframing of experience as those which produced expectations as significant and dramatic enough to provide good material for a story. (ibid.: 374)
Crucially, Mattingly emphasized the importance of conflict as the engine of storytelling, as Kenneth Burke (1945) and Hannah Arendt (1958) had done and as Bruner (1990) was about to do. One of the joys of reading Mattingly's case histories is her exquisite ear for conflict. She was perfectly aware of the limitations of narrative as a therapeutic instrument, observing that the ‘clinical stories therapists projected onto new situations often run into trouble because the new situation was often resistant to being scripted by an old story’ (Mattingly, 1989: 380). This meant that meaninglessness became an inevitable dimension of narrative-based clinical practice.
The possibility of doing something and nothing happening, of moving through time and losing it, because time moved while actors did not, of intending things, but having a senseless experience: these are all ordinary expressions of everyday experience of meaninglessness. (ibid.: 386)
Kathryn Montgomery Hunter's Doctor Stories: The Narrative Structure of Medical Knowledge (1991) is in some respects complementary to Mattingly's work. Many of the essays in Montgomery Hunter's book originally appeared as journal articles in Literature and Medicine during the 1980s. Montgomery Hunter spent two years in three major teaching hospitals following ‘willing colleagues on work rounds, to morning report and professors’ rounds, to sign-out rounds and problem conferences, to morbidity and mortality conferences, and to grand rounds’ (Hunter, 1991: xiii). As a literary scholar, what interested her were the numerous ways in which narrative-based reasoning in medicine was filtered through canonical forms of speaking and writing. The core of her argument was that medicine was not a science but was ‘science-using’. She was at one with the Harvard anthropologists that clinical consultations were demands for interpretation, adding the idea that the patients’ interpretative activities were no less searching. Like Mishler, she found that the voice of biomedicine often drowned out the more tentative voice of the lifeworld and she argued that more attention to the latter would result in better care and outcomes. ‘Good decisions about patient care’, she wrote, ‘call for a richer narrative than the traditional medical case’ (ibid.: 149). Montgomery Hunter called on clinicians to develop a more literary sense of the lives of patients, as Coles had done before her.
Psychology and narrative
When we turn to the psychological theorists of narrative, we find a similar split between the social and cognitive perspectives. The former was represented by Elliot Mishler, whose The Discourse of Medicine: Dialectics of Medical Interviews (1984) was arguably the most influential work in the narrative-based medical humanities. Mishler was a radical social psychologist who specialized in ethnography. With Nancy Waxler, he recorded the styles of conversation found in families in which at least one person had been diagnosed with schizophrenia (Mishler and Waxler-Morrison, 1968). He was steeped in critical theory, perhaps most especially conversation analysis.
In The Discourse of Medicine, he studied the psychological impact of routine medical consultations on patients. Biomedicine, he hypothesized, is thoroughly ideological because it eliminates from its discourse the voice of the lifeworld.
The discourse of medical interviews is not to be understood primarily as one dominated by physicians speaking in the voice of medicine with patients intruding their concerns, thereby disrupting its smooth flow. Rather, discourse is revealed as a dialectic between the voices of the lifeworld and of medicine; it involves conflict and struggle between two different domains of meaning. (Mishler, 1984: 121)
Mishler's term ‘the voice of medicine’ is not synonymous with the things doctors say and write. It is the ideological distillation, in the full Marxist sense, of the power of biomedicine. Taking his bearings from the writings of Jürgen Habermas from the early 1970s on instrumental rationality, Mishler argued that medicine transforms practical interests into technical ones, conferring power on the technocratic elite, that is, doctors. The narrative methods Mishler proposes are intended to mitigate an underlying violence in the medical encounter, itself a reflection of underlying violence in the social world.
Mishler states in The Discourse of Medicine that his methods come from sociolinguistics, phenomenology, and ethnomethodology (Mishler, 1984: 29). His phenomenology came from Alfred Schutz's phenomenology of the social world – though, significantly, he takes Schutz to task for minimizing the scale of the social conflict that characterizes even the most ordinary interactions (ibid.: 124).
Mishler's narrative analysis was predicated on the conversation-analytic assumption that social settings bring with them distinct presuppositions and rules that people use when they talk to one another, which are so taken for granted that they pass largely unnoticed. The Discourse of Medicine attempts to specify those rules as they appear in medical consultations. Mishler thought these appeared not only propositionally – through interactants’ statements – but also through ‘features of speech such as intonation, pitch, pacing, volume, filled and unfilled, pauses, non-lexical, vocalisations, full starts, repetitions, interruptions, and overlaps between speakers are omitted from the great variety of printed texts, even when they include quotations’ (Mishler, 1984: 22). Mishler's primary goal in The Discourse of Medicine was to demonstrate that the conversational rules of medical consultations are focused less on care than on social control. In his next book, Research Interviewing: Context and Narrative (1986), he makes a methodological refinement by treating patient responses as narratives. He uses the terms story and narrative in The Discourse of Medicine, but he does not consistently equate patient responses with narratives in that book.
It is, however, in the work of Jerome Bruner that the conflictual account of narrative is expressed most forcefully. In Acts of Meaning (1990), Bruner gave his most detailed statement on the role of narrative in human life. Acts of Meaning attempted to reinvigorate the cognitive revolution that Bruner and others had inaugurated in the 1950s, by bringing the mind and the social world back into psychology. The project had withered, in Bruner's view, by focusing on information processing rather than the cognitive underpinnings of the use of meanings. Narrative was the centre of this reinvigoration. Bruner wanted second-wave cognitive science to enter into dialogue with the other social sciences and with the humanities.
If culture gives meaning to action ‘by situating its underlying intentional states in an interpretive system’ (Bruner, 1990: 34), narrative was how we decoded these. Echoing Geertz and the Harvard group, Bruner asserted that every culture had its own collection of ‘symbolic systems – its language and discourse modes, the forms of logical and narrative explication, and the patterns of mutually dependent communal life’ (ibid.: 35). We internalize these symbolic systems very early on, Bruner thought, largely by assimilating the folk psychology around us. Folk psychology, in Bruner's account, is ‘a set of more or less connected, more or less normative descriptions about how human beings “tick”, what our own and other minds are like, what one can expect situated action to be like, what are possible modes of life, how one commits oneself to them, and so on’ (ibid.: 35). ‘At the core’, Bruner wrote, ‘all folk psychologies contain a surprisingly complex notion of an agentive Self’ (ibid.: 41). In some cultures, such a notion may derive from ‘a world of meanings, images, and social bonds, in which all persons are inevitably involved’ (ibid.: 42). For most of our evolutionary history, we have lived in close quarters with our extended families. The Western pursuit of individuation as the hallmark of fully human status represents a break with those arrangements, but it is not hard to imagine communities in which silence is the norm. In the 1980s US, a different model of an agentive self prevailed, one in which there was a ‘now self’ along with a range of other possible selves – a reflection, Bruner suggested, of ‘the high value placed on keeping one's options open in American society’ (ibid.: 42).
Narrative plays a crucial part in enabling agentive selves to operate in their surrounding culture because it enables them to explain their intentions. Intentions are intrinsically narrative in structure. If we behave in a way that could elicit opposition or conflict, we need to be able to explain our point of view. Narrative deals with the very ‘stuff of human action and human intentionality. It mediates between the canonical world of culture and the more idiosyncratic world of beliefs, desires, and hopes.… It reiterates the norms of the society without being didactic. And … it provides a basis for rhetoric without confrontation. It can even teach, conserve memory, or alter the past’ (Bruner, 1990: 52). ‘Stories’, according to Bruner, ‘make “reality” a mitigated reality’ (ibid.: 97).
The meaningful negotiation of experience in the world with others protects us against stressors and the costly physiology of high stress. It is a biopsychosocial model of narrative that is still serviceable today.
Bioethical approaches to narrative
Of the four major strands in narrative thinking in the medical humanities in the 1980s and 1990s, the bioethical one was most responsive to the wider narrative turn. The first edition of Howard Brody's Stories of Sickness (1987) was the trailblazer – the second edition, published in 2002, was almost a different book. Brody's general approach to narrative was inspired by Alasdair MacIntyre's claim in After Virtue (1981) that a life can never be considered independently of the context in which it arises and contexts change with time. As MacIntyre put it, ‘The notion of a history is as fundamental … as the notion of an action. Each requires the other’ (MacIntyre, 1981: 253). Or again: ‘We cannot … characterize behaviour independently of intentions, and we cannot characterize intentions independently of the settings which make those intentions intelligible both to agents themselves and to others’ (ibid.: 242). Macintyre asserted that ‘stories are lived before they are told’ (ibid.). Modernity encourages us to tell fragmented stories about ourselves because ‘work is divided from leisure, private life from public, the corporate from the personal. So both childhood and old age have been wrenched away from the rest of human life and made over into distinct realms.’ To grasp one's life as a narrative required formidable effort and deliberation, in his view. He did not think humans were ‘naturally narrative’. Brody conjectured that major illness often enforced such an effort upon the sick. A person can only respond to a major diagnosis by asking him- or herself, ‘To which stories do I belong? How can I take my place in them, or must I fashion a new story?’ The core of the narrative process, for Brody, involved the cultivation of ‘self-respect’, a term he took from John Rawls (1971). Sickness alters people's self-respect dramatically in Brody's view – partly because the healthy despise sickness – and much of his book is given over to readings of dozens of literary works in illustration of this claim. Rawls thought it was the job of society's institutions is to support the self-respect of its members (there is a clear overlap with Axel Honneth's [1995] thinking on ‘recognition’). Brody thought that Western societies were especially remiss at supporting the self-respect of the sick.
A noteworthy feature of Brody's book is his inclusion of Jungian accounts of the self (even if he devotes more time to Freud). In Stories of Sickness, the chief Jungian interlocutor is James Hillman, who held that ‘the soul has a basic need to tell autobiographical stories, to historicize, whether or not there is an audience; historicizing is a way of “maintaining an event intact but removed, in a glass vessel so that it can be puzzled over without being identified with.… We historicize to give the events of our lives a dignity that they cannot receive from contemporaneousness”’ (Brody, 1987: 14).
In Studies in Pathography (1993), Anne Hunsaker Hawkins produced an unprecedented comprehensive survey of first-person accounts of illness. Her contribution was fundamentally bioethical. I have compared it elsewhere to Sandra Gilbert and Susan Gubar's No Man's Land (1988) because it brought to notice a large body of literature to which little attention had been paid, and in the process laid the basis for a new academic sub-discipline (Vickers, 2016). Like Gilbert and Gubar, Hawkins’ main aim was descriptive, but she did permit herself a few generalizations about the illness narrative genre. She was the first contributor to the medical humanities to talk about illness in terms of trauma theory. Chief among these was her observation that it bears witness not only to the pervasiveness of mythic thinking in situations of extreme distress but also to its effectiveness. This emphasis on myth had two main sources. One was Robert Jay Lifton's Eriksonian theory of ‘formulation’. Following the Jungian Joseph Campbell, Hawkins proposed that couching an illness in narrative terms was curative because it enabled the sick to discover an underlying unity of their lives: ‘It gathers together the separate meanings, the moments of illumination and understanding, the cycles of hope and despair, and weaves them into a whole fabric, one wherein a temporal sequence of events takes on narrative form’ (Hawkins, 1993: 24–5).
The most important contribution to the bioethical strand of narrative thinking – it is surely the most influential and certainly the most cited monograph in the entire history of the medical humanities – was Arthur Frank's The Wounded Storyteller (1995). Frank argued that the experience of illness had been altered in subtle ways by postmodernity. The ‘narrative surrender’ to biomedical ideology of post-war medicine had alienated the sick. Narrative was the route to agency. ‘Postmodern times’, he wrote, ‘are when the capacity to tell one's own story is reclaimed’ (ibid.: 7). For Frank, different narrative postures reflect the kind of identity crisis that the illness has imposed on the sick person. The narrative typology for which his book is justly celebrated was heavily influenced by literary theory (Frank was a graduate student in sociology at Yale during the 1970s when Yale English was the cockpit of deconstruction). But Frank was also influenced by Joseph Campbell's account of the hero's journey and this accounts for the ‘archetypal’ inflection that Angela Woods has rightly pointed to in Frank's work.
Although Frank cites Alasdair MacIntyre only once, he shares his view that fragmentedness is the chief characteristic of modernity (or postmodernity, as Frank prefers to call it). We have no agreed vocabulary with which to talk to one another about the things that matter most to us as individuals. The contemporary experience of selfhood is not strong enough to sustain such a dialogue because people already feel ‘spoken for’. Postmodernity was consequently characterized by ‘narrative wreckage’. Illness made that predicament doubly acute. A version of this claim had been made by Brody in 1987, as we saw earlier.
At the same time, Frank also suggests that the person who tells their story changes himself or herself by the act of telling. For, as Shoshana Felman and Dori Laub put it in their classic work on testimony, ‘texts that testify do not simply report facts but, in different ways, encounter – and make us encounter – strangeness … the more we look closely at texts, the more they show us that, unwittingly we do not even know what testimony is and that, in any case, it is not simply what we thought we knew it was’ (Felman and Laub, 1992: 7).
Frank shared the Harvard anthropologists’ commitment to the primacy of the body. He took over from the Kleinmans the idea of the body-self. If for Brody sickness posed a narrative crisis for the self, for Frank it posed a problem of action for the body-self. He described four options for the body-self in Western societies. The disciplined body-self subjects itself to regimens designed to renormalize it to the extent that it can be renormalized. The mirroring body-self cultivates idealization by others. Frank gives the example of Michael Jackson urging Elizabeth Taylor not to have chemotherapy because it would make her bald, which Jackson thought would be too unsettling for her fans. The dominating body-self seeks to control others. The communicative body-self realizes the ethical ideal of existing for the other. Frank follows Emanuel Levinas in stressing the face-to-face encounter with the suffering other as the precognitive core of all ethical witnessing (Levinas, 1988[1969]: 86).
Discussion
By the early 1990s, these four strands had begun to influence one another on an unprecedented scale. Narrative no longer featured solely as a hermeneutic framework for bringing the cultural margin of illness into view; it was increasingly also studied as a dynamic feature of daily intercourse that shaped the margin. The job of synthesizing these diverse strands of narrative thinking was accomplished to a large extent in Mattingly and Garro's special number of Social Science and Medicine in 1994. Those authors emphasized that narrative was not only a heuristic device but a powerfully concrete mode of thinking, similar in some respects to metaphor, capable of shaping our experience of rapid, as well as prolonged sequences of action.
Three fundamental characteristics unite the projects I have described in this article. First, in each case, narrative was a way of talking about (and reinstating) hidden contexts: as Arthur Kleinman put it in The Illness Narratives, he and his colleagues were interested in ‘a covert reality that is usually dealt with either indirectly or not at all’ (1988: 9). The focus was not narrative itself but that which could not be laid bare by ordinary observational methods; narrative was the chosen method for overcoming this impediment. Second, they all turned to narrative as a way of recording and ordering the workings of highly dynamic processes. What chiefly differentiated them from one another were their chosen analytic scales. Some used it as a macro-concept (covering an entire cultural situation), some as a meso-concept (to describe a common cognitive occurrence), and some as a micro-concept subtending the extremely rapid interactions – and some as all three. Narrative enabled all such theories to be laid out intensively and extensively, against a single yardstick. Third, all of the works I have considered used narrative as a means of documenting multiple coexistent realities. It is not surprising that as the 1990s progressed, they began to take more and more notice of one another.
Because of the pre-eminent influence of anthropology, the narrative-based medical humanities soon found themselves rehearsing one of the core debates in anthropology and sociology, but using narrative as outlined above to make their case. The debate in question was started by Emile Durkheim in 1893 in La division du travail. It has to do with how dependent social reality is on what Durkheim called ‘constitutive practices’. Constitutive practices are practices that constitute the knowing subject and the object(s) s/he thinks s/he knows (Durkheim, 1984). Take the case of a patient consulting a clinician, an example that frequently cropped up in medical humanities contexts. Is the patient's status as a patient and the clinician's as a clinician guaranteed by some pre-existing agreement, or is it put in place, microsecond by microsecond, by means of a negotiation – not all of which takes place in the interactants’ own awareness – and requiring the consent of each? Durkheim thought that at best there was only a fragile pre-existing consensus in such situations. The patient might decide the clinician was incompetent, or the clinician might conclude that the patient was shamming. The consultation was a situation from which order emerges. It is this dimension of emergentism that is so striking in the narrative-based medical humanities. Many sociologists and anthropologists who came after Durkheim (such as G. H. Mead, Kenneth Burke, Claude Lévi-Strauss, Erving Goffman, Harold Garfinkel, Jürgen Habermas, and of course Clifford Geertz) brought new conceptual tools to address this problem, but they were all at least implicitly committed to the idea of constitutive practices. It is surely no coincidence that all these writers feature prominently in the works of the founders of the narrative-based medical humanities. The core problem that the narrative-based medical humanities sought to address – which I suggest is also the source of their underlying unity, despite the many differences dividing the authors in question – was how to describe the constitutive practices to which illness gives rise.
John Guillory defines a discipline as an ‘institutional arrangement for producing and transmitting knowledge’. It offers a ‘rational procedure for what can be known about an object’. A discipline must have ‘primary institutional instruments, its curriculum, its pedagogy, and its program of research’ (Guillory, 2022: 379). Using those criteria, the narrative-based medical humanities raised the field to the level of a discipline, at least for a time. They underwrote a massive expansion of medical humanities programmes in US universities. Many of the books and articles I have discussed in this article played a central role in those programmes’ curriculum, pedagogy, and research.
The sheer quantity of new critical theory they introduced into the field is also noteworthy. The narrative-based medical humanities produced more major books than other period in the field's history. In their introductory chapter to the Edinburgh Companion to the Critical Medical Humanities (2016), Des Fitzgerald and Felicity Callard contrast what they call an integrationist vision of the medical humanities, characterized by attempts to develop a broad understanding of medicine, one that effectively complements biomedicine by placing the latter in a rich multidisciplinary context largely supplied by the humanities, with an entangled approach (a term they take from the philosopher of physics, Karen Barad), distinguished by the effort to bring to light ‘sets of as yet undetermined material-semiotic configurations and alignments (bodily, pathological, cultural, human, and so on)’ (Fitzgerald and Callard, 2016: 41). The integrationist vision, they write, tends to turn the humanities into a kind of ‘handmaiden’ to clinical practice, while the entangled one opens up ways into ‘animacies, vitalities and pathologies, which flow across different practices and preoccupations that then come to be ascribed to the “humanities” and the “biosciences”’ (ibid.: 45). None of the theories I have considered in this article were handmaidens. They were, moreover, ‘entangled’ in a new way. Prior to the narrative turn, the medical humanities were heavily reliant on critical theory (see my argument on Erikson's influence above; see also Vickers, ffrench, and Lefève in this issue). The narrative-based medical humanities effected a qualitative shift in the pattern and scale of the field's embeddedness in other disciplines’ endeavours. Not only were they learning new critical theory, they were learning it from disciplines that were themselves undergoing powerful reconfiguration through various extensions of the idea of narrative. By absorbing their influence, the field acquired an unprecedented opportunity to demonstrate its relevance to these disciplines, as a multidimensional vector of knowledge with strong roots in medicine, one of the most politically powerful disciplines in the modern university system. Without the bridges into other disciplines that were forged in this period, it is hard to imagine that the health humanities would command the wide interest they do today.
The narrative-based medical humanities also carried out the most far-reaching critique of biomedical reality to date. Never before had culture been brought into the domain of illness so powerfully. Many attempts to do so had been made in the 20th century but only in relation to mental health. The narrative-based medical humanities saw disease as an explanatory model in Kleinman's sense. Disease belongs to culture, which plays an essential part in constituting it as a human reality. Biomedicine had no principled way of taking account of psychosocial causation. Biopsychosocial medicine had to. This foundational achievement was enriched by Mattingly, Montgomery Hunter, and Mishler, who showed how culture shaped the reasoning processes of the actors involved in responding to illness to a much greater degree than was generally recognized. Bruner's approach to narrative – especially after 1989 – sees it as one of the chief means at our disposal to mitigate the (psychosocially driven) stress reaction. Bruner's position is in many respects similar to Kurt Goldstein's and Georges Canguilhem's vitalist view that starts off from the notion that living organisms necessarily have a ‘point of view’ on their environment. This point of view is one of the ways in which they construct their world. It has to be narrativized at some level if we are to identify with our own ‘agentive selves’. Second, the importance of this ‘point of view’ for health has been amply demonstrated in large-scale epidemiological studies such as the Adverse Childhood Events studies and the Social Determinants of Health studies (Commission on Social Determinants of Health, 2008; Felitti et al., 1998). These suggest that when our narrative purchase on the world is invalidated, when we cannot measure up to our culture's notions of what an agentive self should be, we become unwell. We lose some or all of the buffering that comes with social belonging. The bioethicists’ vision of narrative was predicated on the idea that major illness brought with it an opportunity to interpret not only the sufferer's predicament but the whole social world in which s/he was enmeshed. In a widely cited article, Will Viney and colleagues have argued that the medical humanities would flourish ‘when the procedural norms and routines of the humanities, the social sciences, and the biological sciences are openly, evenly and creatively interrogated and reworked’ (Viney, Callard, and Woods, 2015). The narrative-based medical humanities cleared that bar handsomely. They were the first wave to take proper account of psychosocial causation. It is striking that the contemporary field is relatively untouched by psychosocial perspectives in general and by epidemiology in particular.
Why have the narrative-based medical humanities faded from view? I want to suggest three reasons. First, the field is largely led by literature scholars. Despite the literary aura projected by the word narrative, literary scholarship was extremely marginal to the formation of the narrative-based medical humanities, drawing instead on assumptions and disciplinary histories that most literary scholars are unaware of. Second, it may be that the narrative-based medical humanities fell foul of an assumption that they were an ‘applied’ version of the narrative turn, the turn to narrative that so many in the social sciences and the humanities made in the 1980s. As I hope I have shown in this article, that assumption is at best half true. The most vital sources for the narrative-based medical humanities were older: the phenomenological tradition in anthropology, microsociology, and conversation analysis, and the cognitive psychology of early childhood. The narrative turn did eventually find its way into the field, in the mid 1990s, when it was put into dialogue with these older sources. But in most cases, these older sources remained more determining.
The third reason has to do with a change that has taken place in the humanities away from universalism towards the encounter with irreducible difference. This alteration can be seen in Kleinman's book Writing at the Margin (1994) and became mainstream in anthropology after the publication of Paul Farmer's work on ‘structural violence’ in Haiti (2004). I am not equal to the task of explaining why universalism has fallen into abeyance in the contemporary English-speaking university to the extent that it has – it is a massive topic – but its decline may be a further reason why the narrative-based medical humanities are so seldom celebrated today.
Of course, there are other reasons too. The Harvard group successfully refounded medical anthropology on Geertzian lines. Brody and Frank along with others reshaped the field of narrative ethics in health care. Mattingly has continued to develop phenomenologically oriented anthropology with a view to linking the political, ethical, and experiential (much as the Harvard group did, after Farmer). Garro has developed her paradigm of cultural learning. Bruner and Mishler were very far into their careers at the time they published their important books. They said what they had to say. Mishler's approach was enormously inspirational for Rita Charon's narrative medicine (2006), though she, like other clinicians interested in narrative, drew widely on the whole of the narrative-based medical humanities (see e.g. Greenhalgh and Hurwitz, 1998; Hurwitz, Greenhalgh, and Skultans, 2004; Launer, 2002).
Bruno Latour famously said in 2004 that critique itself had run out of steam (Latour, 2004). The unmasking upon which it was predicated had become mainstream, in no small part thanks to conspiracy theorists! What was needed was a focus on the real that was more hypothetical and tentative than the grand critical narratives of the mid to late 20th century. Latour urged young scholars to turn their attention instead to (contestable) ‘matters of concern’ rather than (putatively incontestable) ‘matters of fact’.
The narrative-based medical humanities gave the field a visible identity that was recognized by a wide variety of humanities disciplines. This in turn made possible the development of the medical humanities in humanities departments. I do not believe the field as we currently know it would exist without them. In the UK, clinical practice has been displaced from the centre of the medical humanities, for now at least. This is, ironically, a function of the success of the narrative-based medical humanities. It is not merely evidence of their supersession. The narrative-based medical humanities took the first meaningful steps towards integrating humanities and social science perspectives with biological knowledge. The field as it is currently constituted, with its reticence towards biology, would do well to learn from this.
