Dr. Alverson is professor in the Department of Anthropology, Dartmouth College, 6047 Silsby Hall, Hanover, NH 03755 (e-mail: firstname.lastname@example.org). Dr. Drake is Andrew Thompson Professor of Psychiatry and director of the Psychiatric Research Center, Dartmouth Medical School, Lebanon, New Hampshire. Dr. Carpenter-Song is a postdoctoral fellow at Harvard University Medical School, Boston. Mr. Chu is a student in the medical program in the Johns Hopkins University School of Medicine, Baltimore. Dr. Ritsema is assistant professor of anthropology, International Honors Program, University of Massachusetts, Boston. Ms. Smith is an independent ethnographer in Hartford, Connecticut.
The implementation of collaborative, client-centered care in the treatment of severe mental illnesses will necessitate a shift from a clinical focus on securing compliance with providers' unilateral prescriptions to one of building an alliance between client and care providers in which treatment and recovery are based upon a mutual exchange of questions, concerns, knowledge, values, and goals (1,2,3). Interventions built on alliance will require that providers elicit, listen to, come to understand, and work with what Kleinman (4,5) has called patients' "explanatory models" (discursive accounts) of the illness experience, for therein lie keys to genuine collaboration in health care planning and decision making—the person's most fundamental beliefs, values, and the sociocultural conditions of his or her world outside the clinic walls.
As Montori and colleagues (6) have made clear, "In chronic disease self-management, most of the decision implementation takes place in the patient's space under the patient's stewardship… . [So] it is crucial for the clinician to gain an understanding about what the patient is up against when implementing a given treatment strategy." The authors noted that this may be particularly important when clinician and client "share little in common—live in different towns, belong to different socio-economic groups, and come from different cultures." The client's lived experience of illness, the life-world context of suffering and healing, and the mental health system can be brought to the fore only by engendering dialog in which the person's illness discourse is encouraged (7).
From the perspective of a "phenomenology of illness" (8), clients' "illness accounts," as built up with the help of, and told to, friends, family, colleagues, case managers, or doctors, are not about their illness as it is actually experienced in real time. The accounts are under constant construction and unfold as experience accumulates or changes. Often, accounts will contain "stories" in which experiences are "emplotted," earlier parts adumbrating and leading to later ones by means of forces or agencies posited by the author. However personal they may seem to be, they are generally built upon culturally specific templates (9,10,11,12,13,14,15). Although not necessarily veridical, illness stories and discourse become very real in their consequences. Good (8) called this the "narrative shaping of illness"—the bestowing of "meaning" on events or circumstances accomplished through retrospective thought or narration, which supplies culturally scripted purposes, justifications, feelings, or motives.
If clinicians ignore clients' illness accounts, then help-seeking behavior (14,15,16,17), treatment adherence (18), satisfaction with care (19), and culturally informed clinical practice may be negatively affected (20,21). Inscribing treatment guidance into a client's own discourse of illness experience can be an effective way to motivate the client's engagement in treatment as an active agent and not simply as a recipient. The efficacy of building an alliance was highlighted in at least one large study (22,23) in which "patients' perceptions that their physician knows them as a person [that is, knows their experiences] were more highly correlated with patient adherence to medical treatment than any other dimension of care measured." Further, in clients' illness discourses, health care providers will likely discover empirically for each individual how such background characteristics as ethnicity, class, gender, community, family, and religion—the sum of which are sometimes blithely called "culture"—may affect and in turn be affected by chronic illness (24,25).
Flowing from these considerations, this ethnographic study pursued two aims: to discover in a heterogeneous sample of clients, diagnosed as having severe and persistent mental illnesses, the varying characteristics of their illness discourse and to determine whether the content and structure of this discourse covaried with important ethnocultural characteristics of study participants' backgrounds.
Over the course of 18 months, we worked intensively with 25 individuals, a subset of a larger sample of 204 ethnically diverse, unemployed, inner-city residents with severe and long-term mental disorders who were participating in a randomized clinical trial of three approaches to rehabilitation (26). In the larger study, 46% of participants were African Americans, 30% were Latinos (mostly Puerto Ricans), and 24% were "others" (almost all European Americans).
The 25 participants in the ethnography included nine European Americans, six men and three women with a mean age of 41 (range 23—55); six African Americans, three men and three women with a mean age of 37 (range 29—52); and ten Puerto Ricans, four men and six women with a mean age of 34 (range 22—47). These participants were recruited on a stratified random basis from each of the three ethnocultural groupings in the larger sample and from two of the three interventions—either a supported employment program or a psychosocial rehabilitation program (7).
In the study sample seven participants sometimes exhibited delusional or incoherent talk. The clients' major psychiatric diagnoses were subtypes of schizophrenia (21 clients) and major depression or bipolar disorder (four clients). Thirteen of the 25 participants had not held a competitive job in the past five years. Mean monthly individual income, mostly from federal, state, or municipal transfers, was about $700, ranging from about $550 to about $1,300. With the exception of one African-American college graduate and one European American who had completed some college, none of the participants had received education beyond high school.
This ethnographic study was approved by institutional review boards of Dartmouth College and the State of Connecticut. All 25 individuals gave written informed consent for their participation.
We collected extensive ethnographic data on how study participants described experiences of both mental illness and interventions. Field ethnographers shared in activities with participants for at least one year over the course of the study (August 1996 through March 1998) at many of their regular haunts, engaging in observation, conversation, and informal interviewing in the many real-world contexts and settings in which their lives—including mental illness—are situated and expressed. Details of the ethnographic methodology and some findings concerning vocational outcomes have previously been reported (7).
Expanded notes from interviews, conversations, and ethnographers' observations were entered as "free text only" Word files on Macintosh computers according to individual client and date. The text files were later set up singly and in folders organized by client so as to be read, indexed, and searched using Sonar: High Speed Text Retrieval Version 10 (27). Participants' discourse ranged over many topics, including the experience of long-term mental illness. For purposes of this study, we selected all recorded discourse with the following attributes as instances of "illness discourse": first-person accounts of life before illness, during the onset of illness, and after illness onset; claimed effects of illness and interventions; coping strategies; and interactions with the treatment system. Note that this "search algorithm" was not used to identify or organize the substantive themes of the discourse. The ethnically distinguishing themes were discovered only after all of the relevant discourse had been identified.
For each of the three ethnic groups, the themes that were most frequently, freely, and fulsomely attested in participants' discourse are presented. These are tendencies and generalizations only. Although significant interindividual variation was noted among all participants in regard to the themes presented here, there was also significant commonality within ethnic groups and significant contrasts between ethnic groups in the illness discourse and experiences of illness and of treatment.
Three examples of the illness discourses collected in this study have been posted as an online supplement to this article at ps.psychiatryonline. org. The transcripts serve as an extension of this summary report. There is one each from an African American, a Puerto Rican American, and a European American. The sample discourses illustrate how many of the generalizations and distinguishing characteristics summarized below are manifest in particular, but representative, cases.
African Americans' illness discourse
In this sample, African Americans' accounts of their illness tended to be vague as to origins, causes, time of onset, or inaugural scene. The illness was often referred to as "getting sick." Other accounts by African Americans described being picked on or bullied when young. Most accounts were neither long nor loquacious. Illness was described as a network of discrete troubles and vexations. Dealing with ordinary daily problems—housing, food, transportation, unemployment, financial desperation, disputes, and racial disparagement—loomed at least as large as did coping with mental illness. African Americans averred that racial discrimination and economic disadvantage were barriers to competitive employment at least as great as mental illness.
These participants were more likely than either the Puerto Rican or European American participants to see mental illness as stigmatizing and to want to keep it secret. They feared that others would condemn or reject them as friends or associates if mental illness were made known or discovered. In fact "privacy" (for example, "family business") seemed a pervasive aspect of folk ideology. Families of African-American participants tended to treat psychiatric symptoms as "character issues," such as being lazy, mean, wild, irresponsible, or sometimes possessed. Family interactions, positive and negative, were a significant aspect of African-American participants' illness discourse and a lived-in context of much illness experience.
Most African-American participants were not happy about taking medications. They often saw themselves as "guinea pigs," being experimented with or controlled through "all those damned pills." Most complained of negative physical side effects of medications. Many wanted more "talk therapy," which in their view should provide concrete practical advice for solving very specific problems. However, they seemed not to receive any talk therapy. They often saw themselves as victims of bad or ineffective treatment—being controlled by overbearing professionals, not being understood, or wanting to talk about very specific problems but getting only superficial responses. They resented the high turnover of providers. When therapists changed, participants complained that they had to start all over again getting comfortable "spilling their guts out" to strangers.
The ethnographic data showed that some standard clinical practices led to failures of alliance with African-American clients. Examples of such practices were encouraging clients to participate in therapy groups made up of strangers; inviting clients to disclose problems, to which providers responded minimally (for example, "Uh-huh," "OK," or "Is there anything else you want to talk about?"); focusing on the client without due regard for his or her family; and prescribing medications without sufficient explanation and choice. These tactics were in general not consonant with clients' core values and were seen as disrespectful and unhelpful.
Puerto Rican Americans' illness discourse
Puerto Rican participants described the onset of what they mostly called "nervios"—a concept that has meanings and semantic entailments quite different from those of the biomedical concept of "mental illness"—in rather specific terms as to time in life and frequently as to surrounding circumstances. These often involved actions of other people who were important in their lives. Someone left, died, had an affair, bullied, or neglected something. The sense of being victimized was minimal. There was little denial, and there seemed to be little apparent stigma associated with being afflicted with "nervios—or "problemas emocionales" or "debilidad mental." None of the Puerto Ricans referred to themselves using medical terms such as "mentally ill," "schizophrenic," "depressed," "in treatment," or "recovering." Puerto Ricans tended to see their behaviors or those of others as causes of feelings, which are engendered when one does something—such as violating the Holy Spirit—or which others do, such as "dissing" (disrespecting) someone.
Angry or sad feelings were generally seen as the result of external happenings, not internal forces. Agonistic or panicky acting out, although sometimes seen as a result of one's "problemas emocionales" was more often explained as a reasonable, justifiable response to perceived ridicule, dishonor, rudeness, disrespect, or neglect. Clinicians often interpreted such "ataques" as symptomatic confirmation of mental illness, a claim most Puerto Ricans did not endorse. Most of the Puerto Ricans tended not to see how their "problemas emocionales"—or the underlying cause of these problems, which was the misbehavior of others—could be helped by talk therapy. Most saw talk therapy as a ceremony one went through to get pharmaceuticals, which helped with specific problems such as poor concentration or insomnia, or to receive certification of disability status.
Puerto Ricans' illness discourse did not suffuse their life stories. These participants assimilated or incorporated the discourse of "illness" experience to a wider-ranging discourse about problems and troubles generally. (The European Americans, as noted below, tended to assimilate discourse about diverse difficult life experiences to a discourse about illness.) To the extent that their accounts were bundled with other issues, the accounts entailed putatively causal events, such as a rocky marriage, child abuse or neglect, "messing up," "gaming the system," and transgressing against spiritual, religious, or other "supra-human" agencies. These participants resented and rejected most "medicalese," such as psychiatric diagnoses. Illness symptoms and use of street drugs were seen as events or acts, not inherent dispositions. Most Puerto Ricans were quite at peace with pharmaceuticals and physicians' constant tinkering with them. In addition, most of these participants used or dabbled with "remedies botánicos," "tratamientos botánicos," and "remedios de las botánicas."
Puerto Rican participants supported or were supported by family, which in the Puerto Rican cultural context meant remaining incorporated within the normative patterns of behavior, expectations, and responsibilities of the family or household. The ill person was not extruded from circles of kin and friends because of symptoms of mental illness or many other deeds that would normally be seen as discrediting by the larger community, such as macho or panicky "self-assertion." Disability payments were generally distributed as contributions or "gifts" within the family sphere of exchange. Disclosing and family discussion of "nervios" seemed often to be done in a rather matter-of-fact manner.
The Puerto Rican participants were trying to live their lives, and their mental illness was just one part of it. The illness helped them financially through disability payments, and they willingly took medications to deal with symptoms. In the main they did not see their social and economic troubles stemming from their condition of "nervios," which was seen neither as an impediment to work nor as a justification for not trying to find it. None believed that their lives would be very different if they had not been afflicted with nervios. They spoke often of discrimination against, and mistreatment of, Hispanic-Latino Americans, which they viewed as creating far more difficulty for them than did emotional or mental problems.
The ethnographic data showed many unsuccessful attempts to create alliance with Puerto Rican clients. Failure to appreciate the family's influence and insistence on dealing with clients in terms of a biomedical model of illness and behavior were associated with failure to engage these clients as motivated collaborators in their health care.
European Americans' ("Anglo") illness discourse
European Americans' illness discourse contained many stories, which were often relatively long and complex and which tended to dominate life stories as a whole and to appropriate extensive biomedical jargon. Most participants showed considerable facility in labeling themselves in all aspects of diagnosis, treatment, and prognosis: "schizoaffective," "reached a plateau," "not medically stabilized," "in crisis," and "relapsing" were common terms. Anglos, unlike Puerto Ricans or African Americans, were often very interested in what their mental illness was called and how the diagnosis might be changed. Their descriptions of illness onset were generally couched in terms of some specific trauma or some vaguely genetic cause or predisposition or by reference to finding out as a result of being diagnosed. Mental illness was seen as permanent and disabling.
European Americans, much more than Puerto Ricans or African Americans, were ready to explain their moods—emotional ups and downs—as internal ("intrapsychic") problems. They tended to see wild or weighty emotions and feelings as a root cause of their problems, or they were the problem, which in part motivated them to seek out palliation in pharmaceuticals and talk therapy, which above all should make one "feel better."
Most Anglos in this study seemed comfortably intertwined with mental health professionals and institutions. Although they complained about mental health services and other services in a general way, among the three populations studied they were most comfortable with the biomedical model of their illness and the treatment approaches based on it. Their illness and treatments seemed to be a central part of their individual identities and daily routines.
Typically, the Anglos in this sample seemed to be more alone in life and with their illnesses than were those in the other two groups. They were far more likely to be cut off from significant social support and social capital—that is, morally and materially significant exchanges and reciprocities. With one exception, none of the European Americans claimed to have moral obligations to support others or to expect support from others. Anglos spoke often of being alone or lonely—something rare to hear from individuals in the other two groups.
When the physician or other provider asked Anglo clients, "How are you feeling?" or "Do you have anything you want to discuss?", they generally heard this as a sincere invitation to dramatic recounting. They seemed ready and willing to deliver—with minimal prompting—personal testimony about troubles and suffering and were generally content with clinicians' minimal empathic responses, as long as they also appeared to be listening, making reference to diagnosis, and discussing or adjusting medications.
European Americans generally took a pragmatic and individually variable attitude toward disclosing or discussing mental illness—to whom, under what circumstance, for what purpose. There was some self-stigmatization or fear that others would see mental illness as discrediting but not nearly as strong as that observed among African Americans.
Failures of alliance with European-American clients were also documented repeatedly in the ethnographic data. Not listening to and working with their illness stories and not challenging them to move toward recovery often led to continuing dissatisfaction, feeling miserable, and acquiescence to the belief that mental illness is disabling.
This study has brought to light several variations in how a small number of individuals from each of three self-identifying ethnic groups in one city experienced mental illness and the services provided by the mental health system. On the basis of the premises of the literature cited in the introduction and the specific information we have gathered here, we conclude that mental health service providers will develop or fail to develop a working therapeutic partnership with their clients to the extent that they become aware through context-sensitive, context-informed dialog how those individuals "en-story" their illnesses. Clinicians should be alert to, capable of eliciting, and competent to work with individually and ethnoculturally variable discursive information about illness experience, which clients provide by means of culturally specific tropes, schemata of thought and action, and the ways these are deployed in narrative performances.
We observed numerous client-provider interactions in which one party's intended meanings and messages were misunderstood or ignored by the other because of discourse failure arising from lack of awareness that the other might be coming from a very different perspective. One aspect of apprehending individuality in the clinic is to discover empirically through dialog how the client's ethnic, cultural, gender, economic, age, and other biographic circumstances are ineluctably constitutive of the illness experience.
Failure to recognize this reality could, as Kleinman and Bensen (25) have clearly noted, lead to simplistic cultural or ethnic stereotyping and to simple-minded notions of "cultural competence." On the other hand, appreciating a person's ethnocultural background and being alert to how it may inform individual clients' illness discourse might sensitize clinicians to the myriad qualitatively different ways that individuals understand and cope with mental illness and relate to the mental health system.
We note one important example of this point. The "minority" participants—African Americans and Puerto Ricans—seemed alike in manifesting an outward, social focus on life, which was filled with problems, among which the biomedically defined symptoms of severe mental illness were culturally positioned. Kleinman and colleagues (28) have called this pattern one of "social suffering." This pattern contrasted quite strikingly with the inward and individuated locus and focus of illness experiences (moods, impulses, and distressing beliefs) of the European-American participants, among whom much of the description of life's diverse experiences was inscribed within the story of being mentally ill.
Several caveats warrant attention. We studied a small number of individuals from one city who were served by a single large mental health center. The study participants therefore did not constitute a sample of any coherent larger natural population or even hypothetical aggregation. Therefore, generalization to larger numbers of other people is not warranted. Moreover, this mental health center served clients who constituted very different, and differently skewed, samplings of those with severe mental illness found in the three major ethnocultural populations within the Hartford metropolitan area: a large cross-section of the Puerto Rican population with severe mental illness; a large number of African Americans with severe mental illness who were economically not well off but who otherwise were heterogeneous; and a skewed sample of European Americans with severe mental illness, most economically very poor and several of whom sometimes showed severe symptoms. Ethnographic methods, being qualitative, are designed specifically to bring "qualities" of phenomena and issues to light, not to document incidence, frequency, or prevalence or to create a statistical picture of distributions and relationships.
Despite these limitations—indeed because of them—this ethnography illustrates some of the powerful ways in which ethnocultural background can influence how people understand their own mental illnesses and seek and respond to treatment. "Seeing the world through the client's eyes" is likely an essential step in building the trust, confidence, and partnership, which are—we aver—critical for effective mental health care.
The projects were supported by cooperative UD7-SM-51818 from the Center for Mental Health Services, Substance Abuse and Mental Health Services Administration, as part of the Employment Intervention Demonstration Program. Additional support was provided by grants MH-00842 and MH-56147 from the National Institute of Mental Health.
The authors report no competing interests.