ARTICLES

PAUL FARMER Department of Anthropology Harvard University

Sending Sickness: Sorcery, Politics, and Changing Concepts of AIDS in Rural Haiti

In this article I trace the emergence of a collective representation of AIDS in a village in rural Haiti. I initiated investigation of local understand- ings of AIDS years before the advent of the illness to the community itself and continued documenting the subsequent elaboration of a fairly de- tailed and widely shared cultural model of the new disorder. Through following serial interviews with the same persons over a period of six years, one can discern the rate at which consensus was achieved, the events which led to it, and the sign8cance of preexisting interpretive frameworks for current understandings of AIDS. This case contributes to the anthropological study of cultural meaning in formation and transfor- mation.

IDS presents new challenges to medical anthropology. Some are theoret- ical and not substantially different from the challenges faced by other eth- A nographers who seek to study, comprehend, and describe new phenom-

ena. Others involve the ethical dilemmas inherent in both the study of a terrible new affliction for which there is only limited therapeutic recourse and the deeply vexed question of how anthropologists might best contribute to the effort to pre- vent transmission of HIV. What follows is a processual ethnography of the advent of AIDS in Do Kay, a small village in Haiti’s central plateau. It is primarily a descriptive exercise, and the theoretical questions posed relate to the description of a new illness. Its chief goal is to call attention to the problems inherent in study- ing cultural meaning while it is taking shape.

The need for a more processual approach to the study of illness representa- tions is most dramatically illustrated when one is witness to the advent of a new disorder or one previously unknown to one’s host community. Some of the steps in this process of growing awareness are easily intuited. Before the anival of the new malady there exists no collective representation of the disorder; then comes a period of exposure, if not to the illness, then to rumor of it. With time and ex- perience, low interinformant agreement may give way to a cultural model shared by the majority of a community.’ What determines whether or not consensus is

6

CHANGING CONCEITS OF AIDS IN RURAL HAITI I

reached? In studies of illness representations, medical anthropologists have usu- ally asked, “To what degree is the model shared?” But when studying a truly novel disorder, a new set of questions pertains. How does cultural consensus emerge? How do illness representations, and the realities they organize and con- stitute, come into being? How are new representations related to existing struc- tures? How does the suffering of particular human beings contribute to collective understandings, and how much of individual experience is not captured in cultural meaning? My recent fieldwork in rural Haiti addresses these questions. Though primarily a study of a cultural model, the following account is also the story of three individuals with AIDS, for their experience is what made AIDS matter in Do Kay. This account is distilled from a series of interviews dating from 1983- 84 to the present. These reveal not just the role of culture in structuring illness narratives-we already know a great deal about that-but the ways in which those narratives are elaborated, how they change over time, the embeddedness of rep- resentations (also changing) in narratives, and their significance to the experience of illness.

The Changing Significance of AIDS

The Republic of Haiti’s role in the AIDS pandemic has been unique and unenviable. Like many other countries in the Caribbean, it has been gravely af- fected by HIV. Based on the number of AIDS cases per 100,OOO population, Haiti is among the world’s 20 most affected nations. Many more Haitians have been exposed to the virus. Although no large, random surveys have been conducted, a series of epidemiologic studies conducted between 1985 and 1987 indicate that fully 9% of 2152 “healthy urban adults” were seropositive for HIV (Pape and Johnson 1988). AIDS has recently been reported on Haitian radio to be the leading cause of death among Haitian adults between the ages of 20 and 49. While some contest this assessment, it is clear that HIV disease will mean great suffering in a nation that can ill afford yet another health burden.

For the inhabitants of Do Kay, the village in which most of the ethnographic material presented here was collected, the advent of a new and fatal disorder was, in the words of one person who lives there, “the last thing.” The last thing, that is, in a series of trials that have afflicted the rural poor of Haiti. The inhabitants of Do Kay, which stretches along an unpaved road cutting through the country’s central plateau, have had more than their share of trials. During the rainy season, the road from Port-au-Prince can take several hours to traverse, adding to the impression of isolation and insularity. The impression is misleading, however, as the village owes its existence to a project conceived in the Haitian capital and drafted in Washington, DC. Consisting in 1989 of fewer than 1 ,OOO persons, Do Kay is composed mainly of the families of peasant farmers displaced some 30 years ago by Haiti’s largest dam.

Before 1956, the village of Kay was situated in a deep and fertile valley, near the banks of the Rivi5re Artibonite. For generations, the villagers farmed the broad and gently sloping banks of the river, selling rice, bananas, millet, corn, and sugarcane in regional markets. Harvests were, by all reports, bountiful; life there is now recalled as idyllic. When the valley was flooded, the majority of villagers were forced up into the hills on either side of the new reservoir. Kay

8 MEDICAL ANTHROWLOGY QUARTERLY

became divided into “Do” (those who settled on the stony backs of the hills) and “Ba” (those who remained down near the new waterline). By all standard mea- sures, both parts of Kay are now very poor; its older inhabitants often blame their poverty on the massive buttress dam a few miles away, and note that it brought them neither electricity nor water. The sole improvements in their lives, they ob- serve, have been the construction of a school, a clinic, and other amenities built and managed by a Haitian priest who has been working in the area for over 30 years.

Early in 1987 the first case of AIDS was registered in Do Kay. Because in- vestigation of local understandings of AIDS had been initiated four years before this unfortunate occurrence it was possible to document the subsequent elabora- tion of a fairly detailed and widely shared cultural model of AIDS. By conducting serial interviews with the same people, it was possible to document the rate at which consensus was achieved and the events that led to it.* Another important event also occurred during the course of this study. In 1986 Haiti’s longstanding family dictatorship collapsed, which led to changes that were keenly felt in village Haiti. These changes also had a profound effect on the process of illness repre- sentation, for they altered substantially the ways in which illness and other kinds of misfortune were discussed. The following account attempts to illustrate the forces that were significant in defining a collective representation of AIDS and also to suggest how these forces were revealed to the ethnographer.

1983-84 “A City Sickness”

In 1983, when my research began, the word s i d ~ , ~ from the French acronym for syndrome d‘immunod&icience acquise, was often heard in Port-au-Prince. The term gained currency following the association of this syndrome with Haiti in the North American press. By early 1982 a number of Haitian immigrants had been seen in Florida and New York hospitals with infections characteristic of a new syndrome. Unlike other patients meeting diagnostic criteria for AIDS, the Haitians stated that they had not engaged in either homosexual activity or intra- venous drug use; most had never had a blood transfusion. The United States Cen- ters for Disease Control (CDC) inferred that Haitians as a group were in some way at risk for AIDS. The popular press drew upon readily available images of squalor, voodoo, and boatloads of “disease-ridden’ ’ or “economic” refugees, and painted Haitians as the principal cause of the American epidemic (see Centers for Disease Control 1982; Nachrnan and Dreyfuss 1986). As Dr. Robert Auguste of the Haitian Coalition on AIDS remarked in a Miami Times article in 1983, “In the annals of medicine, this categorization of a nationality as a ‘risk group’ is unique. ”

The effects on Haiti of this association with AIDS were quickly felt and far- reaching4 Throughout the 1970s, as international memories of “Papa Doc” Du- valier began to fade, tourism had begun to assume increasing importance in Hai- ti’s economy. By 1980 it had become the country’s second largest source of for- eign currency and generated employment for thousands living in and around Port- au-Prince. The effects of the AIDS scare were dramatic and prompt: the Haitian Bureau of Tourism estimated a decline from 75 ,OOO visitors in the winter of 198 1- 82 to under 10,OOO the following year. Six hotels folded, and as many more de-

CHANGING CONCEPTS OF AIDS IN RURAL HAITI 9

clared themselves on the edge of bankruptcy. Several hotel owners were rumored to be planning a lawsuit against the CDC. Haitian government officials reacted in a manner reflecting the deep contradictions of the Haitian ruling class. Within months one was hearing the classic mixture of antiracist nationalism, followed by local repression of those held responsible for “spreading AIDS. ” These measures did nothing to counter the collapse of the nation’s tourist industry. As Abbott has recently observed, “AIDS stamped Haiti’s international image as political repres- sion and intense poverty never had” (1988:255).

As thousands of urban Haitians were left without jobs, the word sidu took on specific connotations. Few city dwellers were unaware of the syndrome, though the majority of them could not have known individuals with AIDS. The word sidu was not yet well established, however, in the rural Haitian lexicon. In interviews conducted in early 1984, only one of 17 informants mentioned sida as a possible cause of diarrhea. The term did not occur in unprompted discourse about tuberculosis, the most common infection among Haitians with AIDS, nor did it figure in talk about diarrhea or other disorders. When questioned, 15 out of 20 villagers said that they had heard of sida, and a dozen of them associated cer- tain symptoms or stigmata with this label. But many of these attributes were not, in fact, commonly seen in Haitians with AIDS.

Most of the villagers who spoke of sidu noted that they had heard of the disorder on the radio or during trips to the ~ a p i t a l . ~ There was considerable dis- agreement as to what the chief characteristics of sidu might be. In the 1983-84 interviews, seven out of 20 mentioned three aspects of sidu: the novelty of the disorder, its relation to diarrhea, and its association with homosexuality. The ma- jority mentioned one or two of these attributes. Only five noted that sidu is lethal. Three thought that it was originally a disease of pigs; three were also of the opin- ion that despite the contrary claims of the foreign press, sidu had been brought to Haiti by North Americans. Two others asserted that “sidu is the same thing as tuberculosis.” In early 1984 Mme. Sylvain, a 36-year-old market woman, of- fered the following commentary, which resembles that of several of her covilla- gers .

Sida is a sickness they have in Port-au-Prince and in the United States. It gives you a diarrhea that starts very slowly but never stops until you’re completely dry. There’s no water left in your body. . . . Sida is a sickness that you see in men who sleep with other men.

She had little else to say about the syndrome, although Mme. Sylvain was seldom at a loss for words when sickness was the topic.6 These preliminary interviews demonstrated that in Do Kay, where illnesses were usually the topic of much dis- cussion, sida was not. When one villager was asked if he and his associates were reluctant to speak about sidu, he responded, “Why should that be? There is no one who says we can’t talk about sida. But it is nothing that we have seen here. It’s a city sickness (maludi luvil). * ’ In the first year of my research, all talk about the disorder was prompted by questioning; there were no illness stories or “ther- apeutic narratives” about sidu. For the people of Do Kay, already bent under the unremitting burdens of poverty and sickness, there was little at stake regarding AIDS.

10 MEDICAL ANTHROWLOGY QUARTERLY

Before 1985, then, one would have been hard pressed to delineate a collec- tive representation of AIDS in this part of rural Haiti. Despite several individuals’ elaborate explanatory models, despite the savvy of market women like Mme. Syl- vain, the lack of natural discourse about sidu and the low interinformant agree- ment on its core characteristics suggest that, during the 1983-84 period, no cul- tural model of AIDS existed in the area around Kay.

1985-86: Mklange Adult2re de Tout

During the course of 1985-86 relative silence concerning sidu gave way to discussion of the new illness in the Kay area, and a more widely held represen- tation slowly began to emerge. Illness stories were beginning to be recounted, but they were invariably the tales of someone else, somewhere else-people who had died in Mirebalais, the nearest large market town, or in Port-au-Pnnce. There was rumor, too, of mistreatment of Haitians in far-off North America, and one villager often spoke of a cousin in New York who had lost her job, “because they said she was a Haitian and an AIDS-carrier.’’

Fully 18 of 20 informants interviewed during this period referred directly to “blood” in our discussions of sidu, and for many other residents of Do Kay as well, sidu was a sickness of the blood. Perhaps the most commonly heard obser- vation was that sidu “dirties your blood” (li sul sun ou). There was frequent al- lusion to “poor blood,” usually a gloss for anemia, as a prodrome of side and some referred to the dangers of blood transfusion. For example, when in the course of an obstetrical intervention Ti Malou Joseph needed a unit of blood, sev- eral of her covillagers observed Chat, given the “sickness going around” (mufudi deyo a), a transfusion was tempting fate.7 For some, it was a question of exposing the transfusion recipient to a microbe (mikwob); for others, one of “mixing bloods that don’t go together,” causing reactions that eventually ‘‘degenerate into sidu. ” Several informants began to speak of sidu as a slow but irreversible process that was invariably fatal.

Others interviewed in the summer of 1985 stated that “bad blood” (move sun), a somatosocial disorder widespread among Haitian women, put one at risk for sidu. As Mme. Mathieu put it, “You’re very weak when you have move sun, and you can more easily catch sidu.” Although two of the 20 villagers inter- viewed in 1985 felt that the new illness was a “very severe form of move sun,” the rest of those who mentioned move sun underlined distinctions between it and sidu. The observations of Mme. Kado, a 51-year-old woman who worked with the priest who had founded the school in Kay, were typical of the opinions gar- nered in late 1985.

[Sida] spoils your blood, makes you have so little blood that you become pale and dry. It first causes little blemishes (bouron) that rise all over your arms and legs. That tells you that the blood is bad, and makes you think of a simple case of move sun. But sida has no treatment, it’s not like move san. Anyone can get this, but it is most common in the city.

In much of Haiti, disvalued experiences-shocks, disappointments, anger, fright-may be embodied as disorders of the blood. The significance of this con- ceptual framework led Weidman and her coworkers to speak of the “blood par- adigm” underlying the health-related beliefs of their Haitian informants in Miami

CHANGING CONCEP~S OF AIDS IN RURAL HAITI 11

(Weidman 1978; see also Farmer 1988; Laguerre 1987). It is within this paradigm that are found the causal links between the social field and alterations in the qual- ity, consistency, and nature of blood. During much of the 1985-86 period, preex- isting beliefs about blood lent form to vague understandings of sidu, which was coming to represent an irreversible pollution caused, depending on whom you asked, by blood transfusions, same-sex relations, weakness from overwork in the city, or travel to the United States. As will be clear, however, the contributions of this paradigm to the emerging representation waned with direct experience of the disorder, and the “tuberculosis paradigm” emerged as the more important of preexisting models.

The year 1985 also marked the debut of a preventive campaign conducted by the nation’s health authorities. There were songs about sida and numerous radio programs, all in Creole and targeted toward the peasantry. Less important were the many articles in the print media and the posters and billboards declaring sidu to be a public menace to which all were vulnerable. Although villagers may have known more about the syndrome as a result of these public health efforts, it was not yet a compelling subject of everyday discourse, which was increasingly, if somewhat clandestinely, dedicated to discussion of national level political events. The Duvalier dictatorship, in place for almost 30 years, was beginning to totter, and more and more rural Haitians joined the chorus calling for Duvalier’s re- moval.

After years of silence the people of Do Kay joined in this chorus. Because peasants had long been excluded from direct participation in politics, the shift was a significant one and had an impact on the way that illness was discussed in rural Haiti. At first, talk of sidu was simply submerged in all-important discussions of national politics. When the syndrome was addressed, it seemed that it was often invoked to malign the regime or the United States. On New Year’s Day 1986, several of my friends from Mirebalais joked that Duvalier was a musisi (homo- sexual) who had contracted the syndrome from one of his msisi ministers. Shortly after Duvalier’s departure one market woman in her mid-fifties angrily denounced AIDS as part of “the American plan to enslave Haiti. . . . The United States has a traffic in Haitian blood. Duvalier used to sell them our blood for trans- fusions and experiments. One of these experiments was to make a new sick- ness.

Later it became clear that the fall of the Duvalier dictatorship gave a boost to stories about sidu. To judge from trends observed in Kay and surrounding vil- lages, rural Haitians began to feel that they could speak more candidly about mis- fortunes in general, and this alteration in the “rhetoric of complaint” may have had a determinant effect on what would prove to be enduring understandings of ~ i d a . ~

One of the first slogans to become popular shortly after Duvalier’s fall was bubouket la tonbe. A literal English equivalent would be “the bridle has fallen off,” but the phrase would be better rendered as “the muzzle is off.” Although few began openly talking about politics in Kay until March, and a full year had elapsed before the adventurous were wholeheartedly joined by a majority of the villagers, the transformation seemed complete by the spring of 1987. In Kay and surrounding villages there was either a sudden proliferation of transistor radios or a surfacing of them. Some persons, men especially, spent entire days cradling

12 MEDICAL ANTHROPOLOGY QUARTERLY

their radios, switching from one news program to another. Community councils, drastically overhauled in other villages, were strengthened in the area around Kay; meetings that once drew a score or so, after the summer of 1986 often drew well over a hundred people. New groups were formed and set to civic activities, such as repairing roads and planting trees. All this was worked into the daily round of gardening and marketing, but the changes stood out nonetheless.

The subject of sida, however, was only temporarily submerged. In Port-au- Prince, many knew people who had died or were ill with the syndrome. Hospitals and sanatoria were faced with large numbers of mom sida, as persons with AIDS were labeled. Haitian researchers continued to document a large and growing ep- idemic. Government health officials conceded that sida was not a public relations issue but rather a major public health problem. In the Kay area, too, sida was once again a regular topic of conversation. In the summer of 1986 questions I posed about the sickness triggered long and elaborate responses. Yet respondents expressed many discrepant ideas.

In natural discourse about sida, the number of references to blood declined. In interviews conducted late in 1986, only 11 of 19 informants used the term when speaking at length about the new sickness. Public health campaigns may have contributed to this shift. The more one heard about it on the radio, the less it seemed to resemble other well known disorders of the blood. The declining sig- nificance of the blood paradigm is suggested by a comment from a 1986 interview with Tonton Sanon, an herbalist. “I’m wondering if it is really a sickness of the blood, because we know how to put blood in its place. There’s a part of it that is in the blood, yes, but it is not only in the blood, and it’s not blood that is the principal problem. The problem is in other systems.”

He was seconded by others who spoke as if the blood paradigm had been used to assess the nature of sidu and found wanting. Interviews with other healers revealed a similar lack of accord about the new illness, although many allowed that sida was beyond their competence. “Truly it’s a sickness that is slippery (enpwenab), ” observed Mme. Victor, a midwife known for her efficacious herbal remedies. “To this day, they’re struggling with it, but they haven’t yet found an herbal treatment for it. ” A doktefey (herbalist) predicted that “the herbal remedy that will heal sidu has not yet reached us, but when it does, we’ll learn how to use it. 9 9 10

In summary, it seemed that during 1985 and 1986, when mention of sidu began to stimulate more interest, there was an effort to compare the disorder to other illnesses, especially those involving the blood. But sidu failed to fit neatly into the existing blood paradigm. Lack of a perfect fit between the new disorder and the old framework posed no real problems, as clear and defensible under- standings of sidu were not yet a necessity: no one from Do Kay had fallen ill with the syndrome.

1987: Prototypes and Protomodels

In many ways 1987 was the decisive year in the process leading to a shared understanding of AIDS. During the course of that year a protomodel of illness causation rose to prominence, a model that proved influential in the elaboration of a more stable collective representation of sidu. By the fall of that year, narra-

CHANGING CONCEPrS OF AIDS IN RURAL HAITI 13

tives about sidu were easily triggered, and it was clear that a consensus, albeit tenuous, had emerged. Interviews conducted in 1987 and afterward revealed that the semantic network in which sidu was embedded had changed substantially since 1983-84. In 1987 the syndrome was mentioned by over half of those asked to cite possible causes of diarrhea in an adult. The majority also associated sidu with tuberculosis. Furthermore, ideas about how the new disorder became man- ifest in the afflicted were more widely shared. Equally striking was the increasing frequency with which the social and political origins of illness, including sidu, were mentioned. There are perhaps two primary reasons for this: first, the un- muzzling of the rural poor led to a new rhetoric of complaint; and second and most important, the syndrome had come to matter locally. Someone in Do Kay had fallen ill with sidu.

Comparing early interviews to more recent ones revealed the increasing im- portance of the shift in styles of complaining which was triggered by the large scale political changes sketched above. Although interviewing style and methods were not altered, the narratives, whether relating a case of diarrhea or some other misfortune, became increasingly tinged with a new political sensibility. Yet “pol- iticization of discourse” is an altogether unsatisfactory description of a far more complicated process. The stories told were superficially similar to those heard earlier, but how tellers gave shape and sense to their stories had changed. For example, in speaking of misfortune, informants’ attributions of blame seemed to be changing subtly. Narrative shifts similar to those in the following interview with Mme. Jolibois abound. Mme. Jolibois, a young woman who supports her family by working a small patch of land, had traveled from the Kay area to a clinic in a nearby town in February 1984. Her infant son had had a bad case of diarrhea. When asked what had caused the diarrhea, she answered in 1984, “I don’t know what causes it. Microbes, perhaps, or gas from milk. Microbes, especially- they’re little bugs that can make children sick. Or it could be my milk. I think he must be getting too old for milk.”

In May 1987, over three years after the first interview, she again went to the clinic, a new one in Kay. This time a nine-month-old daughter had severe diar- rhea. When asked the same question, “What caused the diarrhea?” she res- ponded, “It’s the bad water we have in [my village]. We have to drink it even when it’s muddy and full of microbes. It gives the babies diarrhea, and they die, and the government does nothing about it. It’s always promises without action (promet sun buy). ”

The methodologically minded reader might ask a series of important ques- tions. Were the differences related to the severity of the episode? The sex of the child? Are contextual or performative factors important? Did the ethnographer have closer rapport with the informant years later? Perhaps Mme. Jolibois was simply in a bad or accusatory mood? Such questions were slowly revealed to be secondary, however, as I began to note similar trends in the discourse of other villagers.

The collapse of the Duvalier regime also had a palpable effect on the way in which AIDS-related accusation was marshaled and used. Conspiracy theories abounded: the Duvalier regime had caused sidu, asserted some. Others thought that no, the Duvaliers were too stupid to create a sickness, despite a talent for creating zombies. But they had allowed their nation to be used as guinea pigs in

14 MEDICAL ANTHROPOLOGY QUARTERLY

an American plan to stem migration. Referring to the North American suggestion that AIDS originated in Haiti, more than one villager was heard to remark, “Of course they say it’s from Haiti: whites say all bad diseases are from Haiti.”” Indeed, accusations against the accusers were perhaps the most prevalent of these commentaries.

The illness of Manno Surpris was the second reason that the same villagers who were aware of but generally uninterested in sidu in 1984 might be universally interested in the syndrome less than three years later. In 1987 sidu came to be a social drama that left few adults in Do Kay untouched.’* The impact of this change is suggested by the observations of a young schoolteacher, himself a na- tive of the village in which we worked. He was interviewed several times between 1983 and the present. In a 1984 interview he noted, “Yes, of course I’ve heard of [sida]. It’s caused by living in the city. It gives you diarrhea and can kill you. . . . We’ve never had any sidu here. It’s a city sickness.” A long exchange recorded late in 1987 clearly revealed that the man’s understanding of sida had changed substantially. He could now hold forth at great length about the disorder. A chief factor seems to have been that he was now able to refer to the death from sida of Manno Surpris, his fellow schoolteacher. “It was sidu that killed him: that’s what I’m trying to tell you. But they say it was a death sent to him. They sent a sida death to him . . . sidu is caused by a tiny microbe. But not just anybody will catch the microbe that can cause sidu.” Manno’s illness and death made a lasting contribution to the cultural model of sidu that took shape in recent years, and this contribution was not substantially lessened by the subsequent deaths from AIDS of two other villagers.

Manno moved to Do Kay in 1982, when he became a teacher at a large new school established there by a Haitian priest. He was then 25 years old. An enthu- siastic and hardworking man, Manno came to be held in high esteem by the school administrators. He was entrusted with a number of public-and remunerative- tasks, including taking care of the village’s new water pump and the community pig project, both of which were administered by the priest who ran the school. That an outsider would be granted such favors was deeply resented by some of the villagers, as became clear after Manno fell ill.

Beginning in early 1986, he had been bothered by intermittent diarrhea. Su- perficial skin infections recrudesced throughout the summ