|4 A Critique of Objectivity in International Health|
|图书名称：Writing at the Margin: Discourse between Anthropology and Medicine|
图书作者：Arthur Kleinman ISBN：
出版社：Berkeley: University of California Press 出版日期：1995年
This material was originally prepared as a response to two chapters written for a book on social science contributions to international health of which I am an editor (Chen, Kleinman, and Ware 1994). In that volume, a section was organized to review conceptual and methodological problems in the measurement of mortality and morbidity in low-income societies. The chapters included a review by Christopher Murray and Lincoln Chen, two international public health researchers, which essayed the chief difficulties in the measurement of morbidity and their implications for health research and policy, titled "Understanding Morbidity Change" (chap. 4), and a more theoretical piece by Amartya Sen, an economist and philosopher who has written extensively on famine and female mortality in Africa and Asia, titled "Objectivity and Position: Assessment of Health and Well-being" (chap. 5). My commentary tried to balance the section with a contrasting cultural critique of "objectivity": a core concern for these authors. I have further reworked the material in order to include several related themes with a different audience in mind.
Economists and public health experts (Murray, Chen, and Sen included) have traditionally worked with mortality rates. These they have aggregated for countries and regions in order to correlate them with large-scale economic and social changes that also are measured by aggregate statistics. There is increasing awareness even in these disciplines, however, that mortality rates are social fabrications that are based upon often seriously inadequate sets of data of questionable accuracy. They serve bureaucratic purposes of the state and the international community, certain of which come to distort their significance and uses. In spite of this by now well-recognized problem, economists and public health experts are not deterred from working up aggregated quantitative data in elaborate statistical programs that create the illusion, at least, that certain parameters of the knowledge base are well supported and their policy implications clear. (Sometimes they are, though more frequently they are not.)
Chen, Murray, and Sen are sensitive to this problem; indeed each has made important contributions to the understanding of how serious a problem it is and what can be done to control it. Moreover, they recognize that as suspect as mortality statistics are, they do not raise the degree of uncertainty that data on morbidity—the experience of sickness and its disabling effects—do. These authors, though quantitative researchers, wish to see measures of morbidity (and of well-being) added to international health statistics so that those statistics more accurately reflect the serious problems of chronic illnesses in aging populations which produce substantial distress and disability while not usually taking the life of the sufferer until many years later, if at all. That information, they argue persuasively, is needed to more effectively plan for health care and preventive programs. It is for this reason that the question of measurement looms especially large for them. Yet, they have even greater difficulty with anthropological sources of information, because they are largely qualitative and local, recalcitrant to aggregation at national or global levels.
All three authors felt that my commentary painted them into positions that were overdrawn and not sensitive enough to the subtleties of their arguments. Hence I should warn readers that they must read the chapters by Murray and Chen and Sen before they assume that what I have written is fair to these scholars. Nonetheless, I feel the general points I discuss hold for the field of international health and development in the main, and for that reason I have extended the original argument in a final section added expressly for this collection in order to touch upon other points that deserve to be brought into the compass of the critique.
The reader of Health and Social Change might rightly ask why a volume of social science and international health studies should include several chapters on theoretical questions concerning the measurement of health status. The reason, as the chapters by Murray and Chen and Sen should make clear, is that the measurement of morbidity necessitates assessment of subjective complaints and recognition of the situated knowledge of health indexes. Subjectivity and the positioning of knowledge intensify the challenge to a positivist approach to measurement which is already strained nearly to its limits by problems in the measurement of mortality. While both chapters indicate the need and even necessity for taking into account patient reports and practitioner interpretations, they come down on the side of the requirement for some form of objective measures that can be used to make more reliable intrasocietal and intersocietal comparisons. Those measurements in turn can be used to test claims about changes in health and their relationship to social change. Ultimately, these authors argue, there must be a determination of the legitimacy of claims about ill health so that resources can be allocated fairly and justly to those who are in the greatest need.
This is an understandable and salient goal. It is a goal, moreover, whose theoretical grounds and entailments are all too often taken for granted by epidemiologists, public health experts, and many social scientists. One group of social scientists among whom this aim is not taken for granted is anthropologists. Indeed, critiques of the way positivism and objectivism function in the health sciences are central to the intellectual program of medical anthropology. And it is from the vantage point of this intellectual program that I join the debate that the chapters by Murray and Chen and Sen seek to advance. Because Sen offers a philosophically elaborated discussion that seems both generally compelling and particularly challenging to a culturalist analysis of "health transitions," I offer my comments on the objectivist position as Sen so carefully articulates it. Nevertheless, these comments can be extended to the position set out by Murray and Chen, a position that I think accurately represents the dominant orientation in international health and social development studies. (See also Dasgupta 1993.) The critique I offer does not attempt to represent all relevant culturalist responses to this challenge, but simply those that emerge from my own work over two and a half decades as a medical anthropologist who, in the course of doing cross-cultural research on illness and care that is mostly qualitative, even when it mixes stories and statistics, routinely comes up against the demand by public health experts for objectivity.
Amartya Sen's "Objectivity and Position": Assessment of Health and Well-being
Sen (1994) argues that "objectivity requires explicit acceptance and extensive use of variability of observations with the position of the observer." For Sen such "positionality" is an intrinsic feature of objectivity that needs to be distinguished from subjectivity. Sen distinguishes between "positional objectivity," a statement about the objectivity of observations from a certain position, and "transpositional objectivity," which builds on the former but goes beyond it. Subjectivity, in contrast, pertains to the inward-looking nature of subjective judgments and to interpersonal variance.
Observational claims, in Sen's way of putting it, are a specific form of objectivity that is prior to the objectivity of generalized presumptive claims. Although observation and conceptualization influence each other, and observations are dependent on a person's positionality, there is also "some invariance vis-à-vis the subject" (p. 116). Drawing on Thomas Nagel's The View from Nowhere (1986), Sen reasons that some observations or claims based in observations are more objective than others because they are more accessible to a range of subjective capacities, and because some positioned observations have priority over others. To move from "positional" to "transpositional" objectivity "calls for rules of assessment—of 'priority'—that have to go beyond what is needed in examining the demands of objectivity of observations from an amply specified position " (p. 118). Transpositionality can be built up out of congruence of "all the relevant positional views," or out of a particularly decisive positional view (p. 120). Moreover, even if there is a fair amount of subjective influence, a positional observation can be specified. Indeed, such specifiability may help to control subjective influence. But Sen holds that much of what goes by the name of subjectivity in the domain of health is really a matter of positions that have not been specified. Specify the economic, institutional, professional, and ideological positions of practitioner and patient and you will so constrain the clinical construction of reality that there will be positional objectivity even if there is "non-objectivity in a transpositional sense" (p. 123).
Morbidity can be assessed from the positional objectivity of the patient's point of view. But this is a slippery slope because perception may be variable due to interpersonal and intertemporal differences, and owing to systematic social influences. Once we start specifying the positionality of illness perception, where do we stop? Because Sen is concerned with morbidity in settings of deep deprivation where there are few or no formal health services, as in the poorest regions of India, he is especially concerned that systematic experience in the use of health services in more resource-rich settings may both reduce morbidity and increase self-perception/reporting of symptoms. That would render comparisons across regions which are based on questionnaires invalid. Hence the need to view self-perception of health status as positional. It is objective only where its positional qualifiers are specified, and it is of less priority than other positional observations (perhaps the doctor's) which can be made more accessible and specifiable.
Sen is most convincing when he suggests how gender inequality in India creates positional observations that systematically underreport illness because of deprivation in education and also "acceptance of greater discomfort and illness as part of the prevailing mode of living" (p. 125). Those local observations, though capable of being constrained as a kind of positional objectivity, "would not readily translate into transpositional objectivity of women's relative deprivation, nor into positional objectivity from the general position of being an Indian rural woman" (p. 126). Yet, they might well, Sen qualifies, have local "explanatory and predictive relevance."
This brief summary cannot do justice to the precision, balance, and subtlety of Sen's argument. There is much in that argument that I find persuasive and can agree with. But I wish to draw out a tension between Sen's cautious commitment to objectivity, which for all its complexity and qualifications is still a positivist orientation, and the orientation of many social anthropologists who conduct ethnography.
The Primacy of Categories to Observations: Reliability versus Validity
Sen emphasizes the observational basis of verification, though he qualifies the classical example of the mirror held up to nature by addressing the relativity of perspective. All perspectives are positioned. What we observe depends upon where we are. Therefore, we need to control our observations to assure their accuracy. In research parlance, we refer to the verification of observations as reliability. A finding is considered reliable when it is verified with a high degree of confirmation. For example, different raters may observe whether or not a patient demonstrates weakness in the performance of a task, or exhibits unsteady gait when asked to walk in a straight line, or reports feeling sad. Their degree of agreement after observing the same patient sample is frequently quantified as a correlation coefficient of the reliability of their assessment of a particular feature of morbidity.
But observations, as Sen also notes, involve conceptual assumptions about what is being observed. If you are going to measure ischemic heart disease, you need to begin with a concept of the condition and its operationalized criteria (chest pain, narrowed coronary arteries, stress-induced, or other causes of, insufficiency in oxygenation of heart muscle). The concept—be it a professional or lay category—precedes and guides the observation. Thus, both symptoms (complaints) and signs (observable indications) of disease are in fact interpretations . Conceptual categories shape our interpretation of observed states, be those observations subjective or objective. (And these are categories and interpretations from which one can die: biological processes may be understood through cultural construction and shaped by those constructs, but they and their prepotent effects are real nonetheless.) Put differently, validity —which is the verification of the conceptual categories that organize our observations—is both different from and a more fundamental aspect of verification (i.e., measurement) than reliability .
Elsewhere I have given an example from psychiatric categories that applies equally to any other conceptualization of health (Kleinman 1988b:10–14). Suppose ten trained observers are asked to interview the same ten adult American Indian informants who have experienced the death of a spouse, parent, or child in the previous month. The interviewers will determine with a high degree of agreement that the same informants report the experience of hearing and/or seeing the recently deceased family member beckoning them to the afterworld. This experience is very common among adult American Indians. Yet it is distinctly uncommon among other adult North Americans. Now if the observers agree nine out of ten times that the same eight American Indian informants have had this experience, they will have achieved a high degree of reliability indicating an observation about which we can be highly confident. But if they describe this observation as a "hallucination," that is, a pathological percept indicative of mental illness, which it might be for adult non-Indian Americans, they would have an empirical finding that was reliable but not valid. Validity would require the qualifying interpretation that this percept is not pathological—in fact it is both normative and normal for American Indians since it is neither culturally inappropriate nor a predictor or sign of disease—and therefore it could not be labeled as a hallucination. Rather it would require some other categorization, for example, a pseudo-hallucination or a culturally appropriate and personally normal percept, in order for the observation to be both reliable and valid.
When measuring morbidity of any sort, reliability is necessary but insufficient to verify the condition and its effects. Take chronic fatigue syndrome as an example. Are we talking about the same thing when we observe the fatigue noted by an otherwise healthy long-distance runner with a postviral syndrome; a middle-class single father in a North American suburb with chronic fatigue syndrome associated with grieving for a dead wife; a desperately poor mother of six malnourished children living in a hovel in a favela in northeastern Brazil who has experienced chronic starvation; or an elderly man who has struggled with the disabling lifetime effects of childhood polio, to which his body is now giving in? If assessments lack validity, then morbidity has not been adequately measured. The problem with findings that have reliability in the absence of validity is by no means limited to mental, culture-bound, or new conditions; it is a serious, generic problem in cross-cultural assessment of patient complaints as symptoms and signs of illness.
Hilary Putnam (1993:151–152), a distinguished philosopher, drawing on a decade of publications, puts the issue in more general terms that go much deeper and further into the limits of objectivity.
Context, Interpretation, and the Experience of Suffering
For anthropologists conducting studies in international health, the focus of research, unlike that of epidemiologists and economists, is on individuals and small groups in local worlds: a village, a group of villages, an urban neighborhood, a social network. (For public health researchers, it may seem like the study of a single, "uncontrolled" case, even though the case is a community rather than an individual.) The research (ethnography with or without quantitative surveys is what I have in mind) usually involves relatively long periods—often months, sometimes a year or more—living with informants, engaging in participant observation of daily activities, and interviewing and reinterviewing a relatively small group of key informants. Over time the ethnographer gets to know the local setting well, establishes relationships of trust, and comes to understand the long-term tensions, supports, and negotiations that underwrite daily happenings. Such research seeks to describe informants' indigenous categories of the body, the self, health, illness, and healing, but not as abstract responses to questions cut off from social life. Rather the ethnographer interprets the effects of these meanings in the sociopolitical context on the way the group experiences suffering and responds to illness. Or the ethnographer focuses on some other daily life engagements in which these categories are part of social action. Conversely, the ethnographer also describes the way that illness alters social relations, disrupts patterns of interaction with key institutions, and transforms personal behavior; again, within the exigencies and uncertainties of real social contexts. (See chapter 9 below for examples.)
Those descriptions are understood, in current cultural theory (Bourdieu 1990:10–27; Kleinman 1988b:5–17; Kleinman 1992; Shweder 1991:27–72), as positioned interpretations of positioned interpretations. In other words, the ethnographer establishes his own position as a usually nonindigenous field-worker of a particular age, gender, and cultural background and with particular social commitments, who possesses a certain conceptual orientation and focused problem framework from which he or she encounters members of the community who are themselves positioned (on account of gender, class, caste, etc.) participants in local worlds. The empirical result of this utterly human—because contextualized and uncertain—though professionally disciplined engagement is positioned knowledge; that is, a view from somewhere. Context and interpersonal dialogue are understood to shape the knowledge so that it is always particular to a local world. The ethnographer, if honest, is also compelled by the nature of the experience to be aware of the limits of her or his grasp of uncertain and changing human conditions.
From the perspective advanced by Sen, as I have qualified it, the ethnographer's account carries a positioned validity. Often its reliability is untested. But even its validity is not regarded by the ethnographer as "objective," because it is never a transpositional "view from nowhere" (Nagel 1986), nor can it be verified straightforwardly as a positional observation or series of observations. That is to say, it can be neither fully specified (because it is always more than observation) nor made broadly accessible. Rather the ethnographer constructs a professional category based on the interpretation of local popular or expert categories, which are themselves held by native actors engaged in parochial actions vis-à-vis others. The ethnographer interprets a field of interpersonal experience as he or she narratizes (forms a story of) the felt flow of that interpersonal world, a world that is encountered as stories told to the ethnographer (a positioned witness), stories of illness, disability, and therapy. That interpretation is a creation as much as an observation; it is a constrained yet storied observation that is made for a professional discourse and into a genre.
The knowledge the ethnographer produces is never impersonal; it represents not only the public, focused account of informants but also the subsidiary, tacit knowledge that is part of their (and the ethnographer's) practical life activities (Polanyi and Prosch 1975:22–45), what Pierre Bourdieu (1990:66–69) calls a feel for the game, or sense of the field. Thus, the anthropologist's ethnography cannot be, in any strict sense, objective, even when it concerns social "objects" (people, institutions, exchanges, power), because it is embedded in the subtleties and complexities of subjective and interpersonal understandings. The ethnography is constructed from those particular interpersonal engagements in such a way that two ethnographers working in the same community would (and, in fact do) construct distinctive ethnographies.
Anthropologists strive for interpositional knowledge of a local world by comparing understandings of positioned participants. They also struggle to place their findings in a transpositional professional discourse of academic writings, one that is external both to the local world of their informants and to their own institutional and cultural settings. What anthropologists seek to achieve, however, is not objectivity of observation, but a multisided interpretation. Only because it includes multiple, situated interpretations can it be considered "transpositional." Translation of findings into what is more accurately called an interpositional discourse, furthermore, should be a final stage in anthropologists' work, not a precommitment or initial move. Hence, even though their goal sounds much like Sen's, their way of going about achieving it seems rather different, and that difference in practice makes for a difference in outcome. The ethnography reads more like a work of social history or biography than a publication in economics or the health sciences.
Anthropologists in international health credit the knowledge of local informants concerning everyday health practices, illness episodes, and therapeutic actions. Are they mistaken in doing so? Both Murray and Chen (1994) and Sen (1994) offer reasons for regarding this practice as epistemologically unsound. After all, patients do not have privileged access to objective "knowledge" about their underlying conditions that can be measurably verified with a high degree of probability. The patient's "subjective knowledge" is less reliable than the medical researcher's "objective knowledge." Doubt is cast on the informant's (and the ethnographer's) constructions.
Two comments need to be made here. First, the researcher's knowledge—that of the epidemiologist, the economist, the public health physician as much as the ethnographer and clinician—is a construction not so very different in kind, though often quite different in degree of systemization and control, from that of the informant. (I believe Sen's analysis also supports this point.) The histologist interpreting a surgical pathology slide or autopsy specimen of tissue presumed to be cancerous constructs the object of his or her inquiry (see Canguilhem 1989). The histologist works with a historically derived, changing set of categories that defines malignancy, that distinguishes, for example, premalignant or borderline cases, that stages invasiveness and spread in different systems of classification, and that relies on his or her focal and tacit knowledge to diagnose. Anyone who has ever looked under a high resolution microscope—say at cells, certain of which are malignant melanoma cells and others of which are dysplastic ones—knows that one has to be taught the categorization with its operationalized diagnostic criteria to distinguish the two, and even then it takes "experience" to "feel" confident one is making an objective assessment. The same can be said of "reading" a chest X-ray or interpreting a CT scan of the skull, listening to abnormal heart sounds, examining a rash, and conducting an epidemiological survey based on such interpretations. Categories construct experience, and in turn socially constructed experience validates the working knowledge and practices through which those categories are interpreted as "objective." Peter Berger and Thomas Luckmann (1967), in their classic statement of the social construction of reality, regard the internalization of cultural categories, through socialization, into personal experience as involving incorporation followed by objectivization. Thus, in their view, objectivity is, at least in part, a projection of internalized but socially derived categories onto the external world, where they are then "felt" to be "objective" structures of reality, because personal experience, through a grand tautology, does indeed match public definition.
Second, informants at times actually can be shown to possess privileged knowledge of the body that predicts at least certain health outcomes. That is to say, "subjective" knowledge may sometimes be more valid and useful than "objective" knowledge. For example, Richard Lazarus and Raymond Launier (1978) have demonstrated that individuals' "self-appraisal" of stressful events can predict what will be experienced as stressful. More impressively, Ellen Idler (1992) and colleagues (1990) have found that "subjective" self-assessment by the elderly of their health condition is a better predictor of their own mortality than are "objective" biomedical measures. In a similar vein, Edward Yelin and colleagues (1980) demonstrated that self-assessment was a better predictor of return to work among disabled chronic pain patients than more objective clinical or radiological tests. Working with informants in Mexico who believe they suffer or are believed by others to suffer from a "culture-bound condition," susto (soul loss), Arthur Rubel and colleagues (1984) determined that even though biomedical measures could not diagnose the condition or determine the pathology, sufferers had higher mortality rates than matched community control subjects. This indicated to them that subjective complaints and lay interpretations were sensitive to disease processes that biomedical methods could not measure. Finally, the ethnographer's use of positioned subjective accounts is an interpretation based upon multiple accounts so that there is the potential for cross-checking and for consensual validation, for example, of illness narratives from different members of the social network.
The chapters by Sen and by Murray and Chen refer to surveys in which better-educated populations living in higher socioeconomic settings who utilize health care services more frequently, and therefore are presumed to be healthier, report more symptoms than poorer, less-educated populations, who utilize services less frequently and who live in settings with poorer indexes of general health. Thus, Sen reviews survey data showing that people in Kerala, the state that has the best health status in India, report more symptoms than populations in certain of India's poorest states, who would also be expected to be in poorer health, while in international comparisons the self-reported illnesses of Americans in surveys are of greater magnitude than those of Indians, who again would be expected by objective standards to be in poorer health. This finding—which needs to be qualified by surveys in the United States that show the opposite, namely that African Americans, low-income persons, and the elderly do in fact rate their health status lower than do whites or higher-income and "younger" persons in keeping with objective measures (Barker et al. 1991; Idler 1990)—should not diminish interest in the pertinence of personal complaints. It is well known in the social survey literature that perception of the relative significance of a problem in a person's life or in the experience of a group is highly dependent on the context and increases with rising expectations. It is also well known that members of disadvantaged groups will tend to suppress complaints of misery, especially when they are interviewed by members of dominant groups, if they feel there is little that can be done or that their complaints may cause trouble (Guarnaccia et al. 1990; Scott 1990). In fact, I take this to be Sen's point.
The positioned place of patient and family reports of symptoms, then, does not make those interpretations erroneous; it only qualifies that local knowledge. Some patients with peptic ulcer craters in their stomachs may not complain of pain; other individuals with the "classical" abdominal pain associated with peptic ulcer disease may have no gastrointestinal pathology. That does not mean that such patients are not conveying valid information about their experience of bodily processes; it only demonstrates a gap between experience and biology. Social expectation, cultural priority, and personal response fill that gap. Why should the materialistic determination of pathology and its projection into prognostic predictions about the future be considered as more authentic or real than the culturally and personally positioned human expression of suffering in the exigency of the here-and-now? I think Sen would agree on philosophical grounds, but Murray and Chen probably would not, because this distinction is, in fact, an operating tenet in international public health studies.
The tacit health knowledge of informants is used to inform health-relevant choices and actions, such as whether to smoke, wear a condom, wash one's hands, draw upon a social tie for strategic assistance in gaining access to health services, and so on. It contributes thereby to health outcomes. The sufferer's interpretation of suffering needs to be taken into account in the assessment of pain, distress, or dysfunction in order for such assessments to have validity in the experience of real people in real worlds.
Suffering is, moreover, not only a "subjective" phenomenon but also an interpersonal one. The distress of an Alzheimer's patient is part of a micro-context of serious trouble in which the family becomes a locus of the suffering and indeed even of the illness experience. We do not dwell in worlds where dichotomies between real/unreal disease, objective/subjective health problems, valid versus reliable diagnoses, or mind/body are part of everyday experience. These are professional constructions that are imposed on local worlds. Many chronic pain patients, for example, experience serious, disabling pain in the back or the head with objective evidence of only very modest, if any, "real" pathology on the CT scan or MRI. Pain for them is certainty, whereas for their health professionals it is doubt (Scarry 1985). A chronic back pain patient who is told not to worry because there is no evidence of "real" disease may be furious because that disaffirms his or her bodily experience. This is the logical outcome of the overvaluing of the biomedical understanding of objective measures and the devaluing of subjective or interpersonal ones—a characteristic of high-technology biomedicine. It has been shown repeatedly to have a dangerous effect on care (see Good, Brodwin et al. 1992). By delegitimating the interpersonal experience of suffering, this approach manages to be simultaneously inhumane and invalid.
Veena Das (1994), writing in the same volume in which this debate first appeared, criticizes the response of health professionals to the plight of the victims of the chemical poisoning at Bhopal. Her criticism turns on the insistence by those physicians that objective measures, not in Sen's sophisticated and qualified sense but in commonsense biomedical terms, replace self-report and family experience to determine the consequence of exposure. Thus, for the health experts, the complaints of many victims—impoverished adults and children who dwelt in a slum near the Union Carbide plant, and, therefore, whose health status was likely to have been poor before the disaster—were suspect because no "objective" evidence from "before" documented the "baseline" level of respiratory functioning and other physiological performance so that current status could be verified and any change that was documented could be attributed beyond any doubt to the disaster and not some other cause. For these professionals, the absence of objective baseline evidence meant that evidence gathered after the accident could never be objective enough to determine the cause. Thus, the victim was medicalized into a suspect who might be making unjustified or even illegal claims on welfare and disability funds—a profoundly disturbing transformation.
When objective indexes are used, they measure biological change as if it were fungible, separable from the experience of distress and the bearing of suffering. Indeed, "objectivity" in medicine is formulated so as to require such an artifactual, context-independent metric. Yet, can such an objective health index validly measure the consequence of industrial poisoning on the person, and on the experiential processes that integrate the affective, interpersonal, and neurobiological changes that together constitute the human quiddity of that person's suffering?
In one sense, ethnography comes closer than other methodologies to being "transpositional," because it requires specification and analysis of multiple (including self-) perspectives. Yet in Sen's analytic schema, ethnography may not qualify as providing transpositionally objective knowledge at all, inasmuch as the knowledge it constitutes remains positioned, or at best interpositional, experiential even if the ethnographer seeks a transpositional framework. Sen's carefully worded analysis clearly does not require that objectivity leave out human experience. Yet that may well be the unintended result when the quest for transpositional objectivity involves a leap from interpositional to extrapositional grounds and from intersubjective to objective claims. In this sense, an illusive ideal in measurement may become the enemy of a practical modus operandi for international health objectives: too rigorous a requirement for transpositional objectivity may lead the researcher to reject positioned knowledge. When this happens, say through the sole use of HIV sero-status and AIDS fatality rates to measure the AIDS pandemic, something fundamental to the human experience of suffering may be lost: namely, its grounding in greatly different local worlds and local lives (Farmer and Kleinman 1990). One impetus for the Chen, Kleinman, and Ware volume (1994) is disquiet among at least some international health researchers over too narrow an indexing of human suffering. This occurs, for example, when infant mortality rates obscure other measures of child "health." If child survival is accurately assessed but made into an "objective" sign of a community's positive health status, yet is followed by malnutrition, disability, and mental health problems among those same survivors—findings that clearly point in a more negative direction—then this measure is "objective" yet invalid. Perhaps the goal for international health studies should be to combine just that degree of transpositionally objective reliability which can be achieved while still making feasible a high degree of interpositional, experiential validity.
As the chapters in the section "Conceptual and Methodological Issues" in Health and Social Change clarify, classical metrics of mortality and morbidity are necessary but not sufficient to describe health transitions. They need to be supplemented by measures of disability and behavioral and social pathologies, and by humanly grounded descriptions of suffering. Narrative-based evaluations of the latter will differ in a fundamental way from numerical measures of the former. The same degrees of reliability and validity should not be required. Or, if they are, different kinds of measures (surveys enriched by stories of sickness) should be employed reciprocally to assure that both features of illness are assessed.
If these strictures are pertinent to the measurement of health changes, they are even more significant for the assessment of social change and the mediating processes that lead from particular social processes to health outcomes. To understand what urbanization in a low-income African society such as Kenya is about and to assess its effects on health, it is insufficient to treat urbanization as a unified variable that can be objectively operationalized, inquired about, and assigned a quantitative value for any population. Rather, urbanization as a concept needs to be unpacked—separated into subprocesses and understood in the historical context of particular settings of rural migration and urban resettlement.
This social change can be greatly different for a marginalized woman who migrates with her children from a community undergoing serious breakdown into a big city slum without support from husband or kin, as compared with a woman and her children who migrate from a vital rural community to which they retain effective social ties into a market town where husband and maternal kin have occupied a compound in a neighborhood of fellow villagers and are engaged in a thriving business. The characteristics of the rural-urban migrants (their ethnicity, age, gender, class, employment, education, and health status), the community from which they come, their resources, the type of uprooting (voluntary, forced), the forms of migration, the situation into which they migrate—all contribute to qualify the variable, urbanization, as does the history of migration in that society and the largescale political and economic forces that push and pull the displaced. To measure the mediators that transform social processes into health outcomes, the way to proceed would be to parse urbanization as a social category into those components that are locally valid. No social change could be validly assessed without evaluation of its contextual meaning for those who experience it. That meaning will mediate between the larger social world, the microprocesses of family and neighborhood, and the person's psychobiological processes; indeed, as a result of that mediation between the social, the biological, and the psychological those body and self processes are transformed into sociosomatic ones. Meaning will itself be a changing mix of cultural code, what is at stake in social negotiations over resources, personal experience, and conversations with positioned others, including the ethnographer. Measurement will need to canvass meaning in context, social process, and historical event, and their relationship to health indexes. Such relational measures will require both "objective" and "subjective" information.
Clearly, the requirements for this type of research cannot be met unless we move beyond the objectivist/subjectivist dichotomy, even with transpositional and positional qualifications. That movement either can be toward interdisciplinary forms of assessment that integrate quantitative and qualitative evaluations, or toward studies that encompass the separate contributions of narratives and numbers. Our terminology may be different, yet I interpret Amartya Sen's analysis as a complementary call for widening the space of measurement to make room for what is interpersonally at stake for human beings.
Objectivity and the Basic Values of International Health
The search for solid assurances of objectivity, as important as it is for international public health, is by no means an end in itself. If you ask experts in international health to explain just why objectivity is so important, they do not offer a basic scientist's defense of the scientific process. Nor do they launch into a discussion of epistemology and ontology. Quite to the contrary, they insist theirs is an applied field. And it is for a thoroughly applied reason that objective surveys are said to be crucial to their mission. International health professionals like to tell down-to-earth stories about desperately poor people dwelling in miserable urban slums or in villages on the dangerous margin of survival whose family members are suffering from sometimes fatal, yet potentially preventable, disorders—diarrheal diseases, measles, malaria—while the governmental agencies of those societies are expending the funds that should support progressive programs to control these serious problems on other, less essential, sometimes downright inappropriate services, too often ones that benefit those with power and wealth.
The moral of these paradigmatic tales is that "objective" measurement of the magnitude, distribution, and cost of health problems is essential if advocates of the health needs of the poor and powerless are to be able to effectively persuade the representatives of health and finance ministries to expend the resources necessary to control (or eradicate) preventable diseases. A second lesson these professional tales offer is that only objective measurements can establish, equitably and justly, who are the truly deserving and which problems are most serious and would benefit most from specific public health interventions. Hence, objectivity is ultimately necessary to authorize action, and action, based upon rational technical guidelines, is the moral mandate of this profession. (See Dasgupta 1993 for a similar argument within development economics.)
A child of Enlightenment thinking if ever there was one, international health, which is one of the pivotal fields of social development, sponsors a rational technical discourse that regards health as the progressive achievement of societies that modernize. Among the many things that modernization denotes is movement through a transition that separates underdevelopment from development; high fertility from low; poverty and ignorance and passivity and superstition from a modernity based upon education and rational choice; and high infant mortality rates, high infectious disease rates, and compromised longevity from the improved health indexes of industrialized societies (see, for example, Caldwell et al. 1990).
This transition—variously called the demographic, epidemiological, or simply the health transition—is held to be one of the major milestones on the path of progress. The movement along this path is described in the objective measures of improving health and social indices: fertility, mortality, education, gross national product, and so on. Progress should be continuous along axes of increasing industrialization, technology transfer, democratization, professional expertise, individual autonomy and empowerment, and, of course, health. Action is needed to remove obstacles, advance the pace of change, and control or eradicate disease. Thus, the idea that objectivity underwrites practical applications is linked, "objectively" as it were, to the idea of progress. Together they constitute the moral basis of the profession of international public health.
From the perspectives of anthropology and social history this positivist value orientation is obviously more than a little problematic. To begin with, the ideas of progress and development assume a universal standard against which the experiences of different societies can be compared as evidence of greater or lesser success. The particularities of social change in a region or series of local contexts must be ironed out to fit the universal metric. The process of "ironing out" is one of aggregating data at ever higher levels, ultimately ending up in a comparison of the "developed" and the "developing" societies, or the nonindustrialized, the industrializing, and the industrialized, a dual or tripartite categorization into which all societies are made to fit. Aggregated data are also "cleaned" so that what is peculiar to a context is removed from what is shared and therefore generalizable. Cross-cultural ethnographic comparisons, much like social histories, tell a very different story, of course. Here there are few instances of progressive advance toward health, or universal principles guiding the aggregation of greatly different societies and regions. In their place, there are detailed stories of the local history of particular places and peoples undergoing many kinds of social change with all sorts of antecedents and consequences.
Nancy Scheper-Hughes (1992; and see chapter 9 below) describes a favela in a small Brazilian city in the impoverished northeast in which misery is endemic and health problems multiple and seemingly incorrigible, even as the region and the rest of the city "progress" economically through the production of sugar and as certain health problems are improved. The infant mortality rate may be reduced through earlier access to more effective treatment of diarrheal disease with associate life-threatening dehydration, but malnutrition, other infectious disorders, and mental health problems are widespread. Is this an indubitable instance of success? Few would regard it as such since the poverty and violence and suffering of everyday life are so extensive; yet for public health purposes, the improvement in outcome of severe diarrheal disease will be depicted as a story of success. Here medicalizing social problems, focusing upon an individual disorder in a field of interrelated problems, and applying the positivist metrics of objective measurement woefully mislead.
China in 1994 offers another example of the complexities of development. The coastal edge, where perhaps 200 million people reside, possesses the world's fastest growing economy (a 13 percent growth rate). Urbanites and rural villagers enjoy a greatly improved standard of living. But China has also an interior where 750 to 800 million people, mostly very poor peasants, are experiencing "development" at a very different pace. Whereas the coast from Guangdong to Jilin has transformed China into the world's third largest economy (according to Purchasing Power Parity), after the United States and Japan, the huge interior is closer in living conditions to India than it is to Singapore, Hong Kong, or Taiwan. In the southwest and northwest provinces, yet another 200 million Chinese live in conditions of often extreme poverty. Here infant mortality rates can be five times higher than the rate in Shanghai; the poorest peasants rarely eat meat and often get inadequate amounts of vegetables. Some can't afford rice on a regular basis. Housing is also inadequate, and there are not enough clothes to go around. Whereas peasants in the coastal periphery can afford a fairly high level of fee-for-service health care, in the most impoverished interior, they cannot. And as a result they suffer without biomedical care, inasmuch as the rural health care system and its insurance schemes, in impoverished areas, have broken down under the pressure of decollectivization and privatization. Some claim that the situation of the poorest peasants is no different than under the Qing Dynasty. This third sector of the Chinese economy and polity contains most of China's minority groups, whose standard of living is far below that of Han Chinese. Tens of millions of internal economic migrants who are encamped in China's cities come from this "other" China. Economic and health conditions of this "floating population" vary greatly, but many are destitute, their shantytowns sites of violence and sickness.
In many areas of China, crime, prostitution, drug abuse, and sexually transmitted disease have returned to the high rates that prevailed under the Nationalists, after forty years during which they were effectively contained by the diffused political control of the communist state, which simultaneously created political violence and human rights abuses on an immense scale. New problems have arisen related to appalling levels of industrial hygiene, inadequate environmental safeguards, man-made catastrophes such as explosions at munition manufacturers' plants, and the early onset of an AIDS epidemic. Success and failure are systematically related, as in the case of cigarette smoking, which places tens of millions of Chinese at risk for lung cancer, emphysema, heart disease, and stroke, yet also fills the coffers of the state's tobacco monopoly and thereby contributes to economic success.
How are these predicaments of health to be assessed within the paradigm of "development"? There is evidence of economic and health successes, economic and health failures, the dark side of rapid economic reform, and continued underdevelopment. Intrasocietal diversity, and indeed fragmentation, is so extensive as to defeat a single index of socioeconomic and health change. Does it even seem sensible anymore to use aggregate statistics to describe China as a single social unit? Do such statistics describe an "objective" entity? Is aggregation across regions valid? Can there be a "transpositional objectivity" when there are so very many "realities" that stand for China's economic and social conditions? Will a single narrative of health and social change in China do?
What interests are served by the quest for objective data? Clearly, first served are those interests of the professionals whose technical construction of a public problem can then become the basis for developing and evaluating intervention programs (see Gusfield 1981).
Doubtless at times those professional interests will correspond with the needs of the public. The eradication of smallpox is an outstanding example. Yet, frequently the acquisition of technical data is self-serving and does not contribute in a significant way to the health of the community. By privileging questions of the objectivity of observation, public health surveys routinely avoid questions of the "interests" that stand behind and direct the search for objective findings. Therefore, the political and economic interests of the state or of dominant factions are ultimately served.
This need not be the case. Professional formulation of the health components of societal problems can articulate (and occasionally have done so) interrelated problems of powerlessness, poverty, and suffering as a broader social policy horizon requiring political and economic transformation. Indeed, historical, anthropological, and sociological studies regularly do so, as do, less frequently, epidemiological surveys. The question, then, becomes one of discovering how professional public health interests have been subordinated to dominant political economic interests. While students of this topic have examined the politics of health policy and the sociology of the professions (see, for example, Justice 1985; Freidson 1986), I think it is equally important to examine the relationship of scientific epistemology and public health practice to cultural orientations that are closely linked with political economic interests and institutions.
The entire field of social development, involving the academy, the foundations, governmental agencies, nongovernmental agencies, and international institutions such as the World Bank, is oriented around what is called "science-based" development. But what kind of science? In the main, not that kind of science which emphasizes critical engagement with the purposes and concepts of social change and its relationship to health, or which points to the variety and diversity of changes and outcomes. Rather, science-based development all too often builds upon economic, epidemiological, and policy analyses that privilege the transpositionally objective measurement of narrowly articulated professional categories—the view from nowhere (Nagel 1986)—over the anthropology, history, and sociology of the view from a particular somewhere by a particular somebody with some special interest at stake.
In this way, the social development field connects with the idea of a value-free, objective domain of health and human development.Here progress and flourishing are "natural" and social and health difficulties are professionally articulated as "human problems," obstacles to progress that require technical remedies. Thereby, poverty is constructed as a risk factor, unclean water and uncontrolled sewage are precipitating causes, and microorganisms are the agents of ill health for which technical fixes can be arranged. Their systematic relationship to one another and to political and economic structures, however, goes unaddressed.
This approach is not without its utility for the solution of narrowly conceived, though still important, problems such as water and sewage control, a utility which no one could argue against. Yet the emphasis on science-based development, I would suggest, is also at times what is wrong with international development and public health. The sources of human misery are based in the dynamic effects of large-scale political and economic forces working within local worlds. Those causal effects are mediated by cultural and historical processes that place certain categories of persons under great pressure while protecting others, that provide resources for some while blocking access to resources among others. It is not that these forces, processes, and effects cannot be objectively measured, but rather that the narrow technical pursuit of objectivity, as currently constructed by the international health professions, gets in the way of the larger conceptual and methodological task of reconstructing and reinterpreting those social sources of human misery so that their powerful interconnections can be targeted for interventions.
Thus, what at first glance seems to be a fairly recondite question of the epistemology and ontology of objectivity points to deep values in international public health that are themselves a major barrier to understanding the sources of ill health and studying ways to effectively address them. Anthropology and social history offer a needed complement because they critique the deep-grained assumptions that need to be recast. Only through the concrete understanding of particular worlds of suffering and the way they are shaped by political economy and cultural change can we possibly come to terms with the complex human experiences that underwrite health.
And yet, the quest for objectivity also offers certain strengths that cannot be waved away (and indeed are not) by most historians of medicine and anthropologists of health. These strenghts complement and go beyond interpretive approaches and suggest that an adequate study of international health problems requires both narratives and numbers. To more adequately formulate and assess leading international health issues such as AIDS, drug abuse and related violence, forced uprooting and migration, and malaria and tuberculosis, the appropriate goal should be to work out modi operandi that reciprocally use quantitative and qualitative measures, that insist on validity together with reliability, and that balance history with epidemiology and ethnography with policy analysis. That reciprocal engagement is also the way to test the significance and practicality of conceptual as well as methodological practices.
There is, of course, a huge barrier to the integration of anthropological and international public health approaches. Although some anthropologists have learned to work effectively with international health experts with impressive achievement—say, in assisting public health programs to proceed in culturally appropriate ways in local communities (Nichter 1989; Kendall 1988; Nichter and Kendall 1991)—they have usually (and, given their goal, appropriately) not challenged the fundamental orientation of public health. Yet, taken to an extreme, the critique of the cult of objectivity does lead to an almost insurmountable divide, an epistemological and, one wants to say, ideological disjunction between the two professions that I for one find uncomfortable, unnecessary, and unhelpful. To explore the divide at the most extreme point, let's examine a position set out by Nancy Scheper-Hughes (1992:171–172), a medical anthropologist who has employed ethnography to provocatively challenge policy directions and criticize the alleged abuses of medicine.
Explicitly following in the well-trod footsteps of the European tradition of critical theory, she notes that "commonsense reality may be false, illusory, and oppressive" (p. 171). The goal of the critical medical anthropologist is to search out buried truths in order to "'speak truth' to power and domination" (p. 171). This means exposing the role of "traditional" intellectuals and experts who are "bourgeois agents of the social consensus" that maintains "hegemonic ideas and practices" (p. 171). Doctors are such experts, and therefore serve a crucial social function by misidentifying the hidden indignation of the oppressed poor who are sick and its sources in the sociopolitical order. Anthropologists who fail to speak this class truth occupy the same role as doctors (p. 172). The medicalization of folk complaints—that is, the doctor's diagnosis and treatment of folk idioms of complaint—is a kind of appropriation that may treat the symptoms but simultaneously camouflages their social cause. Therefore medical treatment is "bad faith," "an oblique but powerful defense strategy of the state" (pp. 169–170).
This analytic line is not hyperbolic for its own sake. It is obviously meant to draw a dividing line between critical theorists and others, including international public health professionals and anthropologists who do not accept the idea that they are collaborating in "bad faith" with hegemonic projects of the state and transnational corporate capitalism to boot. The critical theorists claim moral superiority and accuse others of immoral behavior; that is, other professionals in the health field are implicated as working with bad faith or false consciousness, mystifying the political reality via medical applications.
So unqualified a revolutionary claim will strike many—especially given the current vicissitudes of revolutions—as mischievously overdrawn. This margin is untenable. After all, has not even Henri Lefebvre (1991:25), one of Europe's leading Marxist theoreticians, pronounced: "Critical theory, after being driven into practical opposition—and even into the most radical form of it—has had its day." Fortunately, relatively few medical anthropologists, including perhaps even Scheper-Hughes in her more generously collaborative style, would go along with this extreme position, based upon which any useful collaboration with health professionals short of revolutionary change would seem almost impossible. (Indeed, the purpose of the extremist position may be to make interdisciplinary collaboration much more difficult.)
And yet, I also think many medical anthropologists will properly regard a critical perspective (including the important findings in Scheper-Hughes's ambitious ethnography of everyday violence in the lives of the chronically poor and hungry) to be an important modus operandi in their work. I certainly do. A critical cultural analysis is needed to unpack international health's value commitment to the objective measurement of change in health status based upon the cultural idea of progress. Cultural critique is also necessary because of our epoch's construction of the great issues in health largely (and sometimes solely) as questions of economic costs and their control. The historical relationship between cultural practices, politics, and economics in the latest phase of capitalism should be at the heart of ethnographic studies, be they in Africa or America. Anthropology's project in international health (not the only one but surely an important one) should be to sponsor critical studies at several levels and also to encourage research that balances or complements objectivity with criticism. That means that besides collaborating with international health professionals in programs that are organized within a public health framework, anthropologists need to organize alternative frameworks for international health studies that conceptualize health, health conditions, and health care in social theoretical ways that can use ethnography and cross-cultural comparison as chief methodologies.
Although useful efforts to do so are under way, it is not at all surprising, given the early stage of research and the difficulty of securing core research funding, that these efforts have not yet created a substantial body of work that is comparable to economic, policy analytic, and epidemiological contributions. Because questions of the objectivity of methodology and findings have dominated the latter approaches to international health research, the conceptual basis of international health has remained remarkably underdeveloped. Hence there is a special opportunity for critically informed social theory, though of a somewhat more circumscribed variety, to define and fill in the conceptual space of international health. It is for this reason that an analysis of objectivity, as remote as it may seem at first from the exigent social problems of human conditions, is an occasion to critique what has been done so far and offers an opening to suggest what could be done in future. It is a rebuke to frustration over established strategies (see Desjarlais et al. 1995).
Hilary Putnam suggests that with respect to objectivity we might as well
|【字体：大 中 小】【打印】【关闭】|