To define relativism is, of necessity, to take a position in the controversy surrounding it. This is because, as Geertz (1984) has suggested, relativism, especially in its oft-derided ‘epistemolo-gical’ form, has become one of those peculiar by-products of anthropological spleen more deplored, and therefore more fully developed, by its opponents than held by those accused of being its adherents. For this very reason, it is helpful to begin by distinguishing three different kinds of relativism: conventional cultural relativism, ethical relativism, and epistemological or ‘cognitive’ relativism. These three notions, defined below, sometimes overlap, as nothing precludes someone from holding two or more of these positions at once; nonetheless, they constitute quite different concerns, and, more importantly, are subject to quite different degrees of controversy.
Many anthropologists since World War II would probably find themselves agreeing with a conventional cultural relativism — and disagreeing with its being called a form of relativism. Less a philosophical position than a vague methodological attitude, the conventional cultural relativism that most anthropologists, British or American, take to work with them is a combination of two notions: first, that insofar as there are behavioural differences between various populations of people, these differences are the result of cultural (sometimes societal) variation rather than anything else; and, second, that such differences as do exist are deserving of respect and understanding in their own terms. In its weakest, most popular version, conventional cultural relativism is agnostic about whether or not some universals, like death, eating, reproduction, or even class conflict and pan-psychological desires, might exist to define the limits of human diversity. Some anthropologists would make this list of universals large, many more would keep it small, and a few would argue against any; but any anthropologist not supposing all human action (or any human behavioural variation) the result of in-built universal forces would still qualify, in this sense, as a relativist. As applied in practice, this kind of relativism has also generally held most Western epistemological presumptions constant, and has often avoided troubling ethical issues altogether by reducing them to matters of local causation or design. This is, perhaps, because not questioning the first allows one to avoid debating epistemological relativism; and keeping the second issue from ever surfacing neatly side-steps the conundrums of ethical relativism. Still, this sort of conventional cultural relativism remains a kind of relativism, since, as per disciplinary convention, it assumes that the behavioural variations which interest anthropologists must be understood within the cultural or social frameworks which contain them.
This conventional cultural relativism is not a very thorough-going kind of relativism, for the strength of its claims tend to fade at the edges of its ethnographic interests; but it is a very safe view to propound. Hence, conventional cultural relativism is what students often take home with them from introductory anthropology classes. Moreover, it has proven a good club with which to batter arguments about race and fethno-centrism, and a practical attitude with which to approach ethnographic research. Still, this form of relativism is obviously not a proper philosophical position — nor, practically speaking, could it be. Rather it is a conveniently sloppy framework — a kind of ‘work-ethic’, or disciplinary ‘common sense’ — within which many anthropologists are comfortable discussing human variation, and which is the residual end-result of the more properly philosophical debates about ethical and epistemological relativism that have swirled around ethnographically based anthropological research since its inception. The arguments about these forms of relativism, thus, are really struggles over what will ultimately constitute the conventional cultural relativism, which is, really, anthropology’s practical core. And so we must turn to these more blatant controversies.
Ethical relativism
Ethical relativism is the notion that the business of making universal, cross-cultural, ethical judgements is both incoherent and unfair because moral values are a product of each culture’s unique developmental history, and can, thus, only be judged in relation to that history. Versions of this form of ethical relativism, varying in strength from a weak call for mutual inter-cultural tolerance to the (rare) strong demand for ethical compartmentalization, have been part of anthropology and the traditions of thought that gave rise to it since, at least, Montesquieu’s De I’esprit des lois (1979 [1748]). This kind of relativism was most popular, however, in the 1930s in both British social anthropology and American cultural anthropology. British anthropology’s various forms of functional-ism, for example, seemed to suggest that any given society’s ethical practices were the result of long-standing structural or practical developments, served complex and subtle purposes, and thus were not to be tampered with. Similarly, in the United States, Boas (1911) and his students Benedict (1934), and "Herskovits (1972), held to a diffusionist view of cultural development which tended to undercut universalist claims, and made ethical relativism seem empirically obvious, methodologically necessary (to avoid ethocentrism), and obviously just. They also found it a useful jumping-off place for launching attacks on American provincialism, racism, and anti-Semitism.
These notions were, perhaps, best laid out by Melville Herskovits. In the late 1940s and early 1950s, Herskovits, inspired by his repugnance for racism and cultural imperialism, claimed that ‘cultural relativism’, as he termed it, was not merely pragmatically useful but empirically proven by the ethnographic record; and then went on to flirt with a form of epistemological relativism (Herskovits 1972), suggesting that perhaps not only mores but perceptions as well were products of enculturation. However, Hers-kovits, for whom relativism was really more a way of criticizing colonial imperialism than a proposal for a thorough-going philosophical doctrine, was soon brought to the dock by philosophers, and some anthropologists, on a charge of paradox: i.e. for asserting, as an absolute truth, the claim that all truths are culturally relative, and was thus, perhaps unfairly, hoisted by his own petard. In any case support for ethical relativism began to erode during and after World War II, for various, sometimes ironic, reasons: the inadequacy of applying relativist tolerance to that period’s sanguinary ideologies; the postwar demand for, and expectation of, development by the ‘New States’; and the devastating postcolonial critique that ethical relativism, and the ‘don’t touch’ approach to intercultural contact it seemed to approve, was merely a Western ideology that played into the hands of colonial masters by justifying their repressive status quo.
Still, this controversy has by no means died out. In the 1980s and early 1990s, for example, a claim by the cognitive psychologist Lawrence Kohlberg to have substantiated the existence of a universal developmental sequence of moral reasoning (one culminating, rather suspiciously, in something like Western liberalism) was met by equally passionate counter-claims by Shweder et al. (1990) that this ‘confirmation’ was merely an effect of the Western presuppositions built into his data-gathering technique. More painfully, current arguments among anthropologists about power, human rights, and ritual genital mutilation reveal an old tension between the cosmopolitan tolerance of ethical relativism, with its corollary suspicion of the motives of those wishing to find or impose ‘universal values’, and an activist intolerance of repressive or violent conditions, with its equally apposite distrust of those who would allow such conditions to continue in the name of cultural autonomy. Nor has the routinization of some reflexivity in much American ethnographic writing effaced this issue. Indeed, by questioning the authorial motives of both those who would remain relativistically neutral and those who would sometimes hazard judgement, reflexivity actually serves to heighten this dilemma. So the debate goes on, revealing an inevitable tension within the field, and discomforting, by extension, the sometimes complacent political/ethical neutralism of anthropology’s conventional cultural relativism.
Epistemological relativism
This stands in contrast, however, to the debate about epistemological relativism which shows some signs, lately, of breaking down. Epistemo-logical relativism, sometimes called ‘cognitive’ relativism, is often defined, by its critics, as the assertion that systems of knowledge possessed by different cultures are ‘incommensurable’ (i.e. not comparable, not translatable, utterly alien, etc.), and that people in different cultures, therefore, are believed by epistemological relativists to live in different, equally ‘true’, cognitive ‘worlds’. Of course, defined this way, epistemological relativism is a straw man, and one embraced by very few of those accused of being its adherents (Rosaldo 1989: 218-24). This is partly because, so posited, it is obviously self-refuting. That is, the thesis that, as different cultures are incommensurable, therefore comparison and universal assertions are impossible, is, as Herskovits found out, one of the very class of statements that it has just declared impossible. Moreover, this definition of epistemological relativism also misses its mark. For its self-refuting character has not in practice entailed (nor, by its own logic, need it only imply) the epistemologically conservative alternatives its critics favour; the epistemologi-cal agnosticism of many symbolic anthropologists, for example, neatly sidesteps this critique.
Nevertheless, this kind of epistemological relativism could be said, with varying degrees of justice, to have arisen in connection with, or even out of, three major anthropological debates. The first such debates were among linguists in the 1950s about the, misnamed, Sapir-Whorf hypothesis. These concerned whether, or if, one could say that the language one spoke, and its capacity for expressing certain kinds of concepts, determined, fully or partly, the kind of experience (or ‘reality’) one could have. These discussions were flawed, however, by ambiguity over whether this was a debate about first principles in epistemology or about an observational finding in linguistics, and were not helped by being concerned with a ‘hypothesis’ that was never actually formalized as such by either "Sapir or "Whorf. The second round of debates, in the late 1950s and 1960s, were about the extent to which rationality, or the way humans think, is or is not universal. This was sparked by the use made by some philosophers, such as Peter Winch and Willard Van Orman Quine, of the discussions of proof and ‘primitive’ thought in "Evans-Pritchard’s ethnographic work (Evans-Pritchard 1956) — work which was, itself (as can be seen in Evans-Pritchard’s discussion of Nuer religious symbols) a refutation of "Levy-Bruhl’s earlier theories that posited a radical difference between ‘modern’ and ‘prelogical primitive’ thinking (Evans-Pritchard 1956: 123—43). Unfortunately, Quine’s (1960) carefully illustrated doubts about the possibility of literal translation — and, therefore, of comparison — from one language to another have received fairly little discussion among anthropologists. More attention has been paid to Winch’s (1958) argument that if rules of thinking are socially constructed; and if, therefore, rationality could be said to differ validly from culture to culture; then the social sciences must be seen as simply obscuring what they hope to understand whenever they try to ‘explain’, in Western scientific terms, the apparently irrational statements and behaviours of, say, Western religious believers, or almost all non-Western peoples. However, the ensuing discussion of whether, or to what extent, rationality was indeed culturally constructed (with most of Winch’s critics answering ‘no’ and ‘only in part’, respectively) was flawed by far too little ethnographic consideration of whether the Western scientific rationality almost all (including Winch) used as a backdrop to this discussion was itself, in daily scientific practice, quite so ‘ rational’ .
Finally, more recently, some epistemologically conservative critics have assumed, inaccurately, that epistemological relativism constitutes the core of many approaches with which they disagree: symbolic anthropology, hermeneutics, reflexivity, sometimes feminism, cultural studies, and the denial of universal epistemological foundations by postmodernists and poststructuralists. This is a rather odd list, however, since it is very difficult to find an anthropologist from this extremely heterodox collection of approaches who also claims, or recommends, epistemological relativism. Moreover, the term ‘relativism’ is more often thrown about among this group, especially by some feminist scholars, as an accusation than as an accolade. What all these groups do have in common, however, is having argued (in various ways) for a critique of conventional Western scientific rationality. And this is the real nub. For it is, indeed, the very idea that Western rationality as such is vulnerable to political or epistemological critique that is seen, by epistemological conservatives, as implying an underlying relativism, and by many of those so accused as implying nothing of the sort.
The resulting disagreement about who is or who is not a relativist reveals, perhaps, that the debate about epistemological relativism has an underlying instability, and, at long last, may be falling apart. Or, at least, a number of reasons have been offered as to why it may eventually do so. Clifford Geertz, for example, has suggested that anti-relativism is really just a symptom of pre-ethnographic nostalgia, an attempt to put the apple of human diversity back into the tree of Enlightenment rationality. But the option of staying home for home truths, despite odd arguments by the philosopher Richard Rorty in favour of a kind of educated ethnocentrism (1991: 203—11), seems to be fading even for the most determinedly provincial. Moreover, although arguing about epistemological relativism by arguing about the presence, character, or number of human universals — the form taken by all the arguments above — made some sense in regard to the Sapir-Whorf hypothesis or the rationality issue, where both the concept of culture and of Western scientific rationality were held fairly constant by both sides, in more recent debates, where this is often not the case, such an approach to ‘refuting relativism’ (along with the concept of epistemological relativism itself) hovers on the edge of incoherence. After all, it is hard to accuse those who question the concept of culture of believing in a culturally based epistemological relativism. It is no surprise, then, that recent post-cultural attempts by theorists in feminism, science studies, and symbolic anthropology to construct alternative ‘objectivities’ (Haraway 1991: 183-202), truly ‘relative relativisms’ (Latour 1993: 91-129), or stances of epistemological agnosticism, though shot through with their own problems, are not affected by the traditional critiques of epistemological relativism. They simply are not good examples of what they are accused of being. Yet, precisely because of this, their approaches to the relativism debate also threaten, for good and ill, the political complacency and comfortable episte-mological sloppiness of anthropology’s old conventional cultural relativism, and thus leave the discipline again uncertain at its very heart. But, perhaps, for anthropology, that is only to be expected.
