The reason I have for taking part in this debate is not to rehash inconclusive arguments over methodology, but rather to revive interest in the objectives of anthropology. It is appropriate to be concerned with the means of acquiring knowledge if we are confident that the established ends of our collective efforts are sound. But, when our social purpose is uncertain and our discipline reflects a general intellectual malaise, preoccupation with means rather than ends becomes self-defeating.
I take British social anthropology today to be marginal, fragmented, confused; to be obsessed with its own internal affairs more than with any larger conception of the purposes of knowledge. The demoralized descendants of Radcliffe-Brown and Malinowski are now increasingly given to insular reflection on the sources of their own anxiety, an anxiety induced by the loss of empire and with it the closure of that window on the world which once gave British anthropology its breadth and global vision.
We are not alone in this. The great project of modern social science with which our century began is manifestly in disarray. The West is now facing for the first time, in the shape of Asia’s resurgence, a challenge to its intellectual and practical ascendency. If we wish to situate our dilemmas within these epochal events, it will not do to dwell on the methodological legacies of those who wrested a niche for social anthropology in the British academy half a century ago. Rather, we must take a broader view of our place in world history, the better to devise a strategy for making a constructive contribution to understanding modern society’s next phase.
There are two great ideas driving modern history and they are inextricably linked—democracy and science. The first of these says that societies fit for human beings to live in must guarantee the freedom and equality of all citizens so that the people may be self-governing. The second says that such societies can only flourish if knowledge in them is based on the discovery of what is objectively real.
Science has two great objects—nature (everything out there that we did not consciously make up) and society itself (which is both a part of nature and the result of human intentions, however misguided). A democratic society has to break down intrinsic barriers to its own development—poverty, ignorance, injustice. To do so it needs science. Whatever we plan to do is more likely to succeed if we employ reason to find out how essential things work.
Moreover, the principles on which society is founded must be common to all of us; they must touch what is ‘natural’ in us, as opposed to what is merely conventional or arbitrary. This is the core of the modern quest for human, civil or natural rights. Equally, science thrives on democratic social organization. It is above all a communal enterprise, relying on the painstaking, cumulative efforts of generations towards shared ends. When science is merely an elite exercise, cut off from the general impulses of ordinary people, it is in danger of atrophying. The idea that links the two sides is education. Free and equal citizens must be knowledgeable. And science must be sustained by a general culture which values truth, learning and practical invention.
There is only one modern revolution and it is far from finished. It began in earnest in seventeenth-century England, which Veblen once described as ‘an isolation hospital for science, technology and civil rights’. The discoveries of Locke and Newton were made general in the eighteenth century by the European Enlightenment and were realized as a living social experiment in the United States of America. Since then, the French and Russian revolutions have dominated the thinking of progressive intellectuals. And, as the Western industrial nations became more wealthy and, it must be said, more equal, the pursuit of knowledge has often become more esoteric and personal.
Of late it has been claimed that we are already in a post-modern, post-industrial or post-scientific phase. I doubt if this is true even of the richest countries; but it is manifestly untrue of world society as a whole, where poverty, ignorance and the starkest inequalities are normal for the vast majority. The task of building a world fit for human beings to live in has barely begun.
Two attitudes predominate among our intellectual and political elites: one turns its nose up at ‘bourgeois’ democracy and science, declaring them a sham, without enquiring too deeply into institutional realities; while the other rejoices in the apparent achievements of ‘free’ Western states whose citizens remain at this time extremely unfree and unequal, being governed by remote rulers for whom science is largely an aspect of the military budget, rather than a means of general emancipation.
It may be objected that my idea of science is old-hat, that the world has moved on in the last three hundred years. And so it has. Keywords like nature, society and science move with history. This is a dialectical process and its principle is negation. What science is supposed not to be, its place in a set of terms referring to what it is not, offers a better guide to historical shifts in its meaning than positive definitions taken in isolation.
There can be little doubt that what science originally was not was mystical beliefs—religion, superstition, stories—uninspected traditions referring human existence to a supernatural cause; in a word, it was not ‘myth’. After five thousand years of agrarian civilization, the main task of modern societies is to found knowledge on a truly secular footing. Even a century ago the political drive sustaining science was largely anti-clerical; and, in a world where fundamentalist Christianity and Islam flourish (not to mention the Catholic Church), this crusade is still necessary.
Yet, in this century, for most Western intellectuals that battle may appear to have been won. What science principally is not has shifted ground to embrace the oppositions which sustain an expanded academic division of labour. The negation of science is now most commonly the creative arts—literature, poetry, the critical imagination—reflecting the division between natural science and the humanities (the separation of matter and spirit) which has spawned, as a hybrid experiment, social science.
Today’s debate could be taken as a referendum on the social sciences and on anthropology’s place as one of them. Most of this audience probably came to it with the word ‘science’ already fixed in mind as a positive or negative notion defined by one of several linked oppositions, all of them retained in present-day usage. For the founders of British social anthropology, our science of ethnography had as its principal negation ‘history’ or Victorian evolutionism. Now ethnography may be appropriated by the advocates of anthropology as writing and reflection, the very antithesis of science. Meanwhile scientific anthropologists are likely to insist that their subject matter is largely historical.
It is for this reason that I have sought to rescue the original and, I would hope, unifying conception of science as one of the two great objects of modern development. I feel sure that, if we concern ourselves with the method of knowing rather than with the object of knowledge, we will repeat the mistake which has led twentieth-century social science into a blind alley; and our debate will be hopelessly confused. Science undoubtedly rests on the premiss that it is possible to know what is objectively real. But to be committed to that idea is not to be forced to sign up for an ossified seventeenth-century epistemology, as, for example, economics has (thereby revealing itself to be more secular religion than science).
The intellectual achievements of the last three hundred years, in both science and the humanities, have necessarily altered our conception of subject-object relations and of ways of knowing. A modern science must incorporate notions of history, reflexivity, relativity, linguistic and logical traps, Western ethnocentrism, the need for self-knowledge and much else. The best twentieth-century scientists have already done so. The ideal type ‘science’—the positivist stereotype of the man in the white coat—cannot capture what scientists, the best and the worst of them, actually do.
The mistake is to emulate scientific method, while forgetting what science is supposed to be for—to be so wrapped up in the problem of one’s own ability to know or communicate anything that the priorities determining what needs to be known are lost. If modern anthropologists can often be seen to fall into this error, they are no more guilty than most modern intellectuals. We have lost our way; and this may be because we can no longer see the connection between the social purposes of knowledge and the pursuit of knowledge for its own sake.
Our problem is that the natural scientists can no longer relate what they do to the complex character of human existence, including their own; while the humanists have given up trying to understand how the world works. The social scientists have proved incapable of spanning the gap, for a number of reasons, but mainly because they tried to behave like scientists without seeking to alter what natural scientists think or being able to learn fast enough from what they have discovered. As a result they (including ourselves, as British social anthropologists) have nothing to say about the reciprocal interdependence of nature and society.
My contention is that our civilization desperately needs to reconstitute the original Enlightenment goal of progress through the systematic application of reason, in a world where nature and human society are understood to be interdependent. The prevailing division of intellectual effort within the universities stands in the way of such a development. But the progress of humanity on a world scale will demand a new concept of scientific knowledge and of its constituent branches. The mechanization of brains is one aspect of our phase of the modern revolution which is already requiring such a reorganization.
Even as I contend that anthropology must be part of the great modern project to institute democratic societies on a firm basis of objective knowledge, I would still argue that the experiment known as social science has been a failure and that we would be well advised to distance ourselves from it as fast as we can. Otherwise we will soon go down with sociology, economics and other benighted ‘pseudo-sciences’ into that dustbin of history reserved for disciplines which failed to move with their times.
The task of our generation is to bring knowledge of nature and society once more into an active, mutually reinforcing relationship. This means mediating and ultimately transcending the opposition between science and the humanities. Anthropologists are uniquely placed to begin such a task. We retain vestiges of an evolutionary anthropology which combines the study of humanity’s nature, societies and cultures. Even within social anthropology we combine both the scientific tradition of social theory and humanistic scholarship, as well as our own distinctive hybrid style of ethnographic writing, of abstract generalization pursued through concrete description. Above all our subject matter is the vital, inclusive middle ground—humankind as a whole.
Our virtue as an eclectic anti-discipline is that we are (or should be) open to all the currents which will make the next intellectual synthesis. It would be absurd to tie ourselves to the analytical relics of the last synthesis, to the social sciences that were formed in the early twentieth century. I do not know what this next synthesis will call itself, but I suspect that it may be ‘science’, perhaps ‘human sciences’. The rhetorical power of the word is too strong for us to abandon it lightly. If Derrida and the deconstructionists are human scientists—and I take their synoptic review of the history of Western thought to be a scientific enterprise—then current academic divisions cannot be taken seriously as a guide to whatever science may become in the next century. It is only through the dialectical synthesis of what science is and what it is not that progress in the pursuit of knowledge can continue. What matters is that we should seek to play an active role in the ongoing redefinition of what knowledge is centrally thought to be in our society. If we do not, we deserve to go under in the global upheaval that is building to a climax under our noses.
The task of a science is to generate replicable knowledge; to help others to do difficult things more easily and reliably; to get something right again and again; and, when it is no longer right enough, to think again. Anthropology as an academic discipline must be a part of science. We are in the public domain and we must fight for our place there. The rhetoric of serious public discourse concerns science. We have plenty to say about what that ought to be. We are not artists, even less priests. We have nothing to gain by declaring ourselves to be against science.
AGAINST THE MOTION (1)
The ‘opposition’ will divide its labour: Judith Okely will deal with the issues of ‘science’; I will limit myself to the question of generalization.
It is undeniable that, throughout most of its history, social anthropology has been a generalizing discipline. Indeed, one might well go so far as to say that if our predecessors did not generalize, they did nothing. But our concern today is not to debate historical facts. Rather, we must address ourselves to contemporary truths.
Our subject is not what the discipline has been in the past, but what it should be now, this normative view informed both by developments which have transformed anthropology especially during the last twenty years or so, and by the intellectual and political circumstances within which we are presently working.
Whatever the ambivalence with which any of us approach this motion today, I am certain that none of us will be tempted into the kind of soggy compromise offered recently by Peacock in his The anthropological lens. There, he tells us that anthropology deals ‘exquisitely’ (his word) with individuals—but sees them as representative of their societies. ‘Ethnography,’ he says, ‘reveals the general through the particular From the Kula ring, we learn about order and integration; from the shaky-handed circumciser, about the interplay of tradition and conflict; and from the cockfight, about hierarchy.’ 1The adjacency of these two latter examples would have intrigued Freud, but there is nothing very intriguing about Peacock’s contention: it is nonsense. We have been taught to make sense of the Kula as if it was ordering process; and to see Ndembu circumcision as an exemplary case of ritual mediating conflict—in other words, to use the general as a matrix with which to de-particularize the particular; and thereby to validate itself. Generalization becomes a self-confirming hypothesis—a classic instance of what Ardener might have characterized as text becoming genre.
While we will not ape the ploys of formal debate, we are entitled to test a little the terms of the motion. While we do not feel that it is incumbent upon us to reject generalization out of hand, we do dispute that generalization is the defining activity or competence of anthropology. Indeed, although we accept that there may be circumstances in which, and audiences to which, generalization might be appropriate, we deny that it is among the more important qualities of the anthropological exercise. Furthermore, we do not regard it as necessary to show by repeated example that generalization can be vacuous. What strikes me as curious is the notion that we should aspire to generality, for generalization is such a dull, such an unambitious, mode of discourse—rather like regarding any performance of a Bruckner symphony as if it were an unvarying reproduction of Bruckner’s notes; or, as in the critic’s version, as if it were merely an expression of the conductor’s reading, ignoring the hundred or so musicians who do the actual bowing and blowing, the plucking and banging.
No wonder orchestral musicians view the maestro through such jaundiced eyes. With what tinge of yellow should we regard our respected colleague who tells us, in a near-orgy of generalization, that Sinhalese abhor individualism as an affront to the cosmic integrity of the State and hierarchy; while Australians regard the State as an offence to the individualism which they venerate, and therefore, drink excessively in celebration of their personal autonomy? 2
Well, if von Karajan can have his Bruckner, we must allow Kapferer his Australia, so long as we are clear that it is a figment of his vivid and ingenious imagination, whatever claims may be made for its ‘ontologies’. We, on this side, prefer to think of Australians rather than of Australia, and see no reason to suppose that they are any more generalizable than we are. Are there groups whose members are less comfortable with generalization about themselves than departments of anthropology? Let me repeat to you an instructive observation, arising out of a consideration of literature on American and Indian kinship: ‘Blanket considerations... involve so much selectivity and systemization by the analyst, that they cannot reflect indigenous thinking to the extent claimed.’ This is a caution I applaud, and was made in an excellent article on the anthropology of kinship by my opponent today, Anthony Good. 3
It may reasonably be objected that generalization does not have to take such crude, all-embracing forms. I agree. What kind or degree of generalization might we then regard as acceptable? To the extent that we think with categories—of gender, age, ethnicity, class, and so forth—we know we cannot eschew typification altogether; thinking and intellectual discourse could not proceed without it. But that is not to say that the proper objective of anthropology is to generalize. I would prefer to say that anthropologists use generalization pragmatically as an essential weapon in their struggle to beat a path through generalities towards some greater sensitivity and enlightenment. In this regard, we can measure ourselves against the politician, the journalist, the advertiser, the survey researcher, whose entire enterprise depends upon their capacity to make the grossest kinds of generalization, and to make them stick. But we may also distinguish ourselves from those scholars who are content to ignore, or to miss, the inconvenient qualification, the exceptions to the rule—those devastating pieces of information which, when once revealed, turn out not to be quite so exceptional after all, and, indeed, to show the general statement to be a travesty. Our own experience surely leaves us in no doubt that the ubiquitous failure of development projects, of urban plans, indeed, of strategic planning of all kinds, often results from the planners’ neglect of, or disdain for, the vital differences among people which their generalizing models obscure. Social scientists, from econometricians to ethnomethodologists, are obsessed with the postulation of pattern, or rules which purportedly govern behaviour, or some contrived regularity, or with testing their unfalsifiable metaphysics of regularity. But what strikes us so forcibly through ethnographic observation are the irregularities among people. If anthropology’s concern is not with complication, with complexity, with differentiation, with non-generalizability, we might just as well retreat to the positivistic pleasures of number-crunching, of social surveys and statistical sampling; take refuge in comfortable, but mindless statements about ‘human nature’; or indulge in the pernicious kinds of formulation that we castigate in less liberal minds than our own as racism, sexism, ageism, and so forth.
Let me make it clear that I am not pleading for a redefinition of the objects of study, from societies to individuals; and far be it for me to wish to bring into disrepute the notion of ‘culture’. But I do insist that we can treat societies, cultures, as barely generalizable aggregates of difference rather than as fictive matrices of uniformity. Plainly, what is at issue here is not simply the highly specific topic of generalization but, rather, the more fundamental, perhaps irresolvable question of the nature of anthropology itself. I am anxious that we should not tumble over that precipice. However, I will put this to you: we can reject generalization as an essential activity without succumbing to the vagaries of ‘post-modernist’ ethnographic representation, if indeed that is a path you wish to avoid, or without leaving material life for the more elusive realms of symbology, if that is the route you dread. It is not a matter of choosing between ‘theoretically hard’ and ‘methodologically soft’ anthropology. It is a matter of recognizing Hobbes’s postulate that societies and cultures are constructs of individuals (not the other way around). If we recognize that, then we have no choice but to slash and burn wherever we encounter the generalizing bush.
The alternative is an anthropology to which people—individuals—are almost purely incidental; indeed, are ignorable if they cannot be generalized into some category or other.
Holy’s Comparative anthropology4has some helpful pointers for us. Holy himself notes the move away from the search for cross-cultural uniformities towards cultural specificities—a progress which, as he and other contributors separately observe, liberates us from the tedious search for variations on a theme; and, by focusing on the relationship within any culture between structure and agency, enables us instead to treat people as culturally creative. There is a major difference between Geertz’s view of culture as ‘webs of significance’ collectively spun by its members; and Parkin’s as the means by which they individually spin ‘endless perspectives’ out of the cultural fleece. 5Generalization is thereby declared redundant—not merely because it is out of fashion or because of political malaise, but because it is intellectually barren, perhaps even bankrupt.
Replying to his invitation to attend this debate, a distinguished anthropologist commented that he was intrigued to learn what ‘concealed solipsisms’ would be employed to oppose this motion. There is no need for me to conceal them: since self-knowledge seems to me quite unattainable, I do not use them. The essential self is frustratingly elusive, the generalizable self insubstantial; considerations which, in themselves, must suggest the absurdity of general statements about societies or their collective constituents. Contemporary concerns with reflexivity do not suggest necessarily a ‘self-indulgence’, or what a respected reviewer recently referred to in Man, perhaps a little intemperately, as ‘the soppy drivel of self-analysis’. The call for an awareness of the anthropological self is not an end in itself (a misrepresentation perpetrated by several writers, such as Friedman) 6but, rather, an injunction to us as ethnographers to recognize that those whom we study are, like ourselves, composites of selves, as complex as we are, as uncomfortable as we are ourselves with generalization about ourselves. The fruitfulness of such sensitivity is superbly illustrated in Wendy James’s account of Uduk personhood in her The listening ebony. 7
What consciousness of the self and of the philosophical problems of personhood should have taught us is that, by failing to extend to the ‘others’ we study a recognition of the personal complexity which we perceive in ourselves, we are generalizing them into a synthetic fiction which is both discredited and discreditable. We fall back too easily on the assumption that in important matters the members of collectivities think alike. That is why we talk blithely about ethnic strategies, about cultural attitudes and values, about how the Azande or the peasants of the Bocage regard witchcraft, about how the Huichol and the Tallensi contemplate their destinies. With what arrogance and insensitivity do we presume to speak about the aspirations, sentiments and sensibilities of tribes, lineages, ethnic groups, sects, or other, even more general categories: pastoralists, hunters, indigenous peoples?
Let me sum up. I do not stand for an anthropology without generalization but, rather, for one which uses it to expose the falsity and superficiality of the general statement. Roger Keesing has recently shown how scratching the surface of Kwaio general statements about ghosts and the Land of the Dead reveals an infinity of contradiction and diversity, a diversity which, he concludes, ‘seems to me to render deeply problematic premises about culture as systems of shared meaning’. 8
Symbols mean different things to different people, different things to the same people at different times. What then can be the status of the generalized claims we make for the most arcane and elusive aspects of social behaviour, wrapped in the mysteries of ritual and myth, cloaked in the fog of kinship ideology? If our aspiration is to reveal and display the genius of those whom we study, rather than our cleverness in inventing them, then we must be bold, ambitious, and look beyond the blandness of the general to the sharpness of the particular—or, at least, know how to treat the general with the very greatest scepticism and caution.
FOR THE MOTION (2)
Anthony Cohen chose in his presentation to focus almost exclusively on stereotyping in formulating his critique of generalization. I agree entirely with his comments on the limitations and undesirable features of cultural stereotyping, but stereotypes are of course not the only kinds of generalization possible in anthropology. I intend to deal with generalization from a rather different, broader point of view. I have been encouraged to be as polemical, unscholarly and controversial as I like, and intend to try and take full advantage of that offer. The motion is that ‘social anthropology is a generalizing science or it is nothing’. I myself wish to claim something slightly different: namely, that if anthropology is not a generalizing science, it is not worth doing.
My main qualification for speaking today is that I have actually done something you will all agree is ‘generalizing science’, namely physical chemistry. Yet the differences between chemistry and anthropology, it seems to me, are not very great. The point is not that social anthropology is more like physical chemistry than you may think—you know as well as I do what anthropology is like—but that physical chemistry is more like social anthropology than you may think.
It is a truism that the crucial distinction between the physical and social sciences lies in their objects of study. Unlike atoms and molecules, people are reflexive: in Weberian terms, they don’t merely behave, they behave meaningfully, they act. Social science is the study of meaningful human behaviour.
Most of us would agree that anthropology is concerned with meaningful behaviour, with action, with actors’ understandings of ‘facts’ rather than ‘facts’ themselves, though this is more true methodologically than theoretically. That is, although we subject much of our information to criticism and re-interpretation by informants, we allow ourselves the option of rejecting these critiques, for sound reasons such as the more systematic nature of our own observations, which transcend the perspective of any one social role.
So far so good: but this argument goes off the rails if it leads to the conclusion that—because of its concern with meaning—anthropology cannot be scientific. For some, indeed, the allegedly ‘non-scientific’ character of anthropology is not merely accepted as a regrettable but inevitable deficiency, but glorified, and made a matter for rejoicing. I find this view unintelligible, and indeed dangerous both for what we teach our students, and for the utility of anthropology outside the academic context. I shall refute it by showing that anthropology is a thoroughly scientific enterprise, if the nature of ‘science’ is properly understood.
First, it stands in a clear relationship to other sciences with regard to its subject matter. Let me illustrate this by means of an admittedly over-simple just-so story, as follows.
There are a number of levels at which the world can be understood. Each is the concern of one of the basic sciences. The nature of the most elementary components of the physical universe is controversial, but for simplicity let us label them sub-atomic particles. Whatever they are, the discipline which studies them is nuclear physics. More broadly, physics generally studies the behaviour of single, monadic particles. When such particles combine to form more complex entities—molecules—these prove to have bulk properties which are more than the sum of their parts. These molecules are the subject matter of chemistry. They combine in turn to form cells, whose properties—most dramatically, the property of life—are again not the mere aggregation of the properties of their component molecules. Cells are the subject matter of biology. They combine into organisms, which have properties and possibilities infinitely greater than mere aggregations of cells. These include the phenomena of mind, studied by psychology. Finally, of course, human organisms combine into societies with attributes such as power, authority, hierarchy, and so on, not present in individual human beings. Such societies also share collective representations—emergent properties whose peculiarities lie at the heart of today’s debate, and which form the most distinctive subject matter of anthropology.
The interrelationships of these basic sciences are clear: each studies those bulk properties which arise when the objects of study of the science preceding them in order of presentation are combined together in distinctive structures. A further methodological principle is that, as the properties studied by each science emerge only at that particular level of structural complexity, they cannot be fully accounted for in terms of properties at any earlier level. That is, they cannot be explained by any form of reductionism. Each of these sciences thus has its own autonomous level of competence, and it should not surprise us, therefore, if it also has its own distinctive methodology. Finally, all these approaches are equally valid ways of understanding events in the real world, though their respective relevance changes drastically according to the nature of the questions asked and the answers sought. Thus, although any human interaction is simultaneously a process of interaction for countless billions of sub-atomic particles, this fact is not particularly relevant if our concern lies with the social implications of what is going on.
So social anthropology stands in a clear relationship to the other basic sciences, because it is concerned with studying phenomena at one clearly discriminable level vis-à-vis those other sciences. This does not mean, of course, that anthropology as presently practised, still less the work of any particular anthropologist, is ipso facto scientific; but it does mean, I suggest, that anthropology has the potential to be scientific.
My next point takes this argument a stage further: social anthropology, I shall show, has already realized its scientific potential. Despite its distinctive subject matter and methodology, anthropology as presently practised fits comfortably under the rubric of ‘science’.
So what do we mean by science? Chambers Dictionary gives the following definitions: ‘knowledge’ (presumably we all agree that anthropology is a form of knowledge); ‘a skilled craft’ (again, we all agree that skill is involved). But presumably the definition closest to what our motion has in mind is the following: ‘knowledge ascertained by observation and experiment, critically tested, systematized and brought under general principles’. With the possible exception of ‘experiment’, if very narrowly defined, it seems undeniable that social anthropology is all these things. Consider them in turn:
(a) observation: participant observation is what we always say is our characteristic method.
(b) critical testing operates at all stages. During fieldwork, there is what Holy and Stuchlik 9call ‘test by praxis’, namely our growing ability to take part in local events, and advance opinions on them which are taken seriously by local people. During writing, our analyses are tested in many ways: against our own data, for internal coherence, and against the ethnographies and analyses of others.
(c) systematized. all ethnography, however ‘reflexive’, ‘autobiographical’ or ‘post-modern’—to mention only a few terms of abuse—is undeniably subject to systematization.
(d) brought under general principles: not all anthropologists set out to do this all the time, but I cannot imagine any justification for anthropology, any reason why it might be worth devoting one’s life to, and trying to teach to others, if it did not hold out the possibility of generalization. If it is in the end merely a means of ‘finding ourselves’ then it is pure self-indulgence and not worth doing.
If we wish to characterize ‘science’ rather more precisely, one of the most satisfactory and influential ways of doing so is that offered by Popper. As you will know, Popper argues that scientific method is founded not upon induction, but upon deduction. Moreover, science is not concerned with ‘proving’—or as he says ‘verifying’—laws once and for all, but only with (temporarily) corroborating them, by showing that certain hypotheses deduced from them account for the observations made so far. There is an asymmetry here, though: one can never finally ‘verify’ a generalization, but one can falsify it. 10
Science as a method applies only to the testing of hypotheses: theory-development itself is wholly non-scientific, and arises through intuition, or genius. Moreover, even the assessment of observations, to decide whether or not they corroborate or falsify the hypotheses in question, is almost never a simple, clear-cut matter, even for the most trivial hypothesis, and has to be resolved by debate within the scientific community. Popper himself, of course, was a methodological individualist, but most of us would wish to see this inter-subjective testing as a pre-eminently social process. The contributions of Polanyi, Kuhn, Ziman and others lend extra force to such a modification of Popper’s account and add further to its general credibility.
All the features Popper describes are explicitly present in social anthropology, where we debate such issues all the time—much more so, in fact, than in chemistry, where the inductive delusion is still widely, if implicitly, held, and where the debate of inter-subjective testing is generally perfunctory. Physical chemistry, I was assured by my chemistry PhD supervisor, was merely colouring in ‘little patches of blue sky’ in the grand design of scientific knowledge. In the fields of debate and questioning, anthropologists behave far more like Popperian scientists than chemists do!Arguments about the significance of evidence, and the inferences legitimately drawn from it, lie at the heart of virtually all anthropological debates. The central process of Popper’s model, falsification, also features clearly in anthropology. If you doubt this, compare our present state of knowledge with that prevailing in the last century. Surely we can say with confidence that—whatever the validity of our present views—certain approaches then adopted, certain assumptions made about specific other cultures, and about ‘others’ in general, were wrong?A familiar, but none the less excellent example is provided by Sahlins’s 11writings on primitive affluence. We do not necessarily have to accept in detail his notion of the ‘Zen road to affluence’, but it is surely undeniable that a necessary precondition for the study of hunter-gatherer behaviour or ideology is the realization that they are in fact not perpetually on the brink of starvation.It is true that anthropologists rarely formulate universal, descriptive and predictive laws, but it would be wrong to regard this as evidence that anthropology is non-scientific, for several reasons:
(a) Most would-be laws in anthropology have ultimately proved to be tautologies: definitions or typologies masquerading as predictions. But precisely the same is true of science: thermodynamics, for example, is a closed logical system in which apparently predictive statements of interrelationship follow automatically from the definitions of such metaphysical notions as ‘heat’ and ‘energy’. Many scientific ‘laws’ are themselves metaphysical statements.
(b) ‘Laws’—even scientific ones—are not merely descriptive and predictive, but morally prescriptive, too. Laws which are primarily of this latter type are found in every society, and occupy a great deal of any fieldworker’s attention. We study such laws, and our conclusions are necessarily different in character from the laws themselves - just as the laws of physics differ from the phenomena which they describe. Social anthropology, you might say, is above the laws!
I have argued that, as a matter of fact, anthropology is a science with regard to both methodology and practice. I shall now argue further that strategically, in our own interests, in the interests of the discipline, but above all in the interests of those we study and with whom we claim such unusual closeness and mystical participation, it is incumbent upon us not only to accept that social anthropology is scientific, but to proclaim from the rooftops that it is.
I am concerned now with anthropology as expertise, with the role of the anthropologist as expert, for example in the field of development. This role is often not particularly glamorous or intellectually exciting, but it is useful. Involvement in such situations may confront anthropologists with uncomfortable moral dilemmas, and certainly it is right for us to question forcefully, where necessary, the underlying assumptions and practices of development. But such dilemmas are inescapable in any practical situation, and it is better for us to be involved than to allow such processes to go on without us. We may at least ameliorate their wilder excesses.
There is a dilettante notion, fashionable in some quarters, that anthropology can have no practical relevance. I disagree: indeed, in the polemical context of debate, I would suggest that such attitudes might justifiably be viewed as almost obscene. Anthropologists above all, precisely because of our claim to specially close, intense personal relationships with those about whom we write, operate under a moral imperative which requires us to involve ourselves in developmental processes initiated from outside but affecting those same people.
The relevance of this for today’s debate is that only the recognizably scientific nature of anthropological expertise lends our advice credibility: the fact that it is indeed based upon ‘observation and experiment, critically tested, systematized and brought under general principles’. If we admit that all we are doing is contemplating our own navels, no one will take us seriously. I am not advocating that we should claim to be doing science merely for strategic reasons, even when we know the claim to be bogus. My contention, to the contrary, is that it is not bogus at all.
The motion is not that anthropology is only generalizing, but that generalizing science is an essential element of the discipline. It allows for the possibility that there is more to anthropology than that, which is indeed so. But personally, I find it both more important and more interesting to learn about other societies than about the emotional responses to them of a single colleague.
Anthropology, then, is a rational, empirical discipline, or it is nothing. It is more than a mere literary genre, and certainly more than psychobabble. As Michael Carrithers wrote recently: ‘it is difficult to see [ethnography] as achieving more than…good writing if it is not grounded in some thought about what is generally true of humans’. 12
If social anthropology is only reflexive, if it is only autobiographical (heaven help us, not even biographical, but autobiographical!), if it is only a form of psycho-therapy for jaded aesthetes, then it is not worth doing.
It may be that as a consequence of doing anthropology we learn more about ourselves, but this is not and should not be its primary purpose. If we wish to do anything to help those we study—concern for whose welfare we parade like stigmata in front of students, readers and the general public—then for both methodological and strategic reasons, we can only do so by regarding our activities as scientific, and convincing others that this is so.
AGAINST THE MOTION (2)
In opposing the motion, I suggest we substitute the word knowledge for ‘generalizing science’. I therefore argue that anthropology is knowledge or it is nothing. The word science is now culture bound, misleading and impoverished. It comes from the classical Greek which means knowledge. By the eighteenth century the meaning of the word was more specific, it included the search for underlying laws.
But even that more precise sense has been debased and confined today. We should consider the current meaning of the word science; it is far from the Enlightenment ideal and even further from the original idea of knowledge. Anthropology risks being unscientific in the original wider sense if it is defined only as science, and as a generalizing one at that. The current meaning of ‘science’ as proffered to social scientists is little more than scientism. It is still contaminated by positivism. Positivism may have been discredited in principle, but it operates in practice. One implication of this motion, you may feel, is that if you oppose it, anthropology is nothing. My arguments offer you more than that, indeed more than scientism. In opposing the motion you do not have to believe that anthropology has no claim to science.
Alternatively, you may think that by opposing the motion you are defining anthropology as mere literature, or—to use a much maligned word—as fiction. It is not my intention to argue that anthropology is fiction. However, in passing, I will clarify a common misunderstanding. The post-modernist use of the term fiction does not mean mere invention, all made up, fairy tale or fantasy. The post-modern definition is closer to ‘social construction’. You may also be misled even about the classical notion of fiction. For example, the nineteenth-century writers of fiction such as Balzac, Tolstoy and Eliot did not sit at empty desks with virgin paper and write from the top of their heads. Balzac conducted painstaking historical and contemporary research into every ethnographic detail of his novels. When setting events in a specific year he made sure the appropriate characters wore the fashion of that season, down to the very shape of the sleeves. Far from inventing characters from thin air, he created them from meticulous first-hand or participant observation. The difference between his characters and individuals for anthropologists is that Balzac’s were composites. Balzac, in La Comédie Humaine, even expanded on scientific theories of his time, believing that humanity could be classified into specific character types. Each type had its own temperament and physiognomy. Balzac’s discredited theories were no more absurd than those of Lombroso, the criminologist, who made explicit claims to be a scientist. This is not to forget some fundamental distinctions between literature and anthropological ethnography. Artistic licence may, indeed must, encourage creative transformations.
Social realism, the seemingly exact replication of life in art, has its limitations. Eisenstein, the Russian film maker, in wanting to convey the moment and meaning of the October Revolution, incorporated imagery from symbolist poetry in defiance of his Stalinist patrons. Professed realism may prove less real in its outcome. By contrast, anthropologists aim at minimum, if not maximum, realism. They make their own distortions for ethical, political and technical reasons. Names are changed, anecdotes edited and disguised. A village may be a composite. Such distortions are not sufficient to call an ethnography literary fiction. In contrast to any novel, we would discredit a field monograph by an anthropologist who had only pretended to have been there. But as we look closer at the development from first ideas to fieldwork, note-taking and monograph, there are value judgements and choices. The necessary selectivity, the omissions, the accretions and theoretical paradigms lead us to acknowledge that the monograph is also a product and construct of the anthropologist’s academic and historical time. The same could be said of Darwin’s The Origin of Species, but we do not say that biology ceases to be a science. In unravelling the conscious or unconscious constructions in a monograph, we do not simply falsify and discredit the monograph as some might fear, we acquire knowledge into how a representative of one culture, usually the dominant imperial one, represents and explains another. The very selectivity of the content can also inform the reader as to what the people themselves chose to reveal to an outsider. To read a monograph as a historical construct enlarges our knowledge and raises the potential in a truly scientific enterprise.
Part of the selectivity of the monograph may depend on the age and sex of the anthropologist, as well as the theoretical paradigm. 13It is obvious that one individual cannot hope to grasp the totality. The monograph gives us a very specific knowledge, but one with which others can engage. The specificity of the fieldworker should not discredit the knowledge acquired. Yet if we followed the positivists’ tenets to the letter this knowledge would be devalued. One such tenet is the interchangeability of observers. Here anthropology has too readily allowed itself to be intimidated by scientism. The specificity of the fieldworker should be explored, not repressed. The extent to which observers are not always interchangeable should be a subject for scientific study.
Traditionally, social anthropology has permitted what it believes to be science to colour its own claims to science. The danger is that scientism, not science as knowledge, sets the agenda. We listen to scientists through keyholes. In practice, discoveries in the so-called hard sciences often occur in ways which are more familiar to the humanities. The difference is that social scientists pretend in print that their discoveries do not happen that way, whereas scientists and mathematicians have the hegemonic power to talk of their experiences without being discredited. It is said that the inspiration for relativity theory came to Einstein when he imagined a man moving so fast in space he could not see his reflection in the mirror he was holding. Thus the ideas were crystallized through the power of metaphor, and a Lacanian one at that. Contrast this imaginative freedom in ‘real’ science with the straitjacket of positivism. It retains its grip, especially on those who would define anthropology as a science or nothing.
One of the tenets of positivism is the unity of scientific method amidst the diversity of subject matter. In another, the natural sciences, more especially mathematical physics, set a methodological ideal for all the sciences, including the social sciences and the humanities. 14Popper and the neo-Popperians, Jarvie and Horton 15would hold with this, despite their alleged break with positivism. These neo-positivists likewise reduce epistemology to methodology. They assert that one objective method is suitable for both the natural and the social sciences. The concept of theory as method reduces a traditional role of theory as critique to the criticism of hypotheses.’ 16The Popperian concept of theory envisages the participant observer as laboratory technician, testing hypotheses and recording facts.
You may ridicule this analogy. The majority of social anthropologists might be reluctant to define themselves as technicians, positivist or Popperian. Why then do we still use the positivist’s language as we go about our work? It is a web from which we should free ourselves. Too often, postgraduates are asked ‘What hypotheses are you testing?’ This is the kiss of death. Yet it is printed in research grant application forms, which postgraduates and we have to address. We continue to collude in the use of the word ‘training’, which reduces our research to a set of techniques. It sets the theoretical agenda. We react rather than initiate. Why do we not ask of others, ‘Why do you bother with hypotheses?’ Hypotheses limit the scope of knowledge. We know our research thrives on what we cannot begin to hypothesize. Agar has responded with ‘the funnel method’. 17This is an inappropriately mechanical metaphor for the creative experience of research.
Anthropologists still struggle defensively with the positivistic ideal of the objective observer, when our unique approach calls for a confrontation with self-awareness. Here reflexivity has magnificent potential in fieldwork practice away from the desk or laboratory bench. A limited science regards the interests of an ego with personal and cultural history as potentially contaminating. We should see it instead as contributing to the meaning of scientific practice. 18Participant observation could be pondered, interpreted and explained. We hesitate to explore how it works, because we are intimidated by the phantom of detachment.
Let us examine the tradition that too often gives us the exact and natural sciences to mimic. Granted that we are still unflustered by any quantification ideal, the metaphor of precise measurement still leaks into our fictive hypotheses. Again, the practitioners whom we are supposed to imitate do not act in the way we imagine. A leading mathematician was asked how he went about solving difficult problems. His answer? Think vaguely. Would we dare write that in a research proposal?
The positivist precision is only one form, for we and those in the humanities also have precise standards. Today we appreciate the metaphorical and narrative style of Malinowski. It was only incidentally learned that part of that productive process came from reading Conrad and from a self-confessed ‘shameful’ obsession with novels in the field. We have rejected Malinowski’s scientific theory of basic needs, meanwhile the ungeneralized anecdotes leave us knowledge. Science, in contrast, defines the anecdotal as pejorative.
A final characteristic of scientific explanation is that it is causal and subsumes individual cases under assumed general laws of nature; something Evans-Pritchard so severely contested. By contrast, a Marxist such as Maurice Godelier would sympathize and, were he here, would probably support the motion. But the Godeliers among you should recognize that in doing so, you are also voting with and for scientism. For a few, this would be no dilemma. A contributor to Critique of Anthropology has called for the end of fieldwork. 19Once freed from what he called this ‘straitjacket’, the discipline could become truly ‘scientific’. Thus the detachment of the observer from human contamination is accomplished at the expense of knowledge and in the name of science. For a more convincing Marxist anthropology, I would suggest the word scientific be replaced by materialist.
To conclude, we have been bewitched, bothered and bewildered by a limited definition of science, even within its Western history. Until we can redefine and extend the meaning of science, we should vote for anthropology as knowledge, global and unconfined.