Epilogue to The Disastrous Ordeal of 1987

By David Graeber, 22 January 1996

EPILOGUE

It might be useful to close with some words on why I chose to write this book the way I did. In purely formal terms, it is a rather unusual ethnography. Its style is at times almost novelistic; at others, it shifts into much more conventional modes of ethnographic or historical writing. The same characters who appear in one part of the text as actors often reappear in others as narrators or analysts, providing (often critical) commentary about customs, local issues, and about each other. When I began writing the book, I had not entirely worked out most of the theoretical ideas about politics and narrative that now appear in it; so I cannot really say that I wrote it the way I did because of them. But I did want to convey the sense I had of the people I knew in Betafo as both actors in history, and as themselves historians. According to the very broad set of definitions I did work out in the course of writing, to represent them in this way is also to represent them as political beings. It is, I have argued, in so far as we all act in, recount, interpret and criticize our social worlds that we are all political beings of one sort or another.(1)

Throughout the book, I have tried, whenever possible, to emphasize such areas of common ground. In fact, if there was one impulse--one might even say one moral imperative--that drove me from the very beginning, it was a desire to explode some of the sense of artificial distance that so many ethnographies create between author, audience, and the people who are being studied. I wanted the reader to be able to think of the inhabitants of Betafo as people they could at least imagine meeting and who, under the right set of circumstances, they might actually get to know. If nothing else, I have tried at least in small ways to always emphasize how--cultural differences notwithstanding--we do inhabit the same world, and ultimately the same history and the same moral universe; or, if one wants to define history in a more culturally specific fashion, then at the very least, that we all could be sharing one.

At the end of chapter 11, I suggested that one way anthropologists could begin to undermine this sense of distance would be to look at what we are doing as more akin to history than to what we would normally consider science. Now, in saying this, I didn't really mean to weigh in to the sporadic debate in anthropology about whether the discipline should define itself as a science or a humanity. To me at least, it seems a silly argument: after all, people who are mainly interested in, say, problems of nutrition or verb structure are obviously going to be relying on a different set of methods than people who are mainly interested in historical consciousness; as long as we all happen to be tumbled together in the same departments, it seems only reasonable to allow that our discipline is a hodgepodge and leave it at that. But the debate does raise some interesting issues. Why exactly is it, for instance, that history is considered one of the Humanities, and not a Social Science? Obviously there are historical reasons--there were people who considered themselves historians long before there were ones who considered themselves social scientists (or for that matter natural scientists). But if it has remained among the Humanities, in the company of the study of literature, art, and philosophy and not that of sociology or political science, I suspect it is ultimately because of some sense that science deals with regularities, if not with 'laws', then at the very least with things that are to some degree predictable, and that history tends to focus on the very opposite, on the irregular and unpredictable, on events that could no more have been predicted, before they happened, than the production of a novel or a work of art.

For some, I will allow, this might seem a rather old-fashioned view of history. Certainly, not all modern historians feel their discipline should even be among the humanities; there are many proponents of a "science of history"--one which can make predictions. In ways the debate within history parallels the one in anthropology. Much of the literature about the nature of narrative which I made use of in the introduction to chapter 7, in fact, emerges from just such a debate (Rosaldo 1989:127-143 provides a useful summary)--in which a set of historians and philosophers of history, having been told that no one had any reason to take them seriously as long as they were simply telling stories and not coming up with any generalizable laws, ended up formulating a defense of 'narrative understanding' as an alternate, and perfectly legitimate, way of knowing. I am, as the reader might imagine, sympathetic to their position. In fact, I would take the argument much further. It seems to me that, at least in anthropology, it is this very concern with science, laws, and regularities that has been responsible for creating the sense of distance I have been trying so hard to efface; it is, paradoxically enough, the desire to seem objective that has been largely responsible for creating the impression that the people we study are some exotic, alien, ultimately unknowable Other.

Let me provide a few examples of what I mean by this.

European documents concerning Madagascar written before the French conquest are remarkably different in tone than those written afterwards. I remember being quite struck by this while doing research in the Malagasy archives. It seemed as if in 1895, the whole character of the country changed. The Madagascar one reads about in 19th century documents documents was a place full of recognizable individuals: politicians, princesses, humble rural pastors or wandering sorcerers, bandits, generals, Christian martyrs. Documents by missionaries, European agents and travelers were all the same in this respect; even political dispatches tended to taken up with speculation about the motivations, affinities, and likely plans of government ministers or potential revolutionaries. The authors were full of all sorts of biases, and most of the portraits are pretty two-dimensional, but at least they were usually making a sincere effort to understand the motives of the people about whom they were writing, for the simple reason that they had to. After all, they were visitors in an independent country, and these were people with some power to affect their lives. Almost the moment Madagascar lost its independence, the human beings also disappear. Documents from the colonial period consist either of vague, descriptive generalities, or (even more) of tedious accounts of administration, along with scientific dispatches and reports.

Margaret Weiner (1995) has recently noted much the same transformation in comparing records from before the Dutch conquest of southern Bali in 1908 to those which came after it; after conquest, accounts of personalities and dramatic events are immediately replaced by bureaucratic "discussions of finance, agricultural production, construction, and public health. It was," she writes, "as if once a region was brought under colonial domination, nothing happened there any longer... The colonial state produced knowledge mainly in the form of statistics and regularities" (Weiner 1995:90).

One reason why individuals disappear from colonial documents is, clearly, because the authors were no longer obliged to take account of them; one of the first things a colonial regime tends to do is to create a political climate in which no single inhabitant of the country is in a position to do anything which could have much of an effect on them. But it was also because they conceived of what they were doing, their mode of rule, in very scientific terms. Hence the 'statistics and regularities'. Colonial governments saw themselves as applying techniques of scientific administration, which could bring the country's economy and society as much as possible under complete, predictable control, and in doing so establish the very parameters within which meaningful action was possible.

But--at least in Madagascar--there was another side to this story. It was also precisely at the moment when the country had been conquered, and these rational-bureaucratic techniques of administration were being put in place, that the new administrators began waxing poetic (in their unofficial writings) about something they called the "Malagasy soul". This "Malagasy soul" soon became a stock theme of French writing on the island. It was represented as the sign of a profoudly alien mentality, full of quirky passions and dreamy fantasies, ultimately beyond the grasp of the understanding of a simple Westerner. As Antoine Bouillon (1981) has pointed out, no one had ever talked this way when Madagascar was independent, and foreign visitors still had to deal with individual Malagasy actors on anything like equal terms.

This is a useful example, I think, because it's so obvious what's going on here. The "Malagasy soul"--in so far as it was anything more than projection--was a mere by-product, a confused amalgam of everything that fell outside the extremely narrow parameters set by the authors' own bureaucratic machinery, or the rationalistic regimes which they now had the power to impose.

Modern anthropology of course took shape mainly within the British and French colonial empires as well. And it too considered itself a scientific enterprise.

This was the age of Structural Functionalism, and one of the main things that made Structural Functionalist anthropology, in the eyes of its practitioners, was the fact that it was concerned primarily with "norms", or regularities. What this meant in practice was that what ethnographers described, and theorists discussed, was almost exclusively those aspects of social life which were predictable, repetitive: the human life cycle, with its age grades and rites of passage, the domestic cycle, ritual cycles, yearly rounds... Even succession to political offices was always treated as an essentially regular processes, which ideally, should always work themselves out in the same way. In so far as individuals and unique events appeared in ethnographies written at this time, they would usually take the form of case studies meant to illustrate more general processes. Here and there, there were efforts to try to find some way of talking about individual projects and intentions (names like Max Gluckman and Victor Turner come most immediately to mind) but it was with the underlying assumption, one could almost call it faith, that individual actors were ultimately irrelevant, that whatever their immediate intentions, they would somehow end up reproducing the same cyclic structure over and over again.

Of course, it was easier to think of such people as living outside of history because, for the most part, they were people living under foreign military occupation, with no political rights. But as time went on, Western observers developed an increasing tendency to confuse causes with effects. Rather than the absence of history being an effect of the way the authors chose to describe these societies, it was because of something profoundly strange about the societies themselves. They were societies that had rejected history. 'Cold cultures' (Levi-Strauss 1966), exotic societies locked in a primal, mythic consciousness. And what did this mythic consciousness consist of? Regularities. Eternal repetition. The faith that everything comes in unchanging cycles, a 'traditionalist' philosophy that actively rejects history, personal idiosyncrasies, the future, and cumulative change in the name of timeless archetypes and the "eternal return" (e.g., Eliade 195-).

It is probably only fair to point out such doctrines tended to put forward most enthusiastically by people who had only read ethnographies, not ones who had written them. But still: notice what is happening. The very attitude which Western observers adopt in the name of science ends up being projected onto those they observe; except there, instead of making them seem scientists, it makes them seen mystical, poetic; strange, profoundly different sorts of human being.

Since the dissolution of colonial empires, anthropologists have rediscovered history. But something of the old attitude remains. There is still a sense that, in order to be considered objective, one must deny certain aspects of the subjectivity of those one studies. Few ethnographies even attain the level of personal engagement one senses in some of most interesting dispatches and reports to be found in precolonial European archives. Indeed, I suspect it is just this sort of denial which is ultimately responsible for the fact that critics can still write of anthropology being basically about drawing the boundaries between an "us" and "them" (Trinh Minh-Ha 1989), to speak as if its fundamental business has always, and must necessarily be, to describe some deeply alien creature--usually referred to as "the Other"--so different from the anthropologist and her audience that anything one says about them is likely to be a mere projection of one's own self (see Said 1982; JanMohammed 1985:59, Spivak 1988, Trouillot 1991, for but a handful of examples.)(2) Again, anthropologists themselves do not often talk this way; it is mainly those who have merely read our books who talk about "the Other". I myself can't think of anyone I know who has actually lived and worked for any length of time with people of a profoundly different culture who left with the impression that they were--to take one extreme formulation (Todorov 1982:3)--"so foreign that they leave me reluctant to admit they belong to the same species as my own".

Obviously, then, by the time of writing, something is falling out. In the field, anthropologists have no trouble recognizing the people we work with as fellow human beings. But somehow, whatever it was that made them so recognizable is not coming through in our descriptions.

Perhaps this is not so surprising, considering what some of these points of recognition are. In my own case, for instance, the most obvious thing which made it impossible to think of the people I met in Madagascar as being profoundly different sorts of human being was the fact that they were all so different from each other. And not only that: the ways they were different from each other seemed pretty much the same as the ways people were different from each other anywhere else I'd ever been. From the moment I started having any sort of prolonged social interaction with people, I found myself sizing them up as individuals: 'Person A seems to be basically well meaning, but she's incredibly self-involved, the sort of person who always feels her life is in a total crisis, tottering on the brink of tragedy; Person B, something dishonest about that guy, like he's always trying to figure out the angles, wouldn't trust him further than I could throw him; Person C: optimistic, playful but probably extremely impractical; Person D: an insensitive, loudmouth jerk...' It is not hard to see why such assessments tend to get left out of ethnographies. Even apart from the last one (anthropologists have a particularly hard time admitting they could have possibly met anyone in the field who they didn't like), it all seems hopelessly subjective. Even a very brief list such as I have just presented would probably inspire some critic to demand to know why they should not think I was simply projecting my own English language categories where they were entirely inappropriate.

It would not be a particularly fair criticism; first of all, because such assessments often involved impressions one could not remotely put into words--and often even the words one used were obviously inadequate, just makeshift approximations that stood in for a much more intuitive sense of what someone seemed to be about. Anyway, the more I got used to using Malagasy the more I started substituting Malagasy words for English, without ever feeling I was crossing any great divide.(3) But more important, I think, that criticism gives the sense that what one is doing is some kind of abstract parlor game, placing familiar categories on unfamiliar objects, and this is utterly untrue. Making such assessments is no game. It is an absolutely inevitable and necessary feature of human interaction. You have to do it because you have to have some idea how people are likely to behave--because their behavior, their actions, are likely to have effects on you, or at the very least have effects on your friends, or people that you like or care about. Often they are based on very immediate concerns, like: what would it be like to spend five or six hours stuck in the back of a truck with this person? Would they at least be interesting to talk to? Would they want to talk to me at all? What's the chance they would spend the time trying to convert me to their religion, or to seduce me, or get drunk and throw up on me (or all three?) For this reason, too, they are also constantly being tested against reality. One often gets it wrong--sometimes, disastrously wrong. Often, too--and this is especially true at first--one misreads the cues because of cultural differences. But sometimes it is just because it is in the nature of such assessments that they are often wrong. This is an art and not a science, and some people are distinctly better at it than others.(4) But it is also worth remembering that even after years of being tested by daily interaction, no one's knowledge of anyone else can ever be quite complete or accurate; it always remains something of an approximation; people will always retain their capability to surprise you.

It seems to me that one of the prime reasons such assessments get left out of ethnographies--even most self-consciously experimental, 'postmodern' ones--is simply because making them implies a recognition that these are people who have--or have had-- some power to affect the ethnographer's life. Obviously, people who are living together, engaged in common projects, or even just in the habit of engaging in conversation, are going to have some kind of effect on one another. But for some reason, ethnographers tend to find the reciprocal aspects of such relations embarrassing. One is allowed to meditate guiltily about the possible ill effects of one's own actions, for instance, but in writing, the other side of the picture tends to get swept away. In the end, the effect is to create a kind of invisible wall, which seems to prevent the people in the book from having any historical agency, any ability to have an effect on the ethnographer and her world,(5) without, however, preventing influence to flow the other way around. By blotting out the traces of character, it effaces even the impressions which recognition of a capacity to affect others' lives will always, necessarily leave behind.

In part, this is probably just an effect of the conditions under which people write, which are generally very different than the conditions under which they conducted their research. Most ethnographers write their books safely tucked away in universities far away from the people they are writing about. By then, those people are usually no longer in a position to do anything that will have an immediate affect on them or anyone close to them-- although the reverse is not necessarily the case. Most ethnographers do spend a great deal of time fretting over the possible effects their writing might have--as I, for example, have worried endlessly over whether I am betraying my friend Armand by publishing information that indicates his ancestors were slaves. It is hard to imagine anything Armand might do, now, that would have any real effect on me. It would hardly be surprising if that reality (which is, after all, the encompassing reality) often tends to be projected back into the way we write about the very different experience of fieldwork.

A small anti-relativist diatribe

I think it's important to consider this possibility because it suggests how often the invisible walls that appear in our texts are really only made possible by the existence of other walls that are perfectly visible--in this case, of a very large and elaborate apparatus of exclusion which involves such things as international treaties, border guards, the price of airplane tickets, and the IMF. Politics does not take place primarily within texts, though one might not know it from some of the more abstract debates about the "politics of representation".

Consider for example the doctrine of moral relativism. By this I mean the doctrine that, starting from the (entirely reasonable) premise that one cannot fully understand any action except in the context of the actor's cultural universe, concludes that as a consequence, no one has the right to stand in judgment over any action committed by someone with a fundamentally different world view. Now it seems to me this is a doctrine that could only really emerge as a product of imperialism. It is a doctrine that would never be produced except by members of an elite population whose dominance over the world was so complete and so reliable that they could live their lives in full confidence that no one with a fundamentally different world view would ever be in a position of power over them. When you find someone arguing that no Westerner has the right to find fault in, say, the cultural presumption that an appropriate response to grief at the death of a close member of one's family is to waylay and kill some random stranger, they can only do so because of the existence of complex and very efficient systems of control, involving armies, police, passports, airport security, immigration laws, and structures of economic inequality, which make it almost inconceivable that anyone who felt that this was an appropriate response to grief would ever end up living in their neighborhood, or be in striking distance of their children.(6) Pretenses to some kind of moral superiority, based on their unwillingness to morally condemn 'the Other', it seems to me, are particularly obnoxious in so far as they are often entirely underpinned by tacit support for real walls to shut real other people out. And by refusing to consider someone as a moral person, one provides a perfect justification to continue to exclude them.

What I am ultimately getting at is that the very least an ethnographer (or anyone else) owes to people they write about is to represent them in such a way that the reader can recognize them as human beings who (as I said earlier) they might not know, but they could know; as people who have at least the potential to inhabit the same moral universe as she. It means recognizing them as people with the capacity to make history. If doing so means that we will have to abandon the pretense of doing some sort of science, then I would say that kind of science is probably not worth saving anyway.(7)

Not that the mere act of writing this way is itself going to solve all that many problems. No one was ever liberated inside a text. Certainly not within a text they did not write. Recognizing the fact that people have the capacities to be historical agents does not itself make it any easier for them to go out and make history. People do not live in texts. Most of the ones who appear in this book, for example, live in Madagascar. There many had, when I knew them managed to wrest a remarkable degree of autonomy for themselves, but are also very poor, had little access to any worldwide networks of influence and communication, in fact, next to no means to affect anything that happened outside of Madagascar. Nor has their situation changed dramatically in the intervening five years or so. Writing about them with more sensitivity is not likely to do much to improve this situation; though it certainly wouldn't hurt.

The capacity to make history

My argument, so far, has much in common with an argument recently developed by S. P. Mohanty (1989) about the political implications of relativism. Mohanty too argues that adopting an extreme relativist position would be politically disastrous: what basis would we have to criticize the structures of power in the world, unless we at least admit that everyone in the world share certain things in common? At the very least, he suggests, we have to recognize that we by now all inhabit a common history (not a series of separate, culturally bounded "histories"), and that we all share a "capacity for self-aware historical agency" (1989:74), which makes us capable of participating in it. All this seems eminently reasonable to me. But one might still reasonably ask: what precisely does this capacity for historical agency consist of? What is 'historical agency', anyway?

The terms I've been laying out over the course of the last several pages suggest at least one possible answer. 'Historical actions,' one might say, 'are actions which could not have been predicted before they happened.' Or, if that is too simple, then: 'actions considered memorable afterwards because they could not have been predicted beforehand'.(8) History, then, is the record of those actions which are not simply cyclical, repetitive, or inevitable.

As with my definition of political action in chapter 7, I am trying to be intentionally provocative--ignoring almost everything that's already been written on the subject, and proposing an alternative so simple that it might even be considered simplistic. Some readers might object that the definitions I propose for both 'politics' and 'history' are so broad that they threaten to make the terms almost meaningless--leaving no way to distinguish a spat between sisters or some other family quarrel and a revolution or a civil war. Perhaps. But this kind of breadth has advantages as well. If politics (and history) is something intrinsic to the nature of social life, even of ordinary, daily interaction, if it is something which everyone is always doing, not just the powerful; then it is possible to see engaging in politics or making history as something which does not necessarily have to involve preventing anyone else from doing so. In other words, rather than assuming that power and exclusion are intrinsic to the very nature of politics, it allows one to at least imagine a politics and a history that could still be going on without them.

These definitions have other implications as well. If it is really true (as Mohanty suggests) that what makes us human is above all our capacity to make history, and if history consists of actions that could not have been predicted beforehand, then that would mean that the fundamental measure of our humanity lies in what we cannot know about each other. To recognize another person as human would then be to recognize the limits of one's possible knowledge of them. Their humanity is inseparable from their capacity to surprise us.

In a way, this fits quite well with what I was saying about why anthropologists are able to recognize they are not dealing with a fundamentally different sort of being. The constant process of assessing other peoples' characters, which I suggested is an inevitable feature of any relation between people, are so many innumerable imperfect ways of approximating something that ultimately can not be known: how exactly that person is likely to behave. (Character, noted Aristotle, emerges from action.) But this is why for all they are necessarily partial, flawed--like bits of cloth pasted over something that's invisible--they nonetheless seem to convey such an immediate sense of common humanity.

Writing

At this point I can return briefly to the question of style. This book contains a number of little dialogues, ranging from simple two-way dialogues between myself and one other person, and scenes in which sometimes as many as four or five different people discuss some issue with each other as I simply sat and let the tape recorder run. In translating these dialogues, I did my best to convey something of the speaker's personalities, tone, and general attitude-- but usually, provided very little in the way of setting, context, how these people had come to be talking around my tape recorder. Dialogues like this are slightly unusual in modern ethnographic writing, but hardly something new. Still, when I showed such texts to fellow anthropologists, they would often be slightly dubious. "Ah, but how can you know what these people were really thinking", or "how can you know everything that's really going on during such a conversation?" It was said in such a way as to imply that there was something very inadequate about such information if one could not. At first the reaction took me quite aback. It had never occurred to me one would even be able to know the full motivations of all the speakers in any conversation. It was not even clear what it would mean to. And even if one had some way of knowing that one of the participants in, say, a transcribed conversation about Vazimba was leaving out some embarassing detail, or was desperately trying to impress a woman across the room he had fallen in love with--or for that matter was wondering about what kind of price he would get for my tape recorder if he stole it--what would be the point of including this information? It would not have much bearing on the subject of Vazimba. I had assumed that the very act of reproducing the conversation in relatively colloquial language would be enough to convey to the reader a sense that none of these voices are absolutely authoritative, that everything said is at least a little incomplete, slanted, and subjective--which is all any further details would have demonstrated anyhow.

James Clifford (1980, 1986, 1988) has insisted that ethnographic knowledge is always, by its nature incomplete and partial; and I will agree that it is a bit disturbing that this point ever had to be made, that it was not always considered self-evident. Since Clifford's work, and the "crisis of representation" which shook anthropology in the 1980s, anthropologists have become used to thinking critically about the process by which ethnographic knowledge is constructed. This--along with an (equally postmodern) concern with the dynamics of social strategies and maneuvering--is of course the real context for my friends' reaction to those dialogues, the urge to immediately start reflecting over the political dynamics that probably lay behind them. Anyone who has read this book will be well aware that I myself am not entirely unconcerned with such questions. But what was really striking was the urge to understand everything, the sense that simply reproducing what was said, with a few telling details to give context, was somehow deceptive. To write in a way which takes for granted that one would never be able to reveal everything, that art is the art of selecting details, is perhaps the hallmark of a literary sensibility. But at moments like this the postmodern sensibility--for all it draws its inspiration from literary theory--seems to move in precisely the opposite direction, and ends up slipping into a kind of perverse, extreme scientism: as if it were only if one could know that precisely what everyone was thinking, every hidden strategy at play, that we could have real knowledge.(9) Not surprising then that many conclude that real knowledge is impossible.

To me the issue is not whether this sort of knowledge is possible (it obviously isn't) but why anyone would even want it. Would anyone really want to live in a world where it was possible to have this kind of total and encompassing knowledge of another human being? By the definitions I've been developing in particular, it would be the ultimate dehumanization.

My own argument is that rather than seeing the limits of our knowledge as a problem, it would be much better to see it as the best basis on which to build a broader sense of human commonality. If nothing else, it would not be an ethnocentric one, prone to the usual criticisms that it was nothing but a projection of a very particular Western idea of the rational Subject. In fact, I doubt it would be possible to find another cultural tradition that even entertains the fantasy that it might be possible to have such comprehensive knowledge of others.

To say this is not to claim that people are not products of their cultures, of social and historical forces beyond their understanding, only to say that they are not entirely so--and also, that it is this fact that ensures that such forces are not entirely beyond our understanding, they are not unknowable in the same ultimate sense as people are. Let me give an example of what I mean. The reader might recall how, in chapter 5, I asked Chantal why it was that people always seemed to cry at a famadihana, even though it was supposed to be a happy occasion, and she asked me how I would feel if someone put my father's decomposing body on my lap. Here was a Malagasy woman, and one who knew me rather well, reduced for the moment to wondering whether I might not really be some sort of bizarre, alien Other after all, simply by the thought that I might not be emotionally traumatized by such an experience. It was not the only time I had seen Malagasy women discuss whether non- Malagasy really had the same depth of emotions as they did, whether they really loved and cared for each other and their families with the same intensity, whether (by implication) they were really quite as human. It was a familiar prejudice, and I had apparently touched it off in her. But the same people--Chantal included--were in other contexts equally capable of holding out the fact that Malagasy people did, in fact, place decomposing bodies on each other's laps as part of the essence of what made them Malagasy (and me, who had never done this, a profoundly different sort of human being). Both-- identities based on the universality of certain emotional reactions, and those based on cultural traditions--are equally constructed. I am not suggesting that we should base our sense of what is fundamentally human on one and not the other. What I am suggesting is that we should base it on whatever it is--that invisible point--that is capable of pivoting back and forth between the two. On whatever it is (in this case marked by the name 'Chantal') that is able to try on such identities, then hold them out and look at them from more of a distance, in the process of trying to define herself by defining her relation to other human beings. It is also what makes it possible for us to create culture--which is something everyone is always doing, since culture is nothing more than the process of its own creation--and which makes it possible to use what we know of how that process of creation tends to work in order to understand cultures we have not, ourselves personally, had a role in creating. Which is what makes it possible for there to be disciplines like anthropology. I might never be able to completely understand Chantal--any more than I will ever be able to completely understand anyone else--but that which makes it impossible for either of us to completely understand the other is also what makes us both capable of sitting down together and trying to make some generalizations about our respective societies and cultures.(10)

I have already noted the Malagasy tendency to represent the source of human intentions and agency--spirits, ghosts, the soul--as something hidden, invisible, hence which cannot be ultimately known.11 This is actually a very common way of representing things. As I have noted elsewhere (Graeber 1996), it appears all over the world in conceptions of that aspect of the person which ethnographers most often refer to, in English, as "the soul"--what Tyler called the "life soul"--the hidden seat of human intentionality which gives us the capacity to act. Even where there is no explicit metaphysical theory, people do take it for granted that one cannot ever really know what another person is likely to do; and usually, that it is from this unknowable place--in the heart, the head, the throat, the liver, wherever one happens to place it--that actions, ideas, new unpredictable things emerge.

Edmund Leach (1982:108) once suggested that what unites all human beings is not that they are in possession of an immortal soul, but they are capable of imagining that they are. Perhaps (aside from the part about "immortality") these are really not such altogether different things.

Footnotes

1 The descriptive approach I've employed, which weaves constantly back and forth between my own accounts and reconstructions, and those of my informants, plays into this as well; whatever one might think of it as an ethnographic style, it does seem appropriate for a book that is largely about the construction and circulation of narratives.

2 It is interesting to note that the problem of "knowing the Other" was originally asked not by anthropologists or their critics but by European philosophers, and not about people of other cultures but about anyone at all. The problem is really the legacy of Decartes, who left Western philosophy with assumptions so utterly, radically individualistic that the very existence of other people--let alone the possibility of knowing anything about them--became profoundly problematic. Any number of philosophers contributed their own proposed solutions to this problem of "the Other" (Theunissen 1986), attempting to prove why it was indeed possible to know that anyone else really existed, engaged in thought, and so on. Sartre, for example, dedicated a large part of Being and Nothingness (1954), in which he laid down his basic sociological theory, to just this problem. The term "Other", in fact, appears to have passed into social criticism largely through Sartre's intellectual allies and associates. Simone de Beauvoir (XXX), for example, argued in The Second Sex that men, in a male-dominated society, tend to define women as a kind of fundamental or perennial Other, which makes it very difficult for women to act as subjects in their own right; Franz Fanon, in Black Skin, White Masks (1956), made a rather similar argument about effects of colonialism. Now, both de Beauvoir and Fanon were mainly concerned with what all this means for the victim, but their work opened the way to taking a terminology originally developed for describing relations between individuals and adopt it as a way to describe the political relations between groups. Hence in a lot of contemporary critical writing one finds the old problem of 'knowing the Other' transposed, with an abstract being called "Europe", "the West", "the European" (or even "Europe as Subject") striving to define itself in relation to an "Other" which seems to include anyone or everyone else. And while philosophers were usually willing to admit that in the end, one could probably know something about other people, on this level, it usually turns out that you can't: the Other always turns out, through various projective mechanisms (usually borrowed from Hegel or psychoanalysis, or both) to be nothing but a shadow-image of one's own self.

In actual practice, when someone accuses an anthropologist-- or all anthropologists--of reducing the people they describe to an "Other", what they mean is not quite so grandiose. It is much more about a certain way of writing. Any author has to assume a certain degree of common ground, shared with her audience--certain assumptions about human life and motivations, what does not need to be explained. When this is combined with an assumption that the ways or attitudes of certain other people cannot be treated this way, that needs to be explained or otherwise accounted for, then this is what can be called The Other. Of course, stated this way, there is no reason that this has to reflect a relation of dominance: but it's also easy to see how it usually will. A woman for instance can write a book for other women, to help them to better understand male psychology, but such a book would be considered a genre book, of specialized interest and not generic knowledge. The universal perspective is, at least in writing, usually assumed to be that of the dominant group. In this sense speaking of "othering" definitely does address a legitimate problem, though rather, it seems to me, in the way of putting a sledgehammer to it.

3 I still occasionally find myself using Malagasy terms in sizing up people here in America.

4 The loudmouth jerk might turn out to simply have had a strange way of reacting to unfamiliar situations involving foreigners (then again, it might turn out he really was a loudmouth jerk).

5 Aside of course from conveying information.

6 Actually the example in the back of my mind here is that of certain critiques I have heard of Renato Rosaldo's treatment of Ilongot headhunting; but I thought best to leave out the term 'headhunting' in order to emphasize that it is not the actual cutting off of heads, but murder, which is objectionable; what one does with the body afterwards--cutting off the person's head, or for that matter, eating it--seems to me a mere question of aesthetics (unless, for instance, such acts are done with the intention of terrorizing or traumatizing someone who is still alive). Here, a relativist would be perfectly right to say there is no difference between this and any other killing.

I should probably also add that, given the prior fact of imperialism, there is no doubt that such relativism can have ameliorative effects (as the work of Franz Boas, for instance, undoubtedly did). But this is a different point.

7 And of course, many would make the argument it doesn't mean that at all.

8 My definition is in part inspired by a definition of history proposed by John Comaroff: that "history is the conceptual space, the time of human experience, in which social scientific knowledge-- and most of all, prediction--is proven wrong" (in Comaroff and Stern 1994:35). But my formulation is sufficiently different that I doubt he would want to be identified with it.

9 See Comaroff and Comaroff 1990:8-10 for an analoguous critique.

10 The doctrine I am describing then is almost exactly the opposite of the way the notion of an unknowable Other is usually invoked. I note too that while I myself would feel distinctly uncomfortable with the prospect of anyone having comprehensive knowledge of my personal, individual mind--I really don't see anything particularly frightening about someone else being able to have encompassing knowledge of my Americaness, my working class origins, my mixed Jewish and German ancestry, my maleness, my avocation as an anthropologist, or any other aspects which are so obviously public and shared with large numbers of other people.

11 In fact, much of the thinking on the nature of hasina discussed at various points in the book could be seen as the rudiments of a social theory, to try to put a name on invisible mechanisms by which a consensus between them becomes translated into a power which, as Durkheim noted, seems to have coercive force outside of any individual.