Why write?
On the study of anthropological texts

Olaf H. Smedal

Doctoral lecture, self-selected topic. Department and Museum of Anthropology, University of Oslo, December 1994


To download, print, or bookmark, click: http://www.anthrobase.com/txt/S/Smedal_O_01.htm.
To cite, quote this address and the download date. Not for commercial use.
1994 Olaf H. Smedal. Distributed by www.AnthroBase.com.
Do not remove this notice from digital or paper copies of this text. 


Table of Contents

The Debate
Why Theory?
The Trick



Since this lecture marks the beginning of a two-day ritual event, I want to begin by recalling certain remarks that the late Roger Keesing once made. Almost twenty-five years ago, at a very prestigious conference in the United States, Keesing introduced his paper by confessing uncertainty about how to proceed. He said,

"Most of what I think I know about giving an address on such an illustrious occasion, I have learned from reading ceremonial lectures by distinguished British colleagues. As a professional student of strange customs, I have sought to distil out the essential principles of ritual lecturing and to follow them here. They seem, first, to be heretical and, if possible, occasionally outrageous; second, to be polemical and contentious, overstating one's cause and creating straw men where rhetorically useful; and third, to exaggerate the importance of one's own work" (Keesing 1972:28n1).

Keesing's strategy probably did secure him an attentive audience, but I imagine that to adopt it on this occasion is hazardous. A major portion of my own work only tomorrow comes under public scrutiny for the first time. To proclaim great importance for it in advance would be not merely hazardous but ludicrous. But there are other reasons, too. For to speak loudly of strange customs is now known as orientalism. To distil out essential principles of anything is obviously essentialism. And to be heretical, well, it is perhaps all right - but it depends on what one is heretical about, and on who is listening. I hope the afternoon's audience will accept, therefore, that I limit this presentation to a few polemical and overstated contentions - directed towards straw men.

"Anthropological texts" is a vast topic. It ranges from poetry to kinship mathematics, from analyses of emotions to analyses of demography, from works on myth, ritual, and indigenous philosophies, to works on ecological adaptation, modes of production, and systems of oppression. What is more, many texts deal with several of these subject matters in combination. The output of books and articles is growing rapidly, and the number of subfields is expanding. The latest addition I have come across is "the anthropology of rhythm" (You 1994).

I cannot deal comprehensively with this vast and extremely heterogenous topic. Not only because one can only say so much in one lecture, but also because the greater part of this literature is unknown to me and is likely to remain so - just as it is, and probably will remain, to most of you.

The heterogeneity, or multiplicity of the topic presents another difficulty as well. For if to talk about texts becomes pretty meaningless unless one talks about what the texts are about (which I tend to think one should - but here is where my straw men and I part company), and if what the texts are about derives from the autobiographical particulars of their authors, the necessarily unique field experiences of their authors, the various theoretical orientations and communicative skills of their authors - and this is only the beginning, because the authors' interlocutors ("informants", "friends", "specialists" or whatever) are not in the picture yet, nor are the cumulative interactions between them and the authors - if all this is relevant for the final, textual product (which I think it must be), it is undoubtedly excessive to even hope to say something of general validity or even interest by examining one or two books. I shall instead, therefore, later on, attempt a trick, and introduce something truly outrageous, which to any anthropologist will seem redundant in the extreme - namely an invented ethnography. If the trick is successful, it will bring home one or two points about what good ethnography may entail, and why it is difficult to write it.

The debate

If it is true, as Stanley Diamond said, that anthropology is "the study of men in crisis by men in crisis" (1972:401), it is only to be expected that many perceive that the discipline itself is in crisis. The crisis that most people write about, is the crisis of representation, that is, the crisis of writing.

"It is a melancholy affair to write about writing about writing about writing ethnography", wrote Michael Carrithers recently (1990:53). Since I agree with him, and am not much given to melancholy, I must explain what prompted my choice of topic.

A while ago I became aware of a debate in the journal Man over a book published by Fredrik Barth in the early 1960s (1961? 1964?); namely Nomads of South Persia. Anthropological works are of course debated all the time, as most of us think they should be, but what caught my interest was the content of this particular discussion. It was set off by an article by Brian Street called "Orientalist discourses in the anthropology of Iran, Afghanistan and Pakistan". Street begins his discussion of Barth's work, innocuously enough, by speaking of it as having the virtue of "offering simple, holistic models based on crisp hypotheses" (1990:247). More particularly, the book in question "was one of the first to describe the political structure of nomadic pastoralists, and it set the style for future work both in its emphasis on this group and in its use of segmentary lineage theory" (1990:247). It is on these two points that opinion is divided. Is the book a work in political anthropology or isn't it? Did Barth apply segmentary lineage theory or didn't he? On the one hand is Street who claims that it is, and that Barth did. In building his case, Street refers to a dissertation by Sue Wright, who conducted fieldwork in the area in the 1980s. (The three main protagonists have now been introduced.) On the other hand is Barth who, in his response in Man, disagrees: In his view, Street plainly distorts the book. With regard to the first point, Barth writes:

"In terms of my own declared intention, and the understanding of most of my readers outside Britain, its theme was rather otherwise, with a focus on ecological analysis and, connected with that, certain economic, demographic and organizational processes" (Barth 1992:175).

With regard to Street's second charge, he says:

"[M]ore startling to me is my supposed use of 'segmentary lineage theory' (Barth 1992:175). A reader must be either grossly inattentive or ignorant to see segmentary lineage organization in the pattern that unfolds (...) [in] my account" (Barth 1992:175).

One might think that this would be the end of the matter. But Brian Street and Sue Wright in turn responded, and it was after reading them that I began wondering "Why write"?

Before I go on speculating aloud about this, I should name a few straw men. I have already remarked that what some consider a major crisis in anthropology is a fairly recent concern with how anthropologists write about those they have studied. This concern is usually referred to as the "literary turn" in anthropology. Its history can be traced to a number of intellectual traditions of variant historical depth and national origin, with the aim to go beyond established paradigms of representation, explanation, and understanding. These diverse ideas fused and flourished rather feverishly in the United States, and became generally known in anthropology through the publication of three books and a series of articles in the 1980s. The broad idea was to subject ethnographies to analyses drawn from literary theory. The men most prominently responsible for this "literary turn" - and its present outcome, known as "an experimental moment" - are James Clifford, Michael Fischer, George Marcus, Renato Rosaldo, and Stephen Tyler; and many would probably add to this list, minimally: Vincent Crapanzano, Kevin Dwyer, Paul Rabinow, Paul Stoller, and Richard Price, although how well Rabinow fits here is debatable, and Price feels uneasy about being included.

A major focus has been on the role of the variously "positioned authors" versus the role of the variously "positioned subjects" the authors write about. This is known as reflexivity. Key terms are dialogue and polyphony. Another focus has been on the rhetorical styles which ethnographers use in order to be believed. To expose these styles is also intended to inspire reflexivity. The most exemplary key term here is probably strategy. A third feature, which has a long history in anthropology, is a systematic scepticism towards the nature of concepts in general, and concepts typical of social science discourse in particular. The key term is deconstruction. On reflection, most would probably label this, too, reflexivity.

As far as I can see, Carrithers's melancholy remark was motivated by a perception that there is a limit to how many layers of reflexivity we can handle before we either lose track of what we are doing, or find ourselves reflecting upon little besides ourselves.

I now return to the debate I outlined.

It is relevant to my argument that I have never visited Iran. I have never spent time with tent-dwelling pastoral nomads. Everything I know about the way of life that the label implies - indeed everything most of you know about it, or think you know about it - derives from what has been written about it. The short and obvious answer to the question I pose: "Why write", is "That we may read". What I propose to do now is to discuss why we should write - and read.

I was first tempted to review Barth's book and the debate over it in substantial terms; to cite ethnographic chapter and verse. But I shall concentrate instead on the more important general issue, namely on how ethnographies are read by the new critics, and on how the critics, in turn, respond to challenges.

So, how did Street react to Barth's comment? First he explains his intentions with the article he wrote by referring to the context in which it was produced: a conference on "Regional issues in ethnographic writing":

"I wished to 'read' the anthropological literature on Iran and other Middle [Eastern] countries somewhat in the way that I had approached the 'ethnographic novel' (...) in an earlier work. This involved (...) uncovering hidden cultural assumptions, embedded as much in tropes and metaphors as in explicit theoretical assumptions" (Street 1992:178).

Commenting on Barth's reaction, he notes that,

Like many authors subjected to the procedure, Barth wishes to re-assert control over 'his' text (Street 1992:178).

After reviewing some of the substantial issues (which I pass over) he concludes that,

"...it is painful for authors if readers keep insisting on their own right to read and to resist the author's intention" (Street 1992:179).

The issue, then, according to Street, is one of the dominance of the author, versus the reader's freedom of interpretation. The issue, in fact, is reduced to one of power. Barth, by responding as he did, is construed by Street as 'protesting too much'. Barth wants 'his' text back, and his claims about having been misread, says Street, only "...serves to reinforce the point" (1992:179).

Sue Wright, the author of the dissertation which Street quotes, soon enters the debate herself (1992). Again I must leave the ethnographic detail aside. The most interesting part of her response is anyway the way she begins:

"Renowned anthropologists arguing over what they think I said or meant brings attention to my work, but it also positions me as an informant. Since an informant is usually silenced once the writing for an academic audience starts, I am glad to have the opportunity to speak for myself.... I would like to state the interpretation of Barth's work that I intended..." (Wright 1992:642).

The irony is palpable. Street contends that Barth's version of what he wrote is not the only one, and that Barth, by referring to his own stated intentions, "wishes to re-assert control over 'his' text" (1992:178). Street then presents his intentions with the critique. And now Sue Wright, the author of the dissertation Street draws on, presents herself as an intending informant. Informants are persons, we should know, who are inherently subordinate to authors who speak on their behalf. But the question we must pose, is why it should be all right for Brian Street and Sue Wright to resurface as intending interpreters of their own texts, as their own informants, if it is all wrong for Barth. Why do they apply one standard to him, and another to themselves?

When a communicator becomes aware that a message has been misunderstood it is not, or should not, be impossible set the matter straight. Barth declares that most readers outside Britain have grasped the book as he intended it - namely as a work with "a focus on ecological analysis" etc. (1992:177). But Street and Wright insist that this is immaterial. The text is autonomous, they say; this is how it is read; Barth cannot do anything about it, and so the matter must rest. To readers of this debate, it seems that Barth is talking past Street and Wright, and they past him.

I suggested earlier that it is reasonable to justify writing by referring to its projected aim: that others read. What this justification presupposes is that these others can read. The great - and no doubt to many disturbing - merit of deconstructionists is their application of rather advanced reading abilities. Deconstruction is hermeneutics pushed well beyond its conventional limits. But it is difficult to see the criticism levelled against Barth's book as anything of that sort. In my view - and I have read the book! - it is as if the critics have not read the first third of it. But this is not all. For instead of attempting to come to grips with the work as a whole, and thus to understand what Barth tried to do in ethnographic and analytical terms (and I think quite successfully), they place before us sentences out of context, and words torn out of sentences, place them in inverted commas, and point their fingers. (This nefarious tactic is obviously contagious.)

Barth's ethnography and analyses are easy to understand. So, when I asked myself, "Why write?" - "What's the point?" - it was with a slightly exasperated feeling that when clearly written texts are, probably wilfully, misrepresented, then the profession faces a problem.

I will just close this debate snapshot by making Clifford Geertz's words mine: "... a more interesting question, after all, than why so many anthropologists can't write is why so many can't read. Or won't" (1990:274).

Why theory?

Moving on now, there are two related points I wish to make with regard to the dissertation on the Ngadha people of Flores in Eastern Indonesia I hope to defend tomorrow. The first has to do with the fact that I am neither the first nor the last to have written about the Ngadha. The second has to do with the concepts and theories that anthropologists bring to the field.

Pater Paul Arndt, a missionary-ethnographer of Kulturkreislehre persuasion, published in the period 1929-1963 some 1500 pages on various aspects of Ngadha people's lives, in addition to a massive dictionary. As I stress in my thesis, I am greatly indebted to much in Arndt's writings, especially his documentation of what Ngadha men (and occasionally, women) say about their beliefs and practices. This is true even with regard to Arndt's major work, on Ngadha 'sociology' or Gesellschaftliche Verhältnisse. But what I was unable to trust were his explanations. Not so much because Arndt was a missionary, but because his 'theoretical' perspective was seriously deficient. As one can read in the only English-language review of his sociological study that I have come across (a passage I have quoted also in the introductory chapter of my dissertation):

The entire corpus makes up a book which is in the nature of an encyclopaedia of Ngadha life, rather than a work to be read through by whoever wants a rounded presentation of this particular culture: the mass of detail makes it very hard for the reader to see the wood for the trees. (...) Arndt presents his data as much as possible in the form he himself obtained it from his informants. The advantage of this procedure is authenticity: one gets facts (at least, facts as seen by the informant) which have not had to pass through the filter of anthropological theory. The disadvantage is that, against the mass of material on some subjects (myth and ritual in particular) there are at times disappointing gaps, even in matters of direct importance for an understanding of social structure (de Josselin de Jong 1956:14).

The curious thing, therefore, is that the absence of an explicit theoretical perspective in Arndt's major work makes it quite hard to read.

For the same reason Arndt was unable to pursue his investigations in any systematic manner. The material was gathered over 30 years, and Arndt expressly passes on to his readers "only that which was said in the way it was said". He has not been selective: "Everything collected should be published" (1954:7).

The work is not disorganised - indeed its table of contents shows it to be very well organised. But it is organised topically; the book consists of a mishmash of topical information drawn from scores of villages scattered over a wide geographical area without the reader getting any sense of how these disjointed details add up. In short, Arndt provides no theoretical context within which to grasp how one "fact" might have anything to do with another. "Facts" cannot properly be localised in what we might call, with a metaphor, Ngadha conceptual space. It is this lack of theoretical focus, lamented by the reviewer, which turns the work into something like catalogue - but a catalogue which is difficult to read.

To the minds of recent critics this should perhaps be welcomed, because a well-defined theoretical perspective is a strategic device. The purpose it serves is to make it difficult for readers to hoist themselves out of a firm narrative grip. The author, in this view, is a schemer, whose hidden ploys should be demasked and brought into the open - something which can only be accomplished by sophisticated techniques, which the recent critics master.

But do they master anything else? When we look back on previous important debates in anthropology we see that they were about which theoretical perspective, analytic model, or interpretive paradigm would generate the most authentic account and authoritative interpretation of those 'Other Cultures' marked out for anthropological analysis (Peace 1990:20).

By contrast, the overriding issue at present is to refocus the anthropological gaze on the authoring of texts, the presentational properties of the monograph, and the colour and tone of the anthropological voice (Peace 1990:24).

Is this all? Is literary criticism the only help we get? Or are there also proposals about what it is important to study, and why? Are there suggestions about how we should think about the world and the people in it? Are there, in short, any theoretical ideas that we might discuss?

One might expect that by now there would be. In fact, one might have expected the argument to be that by employing the advertised textual strategies - such as dialogue, polyvocality, heteroglossia, multicentredness, reflexivity, and so on - the reader might get closer to life "out there", in other words, to understand better how people variously organise their diverse activities, to get a firmer grip, for example, on how power differentials are played out. But according to Stephen Tyler, who for many of the counter critics of the 'literary turn' in anthropology is a favourite whipping boy, the counter critics are actually right in saying that "polyphonic and dialogic texts are not better representations". The mistake the counter critics make is to believe that this is an argument at all because "[r]eprestation is not what [these texts are] about. (...) It no longer matters how we write, because the book is obsolete.... the book is dying out and with it the rhetoric of the book and the language of writing, too" (1989:566).

Since Tyler's eccentric offerings are clearly either too extreme, as here or, which is much more common, too obscure to be useful, even for my sinister purposes, I have consulted instead George Marcus.

To those who were unmoved by the hope he expressed in 1986 that "the play with writing technique brings theoretical insight to consciousness" (Marcus and Fischer 1986:41-42), and who still suspect that the "experimental moment" is mostly "self-aggrandizing, chic, elitist and nostalgic", Marcus now has the following reassurances:

"(...) critiques of rhetoric change the generally understood relationship between theory and research practice in the social sciences. The critiques equalize them as the source of ideas and create dialogue between them - diminishing the hierarchy of authority and subordination that is conventionally understood to hold between theory and research practice. As such, the experimental mood deconstructs the categories of theory and method that organize much work in the social sciences" (Marcus 1994:42).

Depending on how this is interpreted, it is either naïve or banal. It is by no means obvious that "critiques of rhetoric change the relationship between theory and method". But more to the point, who in anthropology thinks of the flow in the relationship between theory and method as being one-way? The point has always been in the anthropology I have learned that theory and method interact, and very intensely, too, and also that the properties of the field of study itself shapes both theory and method. This is precisely why almost every anthropologist I have ever met knows that to design a questionnaire or a rigorous work schedule before arrival in the field is a waste of time. Perhaps wrongly, this is what we tell each other that the sociologists do. The problem, however, that I see in Marcus's prescription, is that it opens for a theory-stripped application of method which can only lead to mindless research. We will have a lot to read and discuss if such research gets under way, only it will be difficult to know exactly what we are discussing, then, because the texts will be - a little similar to Arndt's.

In an article which I found amusing Joseph Tobin (1990) mildly mocks the call for 'radical texts'; i.e. texts of many voices, free from authoritarian pressures about how they should be read. In his quest for radicaler and radicaler and more and more authentic texts he ends up at the Human Relations Area Files which is actually, in Tobin's tongue-in-cheek formulation, "A Text Ahead of Its Time":

Behind its rational scaffolding of topical, subtopical, and area indexing and cross-referencing (...), the HRAF is a teeming mass of discordant voices and narrative disorder. The HRAF is radically authorless, polyvocal, playful, readerly, antitextual, and deconstructive (Tobin 1990:479).

It is authorless because in combining the work of 6000 ethnographers (and the number is growing) any authority that derives from any author's or narrator's persona is stripped away. It is polyvocal because, it is "a compendium of voices, voices of millions of informants and thousands of ethnographers"; reading it "we cannot avoid heteroglossia" (Tobin 1990:479). It is also playful and antitextual because, being a "collection of directories, file drawers, and plastic fiche (...) it is joyfully nonlinear and antinarrative, like a children's flipbook or a postmodern novel that can be read forwards or backwards". It is, says Tobin, "a nonbook" (1990:479) (remember Tyler's pronouncement that the book is dead). Moreover, it is readerly - in fact its sheer size demands of readers "that they make choices, that they determine the text (...); readers are asked, indeed forced, to follow their own intertextual associations" (1990:479). And finally, it is deconstructive in that it "violates the whole" - not, of course, the "whole culture" which remains, if it exists out there, untouched - but the "'articulated whole', the ethnographer's configuration or pattern, 'that distinctive illusion of holism (which) is the most basic characteristic of realist writing' (Marcus and Cushman 1982:30-31)".

In short, as Tobin puts it, 

"The HRAF takes the imposed narrative order of the linear ethnography and explodes it. [It] literally deconstructs an ethnography, breaking up the final seamless ethnography into notecards, returning the text, ashes to ashes, dust to dust, into the disjointed field notes the ethnography was before it was sorted into subject piles and artfully fused into a "whole" (Tobin 1990:480).

It would be unfair to place Arndt's sociological study in the same category as any HRAF section text. But there is a sense in which both are 'radical texts'; both consist of snippets of dialogue, observations of very varying quality, of informants' statements, ethnographers' comments, broad generalisations, and narrow focus - all rather haphazardly collected together in a manner that in no way presents to the reader a 'seamless' whole.

The problem with radical texts is that there is not very much in them to think with or, rather, against. They offer the reader neither guidance nor resistance; and by the same token they leave the reader little wiser. The reason for this should now be fairly clear. And the consequence of producing such 'authorless', 'readerly', 'polyvocal' texts is that any debate over what the texts are 'about' is deflected or defused. It is simply impossible to disagree with the HRAF.

It has been pointed out for some time that the so-called post-modern movement in anthropology has been surprisingly lame in its politics, despite the emancipatory potential it has proclaimed for itself. Its declared retreat from privileging any voice over another, to let everyone speak, implies retreating from a position from which it becomes possible to argue that certain voices, or rather - since the "voice" metaphor is exhausted - that certain persons have more power, and others have less (see e.g. Mascia-Lees et al. 1989). I still think the point is valid. But I now conclude also that, just as it is impossible to disagree with the HRAF, it is impossible to disagree with post-modern texts. To attempt to disagree with its authors is simply a waste of effort.

If I can close this section by briefly exaggerating the importance of my own work, I should like to assert that the dissertation which will be debated tomorrow contains enough data to theorise, enough theoretical structure to merit discussion, and enough conclusions with which it is worthwhile to disagree. I could have done more, I could have done something else, and I could have done it better, but at least it is a work which has been conceived, not merely compiled.

The trick

I change my tack at this point, in the direction of the outrageous.

Rodney Needham has proposed that anthropology is the empirical philosophy of mind. Others have said that anthropology is philosophy with people in place. However little or much one agrees with such statements, it seems to most anthropologists (though less to philosophers) that anthropology and philosophy have something rather profound in common. But the relationship between philosophers and anthropologists is as frequently characterised by mutual dissatisfaction, or perhaps suspicion, as it is by respect and admiration. Anthropologists tend to think of 'pure' philosophising as a drift into the metaphysical and the abstract, while philosophers often dismiss ethnography as either simple-minded descriptions of the empirically given and low-level theorising, or - when anthropologists attempt to go beyond this, that is, when they betray philosophical aspirations - they dismiss it as poor philosophy and perhaps rightly so.

There are exceptions, of course, on either side of the divide, and a philosopher whose attitude to the relationship between 'ordinary life' and philosophy was somewhat different, was Wittgenstein. However, like most philosophers Wittgenstein preferred to invent illustrative material, rather than exploiting examples provided by anthropologists. But it is notable that his numerous invented cases often begin with the little phrase "imagine a tribe..." (e.g. BB, p.103) - that is a tribe whose conceptual relationships with the world differ from our own in some fundamental manner. I am not suggesting that Wittgenstein is an anthropologist manqué. But his many "tribal" examples are meant to show how words are social things, how "philosophical problems arise when language", as he put it, "goes on holiday" (1978a (PI),'38), and how the meanings of words inhere in the uses to which they are put in public (PI,'43).

It is in this sense that many anthropologists share Wittgenstein's central concern. Anthropology, if we understand this not merely as "the study of men in crisis by men in crisis" but as the comparative study of societies and cultures, then its practitioners characteristically look deeply into the use people make of their concepts. On this basis, anthropologists - I think very many of them - attempt to state something noteworthy about the community in question. Such activity: to answer questions of the form "What is the meaning of X? / How is X used?", is only possible after dwelling in each language as it is practised. Plausible interpretations, at various levels of abstraction, are therefore available only after considerable time has been spent in a given social setting. It is for this reason that anthropologists should perhaps remember that what Wittgenstein calls language games "are... set up as objects of comparison which are meant to throw light on the facts of our language by way not only of similarities, but also of dissimilarities" (PI,'130, my emphasis).

And Wittgenstein says explicitly that what he supplies "are really remarks on the natural history of human beings" (PI, '415). But it is because the philosophical confusions we suffer are so close to our everyday language - and therefore so difficult to see clearly - that he permits himself also to "invent fictitious natural history" (PI, II, xii): Fictional cases provide unfamiliar arenas for our words and thereby make us grasp how the way we use them depend on actual "facts and features of our own lives and the world around us that would have otherwise gone unnoticed" (Cerbone 1994:181n20). If Wittgenstein's aim is to make us think about our language, our practices, and ourselves so that we may understand what we say and when it makes sense to say it - because such understanding will free us from philosophical confusions (Cerbone 1994:165) - then hopefully anthropology, by way of comparative juxtaposition of words used in actual language games, played in different cultures, can serve much the same purpose.

It is in this spirit that I now introduce one of Wittgenstein's invented tribes. (In the following I lean heavily on, and quote long passages almost verbatim from, an article by David Cerbone (1994).)

The scenario we are asked to imagine involves a community whose practices of collecting and distributing wood, despite superficial similarities, diverge markedly from our own: Wittgenstein's wood sellers. He introduces the wood sellers after a number of remarks that point out the variety of methods by which one can set the price for a quantity of wood: wood can be sold by the cubic measure, by weight, or, more creatively, by taking into account the age and strength of the persons who cut the wood. One could also allow people to take what they want for some fixed price, or simply give the wood away. None of these methods is particularly unfamiliar or provocative, and it is clear that Wittgenstein does not wish to have raised any issues of particular importance prior to introducing this particular community of wood sellers. He probably expects that the reader has accepted what has been described so far. What follows, however, is intended to prompt more of a reaction:

Very well; but what if they piled timber in heaps of arbitrary, varying height and then sold it at a price proportionate to the area covered by the piles?

And what if they even justified this with the words: "Of course, if you buy more timber, you must pay more"? (RFM, I, 149)

In the methods of determining the cost of any given pile of wood listed prior to this case, each of them fixes a means for justifying the determined cost: (1) The difference in price between two piles of wood sold by the cubic measure, for example, can be justified by showing that the respective piles have different volumes; given a fixed price per unit volume, one can justify one pile's costing more than another, and one does indeed get more wood by paying more. (2) In the case where the price is fixed according to the age and strength of the wood cutter, a smaller pile may cost as much as a larger pile if the former is cut by an older, weaker cutter, and the justification for the sameness in price would appeal to those facts about the respective cutters. One might be unwilling to accept those facts as the best means for determining the cost of a pile, but, given the method, the justification is more or less straightforward and is one we can understand.

In the prior cases in general, one might say that a given method determines an avenue of justification, and so the notion of a correct price for a pile only becomes viable once a method has been specified. Different methods will fix different prices for piles of wood, these different prices will be differently justified by referring to those methods, but we can understand them. Now, however, with these lately introduced wood sellers, something seems to have gone wrong. But what? They certainly seem to have a method for fixing prices: given some price per unit surface area, one simply multiplies this by the surface area covered by the base of any particular pile. Piles with larger base areas always cost more, regardless of their height. So what is the problem?

The difficulty here appears to be that we cannot accept their avenue of justification: a pile with a larger base area isn't always a larger pile of wood, and so to say, "If you by more timber, you must pay more" does not really justify a given price.

Wittgenstein then describes his attempt to point this problem out to them:

"How could I show them that - as I should say - you don't really buy more wood if you buy a pile covering a bigger area? - I should, for instance, take a pile which was small by their ideas and, by changing the logs around, change it into a 'big' one" (RFM, I, 150).

He continues by noting that "this might convince them", and that if this were so, then the case would merely be one in which they were missing something, as though they were a group of children corrected by a well-meaning adult: they just hadn't noticed that a pile with a small base might contain more wood than one with a larger base.

But then Wittgenstein notes that his demonstration might provoke a different response: "but perhaps they would say: 'Yes, now it's a lot of wood and costs more'". He darkly concludes "and that would be the end of the matter" (RFM, I, 150).

With this second imagined response, the wood sellers no longer appear simply naïve, as they would if they were convinced by the demonstration; it is "the end of the matter" in so far as their response signals a failure of communication. We now feel that we are talking past them (and they past us). Wittgenstein closes the remark by concluding: 

"We should presumably say in this case: they simply do not mean the same by 'a lot of wood' and 'a little wood' as we do; and they have a quite different system of payment from us" (RFM, I, 150). 

In other words, given the failure of the proposed demonstration, communication with these wood sellers has broken down (after Cerbone 1994).

To repeat, our first inclination is to ask ourselves if the wood sellers make a mistake. Are they trying to do what we do, but failing?

No. They are doing something else. We hear them saying "more wood" and "less wood" - but we then understand that they cannot mean that: we suddenly realise that we simply have no idea what they might mean by these words.

The problem is in part that they appear to have a different system of payment from us, but the underlying premiss of that is that they seem to have no concept of identity, or at least no concept of identity which we can understand. To us the rearranged pile is "the same", to them it is "different".

I should say immediately that I am unable to resolve the impasse reached here. The example is meant to show precisely that any resolution is impossible; that problems of this kind are insoluble. Still, that this "is the end of the matter" would perhaps not be immediately endorsed by an anthropologist who is more likely to be intrigued than to admit communicative defeat.

If it is true that "Meaning is context-bound, but context is boundless" as Jonathan Culler (1983) has put it, then by appropriately expanding the context it might be possible to establish the meaning of the wood-seller's utterances.

One way to do this would be to return the next day - perhaps the rule was no longer in force, perhaps yesterday was the wood-sellers' equivalent of April Fool's day. Another contextual expansion might reveal that this particular wood-seller market is only one among several - where this particular market is one where "everyone knows" that prices are fixed according to changing rules. Perhaps prices are even fixed according to rules which change in erratic ways: that to come to this market in order to acquire wood is effectively to enter a game where the outcome - as it is in any real game - is unpredictable.

Or one could concentrate on observing, as Wittgenstein anyway urges his readers to do ("Don't think but look", he writes) (PI, 66): What happens when someone else comes to the market? It is a peculiar tribe indeed whose economic participants consist exclusively of wood sellers, and so far we have only the sellers' version of the system of payment. How do native buyers respond to this version? And how are other objects exchanged?

Now I do not wish to trivialize Wittgenstein's point, and to go on inventing fictitious contexts for an imagined case is even worse than to write about writing about writing. So, by way of rounding off this discussion about an imagined "tribe", let us assume that communication in this case really has broken down. To say that communication between human beings depends on the ability to share concepts is to state the obvious. It is a feature of concepts that they are at home in uses, in practices. But I want to contend that ethnography - at its best - manages to overcome such conceptual differences as Wittgenstein is concerned with. The much talked-about anthropological project of translation consists precisely in that very demanding task of elucidating concepts first in their contexts, and then across them. Now it is probably accepted by nearly everyone that concepts and contexts are never absolutely stable. Yet, however transitory they may all be in the long run, some clearly have a longer life expectancy than others.


If after this detour into the imaginary you are eager to learn how I solved a similar conceptual riddle in my thesis, I must disappoint you. I cannot suggest that my own work is ethnography at its best in this sense - the Ngadha communicative challenges were never as fundamental as this one. But, in order to return for a moment to real ethnography, I invite you to contemplate the following anecdote told by Alfred Gell in his book about the Umeda of New Guinea:

"...when walking between villages with a youth, I remarked to him on the rather leisurely pace we were keeping, suggesting that we might not arrive before dark. He (...) assured me that if we were to walk fast, then the sun would go down correspondingly quickly, whereas if we stuck to our leisurely pace, then the sun would do likewise" (1975:163).

We know that the sun sets when it sets regardless of our activities. This Umeda youth knew it to be otherwise. Gell himself summarises by saying: "Astronomical events were conditioned by events on the ground rather than vice versa" (1975:162).

To dismiss the Umeda understanding of the sun's movement as a failure to measure time is perhaps tempting. But it would not help us to grasp how the Umeda may instead be doing something entirely different. What this might be is only possible for us to realise after we have examined their "way of life" in detail and established, among other things, that at the core of their symbolic system is a concern with the organic; that the Umeda 

"...expresses fundamental relationships through metaphors drawn from the organism"; and that "all the separate domains of experience, the most diverse phenomena, [are] related back to a single overriding model: the organism" (Gell 1975:154). 

Thus we may come to realise that the sun the Umeda talk about is not the sun we talk about. I should like to say, with a metaphor I have used already, that it belongs in a different conceptual "space"; it is talked about in terms we are unfamiliar with, and behaves in a way we are unprepared for - which may have practical consequences we cannot anticipate. But that does not mean that it must forever remain incomprehensible. Rather, it is when we have come to realise the conceptual spaces of the Umeda - and, by extension, the conceptual spaces of other peoples - that their practices can begin to make sense to us.

My concluding and very moderate proposal is that we should write in order to make such conceptual spaces available to each other.

To this end, it is counterproductive to condemn or glorify any particular writing style. On balance, I think the varieties, or blurring, of genres we have become familiar with over the last decade or so have enriched anthropology, and as readers we should welcome more of them. But as authors we should never cease to reflect - over why we write.


Arndt, Paul, S.V.D. 1954. Gesellschaftliche Verhältnisse der Ngadha. Studia Instituti Anthropos. 8 Wien-Müdling: Verlag der Missionsdruckerei St. Gabriel.

Barth, Fredrik. 1964. Nomads of South Persia. Universitetsforlaget.

1992. "Method in our critique of anthropology". Man (N.S.) 27(1):175-177.

Carrithers, Michael. 1990. "On ethnography without tears". Current Anthropology 31(1):53-57.

Cerbone, David R. 1994. "Don't look but think: Imaginary scenarios in Wittgenstein's later philosophy". Inquiry 37(2):159-183.

Culler, Jonathan. 1983. On deconstruction: Theory and criticism after structuralism. London: Routledge.

Diamond, Stanley. 1972. "Anthropology in question". In: Dell Hymes ed. Reinventing anthropology. New York: Random House.

Geertz, Clifford. 1990. "Comment on 'Is anthropology art or science?' by Michael Carrithers". Current Anthropology 31(3):274.

Gell, Alfred. 1975. The metamorphosis of the cassowaries: Umeda society, language and ritual. London School of Economics monographs on social anthropology No. 51. London: The Athlone Press.

Josselin de Jong, P.E. de. 1956. "Review of Paul Arndt S.V.D.: Gesellschaftliche Verhältnisse der Ngadha". Man LVI (Jan.):14-15.

Keesing, Roger M. 1972. "Simple models of complexity: The lure of kinship". In: Priscilla Reining ed. Kinship studies in the Morgan centennial year. Washington D.C.: Anthropological Society of Washington.

Marcus, George E. 1994. "After the critique of ethnography: Faith, hope, and charity, but the greatest of these is charity". In: Robert Borofsky ed. Assessing cultural anthropology. New York etc.: McGraw-Hill.

Marcus, George E. and Dick Cushman. 1982. "Ethnographies as texts". Annual Review of Anthropology 11:25-69.

Marcus, George E. and Michael M.J. Fischer. 1986. Anthropology as cultural critique: An experimental moment in the human sciences. Chicago: The University of Chicago Press.

Mascia-Lees, Frances, Patricia Sharpe and Colleen Ballerino-Cohen. 1989. "The postmodernist turn in anthropology: cautions from a feminist perspective." Signs: Journal of Women in Culture and Society 15(1):7-33.

Peace, Ade. 1990. "Dropping out of sight. Social anthropology encounters post-modernism. The Australian Journal of Anthropology 1(1):18-31.

Smedal, Olaf H. Making place: Houses, lands and relationships among Ngadha, Central Flores. Unpublished PhD dissertation. (MS 1994)

Street, Brian. 1990. "Orientalist discourses in the anthropology of Iran, Afghanistan and Pakistan". In: Richard Fardon ed. Localizing strategies: Regional traditions of ethnographic writing. Edinburgh/Washington: Scottish Academic Press/Smithsonian Institution Press.

1992. "Method in our critique of anthropology: response to Barth". Man (N.S.) 27(1):177-179.

Tobin, Joseph. 1990. "The HRAF as radical text?". Cultural Anthropology 5(4):473-487.

Tyler, Stephen. 1989. "Comment on Paul A. Roth's Ethnography without tears". Current Anthropology 30(5):566.

Wittgenstein, Ludwig. 1978a. Philosophical investigations. Third edition. Translated by G.E.M. Anscombe. Oxford: Basil Blackwell.

1978b. Remarks on the foundations of mathemathics. Third edition, revised and reset. Edited by G.H. von Wright, R. Rhees, G.E.M. Anscombe. Translated by G.E.M. Anscombe. Oxford: Basil Blackwell.

Wright, Sue. 1992. "Method in our critique of anthropology: a further comment". Man (N.S.) 27(3):642-644.

1994. "Constructing tribal identity in Iran: response to Ganzer". Man (N.S.) 29(1):183-186.

You, Haili. 1994. "Defining rhythm: Aspects of an anthropology of rhythm". Culture, Medicine and Psychiatry 18:361-384.