The methodological sharpening of cultural anthropology

One way to look at the history of (American?) anthropology is through the rise and fall and rise of the four field configuration of subdisciplines. In The Beginning the four fields were easily combined for a number of reasons: each field was not very specialized, which meant that individual anthropologists could learn a bit about all of them while the Boasian predisposition to particularizing research made close study of a phenomenon using multiple approaches seem attractive. Since then (one possible narrative might go) the subfields have split up and specialized and now are ready to be reintegrated into a new Even More Holistic disciplinary configuration.

Of course this narrative is the line ‘the subfields have split up and specialized’. It does not take very long to conjure up images of the methodological advances made by biological anthropology, archaeology, and linguistics over the past century. But… what about cultural anthropology?

Its clear that we’ve come a long way in terms of our understanding of how culture works, how we might analyze it, and so forth. And its also clear that we know a lot more about the world in 2007 then we did in 1907 (for instance, ‘my’ people were just a blank space on map in 1907). But other than some very modest (but important) developments in taking notes, making sure you’re doing your job and so forth, we just don’t see the same sort of sharpening of methodology in cultural anthropology that we’ve seen in the other fields.

This isn’t to say that people haven’t come up with such methods. Linguistic anthropology has developed both methods and theories that seem to come closest to creating something like what I’m called ‘sharpening’. But of course there have been other contenders before, including The New Ethnography of the late 50s and early 60s. This fell out of favor for many reasons, but I sometimes wonder if cultural anthropology’s failure to sharpen itself isn’t the result of a genuine (if often unarticulated) sense that these sorts of specialized approaches simply remove anthropologists from the nitty-gritty particularism that is at the heart of what we do.

I think that we perhaps like early symbolic interactionists in this regard — we are too suspicious that methodological specialization will lead us away from, rather than closer to, the stuff of social life.



Alex Golub is an associate professor of anthropology at the University of Hawai‘i at Mānoa. His book Leviathans at The Gold Mine has been published by Duke University Press. You can contact him at

16 thoughts on “The methodological sharpening of cultural anthropology

  1. I don’t know what the etiquette is, but I should probably introduce myself first as I’ve never posted on the site before. Briefly, I am finishing a PhD in Social Anthropology at Cambridge on notions of personhood and the idea of ‘the network’ in policies and among university-industry actors in the biotech sector.

    That out of the way…I’m not entirely sure I know what you mean by ‘sharpening’ methodologies. I seem to read ‘sharpening’ above to imply both a sense of greater accuracy in the results one gets, and greater consensus among members of a sub-discipline about the implications of those results. I suppose in the bioanth and archaeology realms you might mean things like the introduction and development of more ‘advanced’ techniques (and technologies) for analyzing biological relationships (e.g.: genetic analysis, etc.), or for dating objects/samples (e.g.: C14 dating methods). On the one hand, these new techniques have enabled more ‘accurate’ or reliable results for the traditional activities of these fields. On the other hand they have both opened up entirely new possibilities for new kinds of investigation, and created new sites of uncertainty in the very problems they are meant to solve. For instance, advances in PCR have made it possible to derive genetic data from Neanderthal specimens for comparisons of genetic variation. However, this data, when used as a dating technique for patterns of human evolution has produced strikingly different results to the dates interpreted from the fossil record based on the strata in which remains have been found as well as C14 dates.

    Far from ‘sharpening’ (if by that one means providing greater certainty or consensus) these new methodologies (or are they methodological techniques?) actually have created more spaces for discourse and negotiation within their respective fields. That is not to imply that bioanth and archaeology are all in a muddle. Rather, the point is that a lack of certainty or agreement on methodological grounds goes hand in hand with ‘advances’ in methodology. In that sense, I don’t see that Cultural anthropology has differed so greatly. Perhaps the main difference is that many (?) cultural anthropologists’ orientation to an inductive ideal for research (if not always in practice) has perhaps shifted beyond Boas’ particularism. But I’m sure someone will disagree with that! 😉

  2. I think aspects of the reflexivity stuff have contributed to what Rex might be calling ‘sharpening.’ I am specifically thinking, for example, of better understandings of communicative and metacommunicative routines (see Briggs) that have helped ethnographers methodologically to shape their research. In addition, the interrogation of the isometrics of place, power, and culture represented by Gupta & Ferguson I think has helped ‘sharpen’ the methodological implications of studying ‘a culture,’ even if it has also too easily been read simply as a rejection of such an endeavor. The whole turn to ‘history’ might be seen as a sort of sharpening, even if, as Mr. Leitner points out, gestures toward precision and accuracy and sophistication and what not can open up their own new uncertainties. In all of this, I distinguish ‘methodology’ from ‘data gathering,’ believing that methodology has to do with a set of concepts for operationalizing theoretical queries the specific means through which empirical data are collected.

  3. I found myself having thoughts similar to Strong’s–namely that the sorts of “sharpening” (dare I say “progress”?) that have taken place over the last 100 years in sociocultural methodology have to do with the conceptual tools through which we understand what we are doing–which are indeed our “instruments” in a way not completely different from the ways microscopes, gas spectrometers, and questionnaires are instruments, in that all these things are ways to constitute data. Among such sharpenings: the pragmatist conception that concepts ARE tools (or the Weberian variation on this idea, the notion of the “ideal type”) and a much more sophisticated understanding of what is at stake in delineating the units of investigation and of the ways those units are constituted. In some ways Boas and his students were not so different as we sometimes imagine, but (and here I agree with Strong), we do know more about why and how to pay attention to the metacommunicative context of our work–and that is a METHODOLOGICAL advance that leads to an advance in the kinds of things we know (here I would cite Bashkow’s _Whitemen in the Moral World of the Orokaiva_ as well as Briggs). A way to frame these questions about conceptual frameworks as methodological advances might be to ask of concepts “what kinds of things do the let us know?” and conversely, to ask of the things we know “what kinds of concepts let us learn these things?”

  4. Suppose, for the sake of argument, someone asserted that, to borrow Thomas Kuhn’s terminology, cultural anthropology remains at a preparadigmatic stage. The evidence is all around us, in a plethora of competing views and an absence of consensus on how to conduct research.

    Thus, the closest we come to sharpening is a shift of focus induced by the observation that in focusing on X we have blurred or ignored completely Y.

    Even when multiple studies take place on similar topics or in similar parts of the world, the product is like the blind men’s impressions of the elephant–none individually wrong, but the whole remains problematic.

    In the British anthropology in which I was trained at Cornell, there had, for better or worse, briefly been a paradigm. Social anthropologists went out, drew maps, hut diagrams and genealogical charts, and described the basic power structure and corporate groups of the peoples they were working with. Methodology focused on separating social facts, constants across some population, from individual variation, a topic handed off to psychologists. Eventually this paradigm came to feel like a straight jacket; Evans-Pritchard coined (or popularized) the phrase “the dead hand of competence” to describe it.

    Gradually more attention was paid to ambiguity, contradiction and change, which made the separation of social facts (shared constants in social behavior) from individual variation (now conceived in terms of social agency, choice and innovation). Still, however, a common paradigmatic mandate to provide details on such topics as kinship, marriage, lineage organization, land tenure and succession to office made it possible to compare institutions in China with those in Africa or peasant villages in Indonesia with their counterparts in Mexico. Following the collapse of that paradigm, what, however, are we left with?

    Arguably the particularism to which Rex points is not a sharpening at all. A new paradigm has not replaced the old one. The old one is collapsed and we are all once again thrashing around in a preparadigmatic muddle, reliant once again on a crude empiricism guided by nothing more than whatever “theories” strike our fancies.

    Our debates resemble those that might occur between practitioners of Ayurvedic and traditional Chinese medicine. Practitioners on both sides may be highly intelligent people with keen empirical insight into conditions affecting their patients health. Since, however, Ayurvedic physicians frame their understandings in terms of

    Vata–the air principle necessary to mobilize the function of the nervous system
    Pitta–the fire principle which uses bile to direct digestion and hence metabolism into the venous syste, and
    Kapha–the water principle which relates to mucous, lubrication and the carrier of nutrients into the arterial system.

    Meanwhile, their Chinese counterparts frame their understanding in terms of Yin and Yang and the Five Agents (Wood, Water, Fire, Metal, and Earth). At the end of the day, they, like us, largely wind up talking past each other.

    Sharpening? No. The muddle continues.

  5. Thanks for the comments all. I ‘get’ what Strong and Comet Jo are saying — for both of them, it seems, the ‘methodology’ means ‘the methods you use when you are writing your ethnography’ and these are distinct from ‘data gathering.’ But I think I’d like to push back against this a little bit. What I had in mind (I think!) when I wrote this piece was exactly what one does when one is ‘data gathering’ and not what one does when one ‘uses concepts to analyze the data’. What does it say about cultural anthropologists that we have Armchair Sharpening but not Field Sharpening? Why does it seem unproblematic to us that we have one and not the other?

    This gets back to David’s point about the imagery I used. Knowledge, as Foucault said, is for cutting, and I suppose I was trying on an ‘efficaciously honed’ versus ‘unspecialized and ineffective’ metaphor. But David is absolutely right — another way to describe it would be ‘opening up’ in the sense that once you have a method for gathering data that produces (as it were) an ‘experimental system’ which can be tweaked and tweaked again to produce an entire realm of exploration and activity.

    I think its telling – although of what I’m not sure — that for cultural anthropology our ‘experimental system’ is the ‘archive’ of material we bring back from the field rather than whatever sort of mis-en-scene we have while we are actually doing research. Is this a _bad_ thing, as John suggests? I’d be willing to say so, although I’d prefer us to be ‘unparadigmatic’ rather than use the teleology-laden ‘preparadigmatic’ and all that term implies.

  6. “Unparadigmatic” is so wimply. Why not go whole hog and say candidly “paradigm-free”? Then, we can define ourselves as anything that restricts our God-given individual rights to do whatever we please.

    Or, in a more constructive vein, we might consider the thoughts of David Kelley, managing director of IDEO, who lists the anthropologist first among the _Ten Faces of Innovation_ described in his latest book.

    The following is cross-posted from a discussion of design ethnography on Anthro-L:


    Kelley writes that IDEO (the design firm behind such products as the iPod) has embraced the Anthropologist role as central to its design process. “So what makes Anthropologists so valuable? At IDEO, people in this role typically start with a very solid grounding in the social sciences…But what’s apparent when you work with them is not their academic knowledge so much as a sense of informed intuition, akin to what Harvard Business School professor Dorothy Leonard calls ‘Deep Smarts.'”

    Kelley goes on to discuss half a dozen characteristics of the Anthropologist.

    1. Anthropologists practice the Zen principle of “beginner’s mind.”

    2. Anthropologists embrace human behavior with all its surprises.

    3. Anthropologists draw inferences by listening to their intuition.

    4. Anthropologists seek out epiphanies through a sense of “Vuja De.” (a sense if seeing something for the first time, even if you have actually witnessed it many times before).

    5. Anthropologists keep “bug lists” or “idea wallets” (in which they write down bits and pieces of everyday experience that surprise them).

    6. Anthropologists are willing to search for clues in the trash bin (seeking inspiration in unusual places).

    The chapter elaborates these ideas.

  7. Whoops! In first line of previous message, please change “wimply” to “wimp” and change second sentence to

    “Then, we can define ourselves as rejecting anything that restricts our God-given individual rights to do whatever we please.”

  8. Let me get this straight (in case I’m confused): the professor whose Ethnography Of The State course included Hobbes, Thucydides, and the Book of Genesis is being called out for claiming that “we can define ourselves as rejecting anything that restricts our God-given individual rights to do whatever we please” while the ivory-tower dropout who advocates combining Chinese and ayurvedic medecine with business school books is the one bemoaning the loss of ‘scientific’ paradigms…? Physician, heal thyself!

  9. Rex,

    Go back and reread the message to which you refer again. I think that you will find that the conversation between Ayurvedic and traditional Chinese physicians was not being recommended as a model. It was, on the contrary, intended as an illustration of what happens when people, who may be quite intelligent and insightful individuals, discuss topics perceived through different worldviews–with no paradigm in place for how to resolve their differences.

    And I wasn’t calling you out on anything but choice of words. To me “unparadigmatic” means that we operate with no constraints but personal inclination. That sounds pretty weak to me, especially coming from a guy a respect, not least for raising the question of how the methods we use in the field might be improved.

    My own preference, for what it’s worth, is to get as close as possible to scientific method, while recognizing that experimentation is mostly out of bounds for ethical and other reasons and collecting samples that meet size and independence criteria for statistical hypothesis testing is rarely a practical possibility.

  10. Aparadigmatic. In _Anthropos Today_, Rabinow notes that there are many many logics, logoi?, that govern inquiry in the human sciences today, that many of these are directly in conflict with each other, and that no one cares — they just go right on doing what they do. Institutionally, anthropology and sister pursuits are ‘outside’ any configuration (assemblage!) that could police them as normal science. This relates to the question of social anthropological methods as candidates for ‘sharpening’ or not; pace John, I don’t think this means that we are talking simply about rival tastes. I think the measure of this ‘advance’ would simply be usefulness: what helps us understand ourselves today? And see again what Comet Jo wrote, and then remember that Rabinow & Co seem obsessed with ‘equipment.’

    One question: What about Google?

  11. A question for Strong: If usefulness is the measure of advance, how do we measure usefulness?

    One approach is academic and evaluates competing “models-for” phenomena being studied. Here the best that I have come up with is the scheme I propose in my 1995 article in American Ethnologist “Negotiating with Demons: The Uses of Magical Language.” I count as superior models that (1) account for more detail and (2) the sequence in which details occur. Consider, for example, a classic murder mystery by Agatha Christie: Hercule Poirot’s account of who committed the murder is superior to that of the bumbling policeman, because Poirot has noticed details the policeman missed and the order in which they occurred.

    Another approach, exemplified by Tom Kelley’s account of how IDEO conceives of the anthropologist, might be described as entrepreneurial and directed at better “models-for.” Here, the anthropologist’s informed intuition is valued because it offers fresh insights, hints that designers can turn into products or processes that deliver higher value. If the language here appears too business-oriented, it may be worth considering the work that IDEO has done on improving emergency room procedures, a worthy goal, indeed.

  12. Whoops, again! The first “models-for” should, of course, be a “models-of.” The reference is to Clifford Geertz’s distinction in “Religion as a Cultural System.”

  13. At the risk of being accused of casuistry, perhaps like Poirot I would evaluate usefulness (or sharpness, I guess) on a case-by-case basis! A criminal investigation and a design studio would probably find different sorts of knowledge useful or illuminating, as John I think indicates here. What makes this question interesting and vexing I think is precisely that the two types of reasoning that John identifies are embraced within a single discipline, a discipline that in refining its concepts nevertheless displays promiscuous openness with respect to methodological and epistemological styles.

    Also: What about coding software?

  14. Strong,

    Allow me to push a bit: What do you mean by “case by case”?

    Let us suppose (using a scheme devised by Noam Chomsky) that our methods are not discovery procedures; we can not apply a formula and, voila! discover Truth. Our methods are not decision procedures; given a body of evidence and a theory that purports to explain it, we can never be sure that the theory in question is Right. The best we can say is that our methods are evaluation procedures. That is, given a body of evidence and at least two competing theories, we can rank the theories as better and worse, based on the evidence in hand. New evidence may emerge, new theories may be proposed. But, based on what we have before us, a body of evidence X and theories A and B, we can say confidently that B is better than A.

    Since I find this way of viewing science (and, more broadly, scholarship) compelling, I find myself asking what, after all, is a case. Is a case simply whatever body of evidence we stumble upon and find ourselves interested in? Or are there criteria for better and worse evidence?

    Or is a case a body of evidence and a theory about it? Don’t we need at least two theories (e.g., Poirot’s and the policeman’s)?

    Arguably, most anthropology does involve two theories: a conventional view and a second perspective that reveals something interesting not included in the conventional view. The conventional view may be that of the people whose lives we study, e.g., traditional language ideologies vs. the linguist’s descriptions of phonology, morphology and syntax. The conventional view may also be that of people who make assumptions about the people whose lives we study, e.g., they must have nuclear families, they must have an Oedipus complex, to which the anthropologist replies, not these people.

    So far, so good. But this brings us back to the meta-question: What counts as a better or worse case? Is a case a good one if the author’s theory is superior only to a straw man? If the evidence offered in support of the author’s theory is thin and sketchy at best?

    Lots of interesting stuff to think about here.

  15. Rex wrote:

    bq. I ‘get’ what Strong and Comet Jo are saying—for both of them, it seems, the ‘methodology’ means ‘the methods you use when you are writing your ethnography’ and these are distinct from ‘data gathering.’ But I think I’d like to push back against this a little bit. What I had in mind (I think!) when I wrote this piece was exactly what one does when one is ‘data gathering’ and not what one does when one ‘uses concepts to analyze the data’

    I think the distinction between sharpening data-gathering and sharpening analysis is not so clear, and I think we have in fact sharpened the former. In the introduction to Sharon Hutchinson’s Nuer Dilemmas, she talks about the differences between her approach to participant observations and Evan’s Pritchard’s: where he counseled hiding one’s disagreements and reactions to people, and a form of participation that amounted to “trying things out” Hutchinson expressed her “uncertainties” and “experiences of culture shock” and saw her fieldwork as requiring an engaged interlocutor. It seems to me that there are parallels between an approach to culture that sees it as contested and reflective, and Hutchinson’s sort of fieldwork–I suspect that most of us practice different styles of fieldwork than our ancestors did and that these are connected to our different theoretical commitments. One could ask why this isn’t talked about so much in terms of methods, but one of the problems is that how exactly one goes about being an engaged-but-not-obnoxious interlocutor varies from place to place–this puts a limit on our ability to generalize the shorts of sharpening the sociocultural data-gathering has undergone, though I do think such sharpening exists.

  16. Comet Jo writes,

    ” one of the problems is that how exactly one goes about being an engaged-but-not-obnoxious interlocutor varies from place to place”

    Still, we might make a beginning by sharing ideas and experiences. When I was preparing for the field, the message I heard repeatedly was to be careful not to become too closely identified with any particular segment of the group whose lives I was studying. The fear was winding up with too one-sided a few of what was going on by losing alternative perspectives.

    Later (my age is showing) it became clear, as feminist anthropology emerged, that the prototype for this problem was a male anthropologist going to the field and hanging out with the guys, thus missing the women’s perspectives on whatever he happened to be studying. Annette Weiner’s restudy of Trobriand exchange was a sharp reminder of how much a purely male perspective can miss.

    What I found worked for me was what I might now label “cunning stupidity” (for years I have thought of it as doing my Columbo act). Thus, one of my papers begins with my annoying practitioners of Chinese folk religion with questions like this.

    1. You’ve told me that when spirit money is burned, it turns into real money in the world of the spirits, right? Right.

    2. You(ve told me that the spirits take only the essence of the food offerings, leaving the rest for us to eat, right? Right.

    3. Well, why can’t we make paper food offerings, burn them and have them turn into real food in the world of the spirits?

    4. Conversely, why can’t I lay some regular money on the altar and have them just take the essence of that? Wouldn’t that work as well.

    The inevitable “No, no, no…that’s not how it is done” in response to 3. and 4. pointed toward an ethnographic problem: Why should there be two kinds of offerings, one burned and one not, one mock money, the other real food.

    The answer I proposed has to do with the standard sequence of a Chinese ritual transaction: (1) Invite the spirits to a banquet; (2) present your petitions or demands to them; (3) send them on their way. The food affirms a close enough relationship that the spirits should listen to what you say. The money restores an appropriate distance once the transaction is over. Together the offerings enact a classic Chinese attitude toward powers-that-be. Deal with them when you have to, but keep them at arms length. Knowing this becomes the key to understanding a wide range of ritual variations, from all money (Exorcisms, Ghosts be Gone!) to lots of food and just a bit of money (Confucius birthday).

    Be that as it may, I have often wondered how this “cunning stupidity” approach would work in other settings.

Comments are closed.