In the history of anthropology, comparative analysis has been the main analytic tool used to move beyond the particulars of ethnography. Various types of comparison has been used, from “controlled comparisons” of neighbouring communities, such as those of S. F. Nadel and F. Eggan, to far flung comparisons, such as those of A. R. Radcliffe-Brown, Margaret Mead, and C. Levi-Strauss, to the worldwide samples of G. Murdock and colleagues.

What has comparative analysis been used for? Our anthropological ancestors used comparative analysis to find patterns, more specifically associations–called by some “concomitant variations” or “co-variations"–between factor or elements or, to use the term they preferred, variables. It was their hope that knowing what went together and what did not, would aid us in explaining why things are present here but not there, or why certain patterns arise here and not there. They were trying to explain, and this meant identifying causes and effects.

An example of a concomitant variation, Radcliffe-Brown in “The Mother’s Brother in South Africa” showed a correlation in widespread cases between the indulgent relations between mother’s brother and sister’s son, on the one hand, and, on the other, the authoritarian relations between father’s sister and brother’s son which are characteristic of patrilineal descent systems.

An example of a causal explanation would be Robert Carneiro’s argument in “Slash-and-Burn cultivation Among the Kuikuru and Its Implications for Cultural Development in the Amazon Basin” that productive potential itself does not lead to greater social complexity and development, but rather necessity, as arising in physically bounded regions where population outstrips production, leads to more complex levels of organization and more intensive forms of production.

Other anthropologists have used comparative analysis somewhat differently. Claude Levi-Strauss, in The Savage Mind, compared totemic organization with Indian caste, in order to show that they can be seen as transformations of one another, or, as he (p. 127) famously put it, “castes naturalize a true culture falsely, totemic groups culturalize a false nature truly.”

Carol Ember, in “Myths About Hunter-Gatherers,” using a Murdock style statistical comparison, addresses such popular ideas (largely stemming from Richard Lee’s account of the !kung or Ju/hoansi) that among hunters-gatherers, organization is typically bilateral, women bring in most of the food, hunters-gatherers are peaceful, etc., by examining a world wide sample of hunters-gatherers. Here comparison is used in aid of accurate descriptive generalization, of discovering what is statistically normative among a particular universe.

It has been said that comparative analysis is what anthropologists can pursue in the absence of experimental laboratories. Or, to put it another way, the world and its workings are the laboratory of the anthropologist, and comparison is the experimental technique.

So much about the past. What now? What place does comparison have in contemporary and future anthropology? And how would you use comparative analysis to advance your own research and understanding?

Views: 2185

Replies to This Discussion

Comparison is unavoidable. In even the most interpretive and particularistic modes of analysis we cannot say, "A is B" without the contrast between A and Not-A, B and Not-B that makes the proposition intelligible. Every act of focusing attention contrasts the focus of attention with the Not-that background from which it is extracted. Judging from the examples, however, the issue is what and how the anthropologist compares. Are comparisons based on ethnographic descriptions likely to lead to fruitful theorizing? How can we know unless we know how "fruitful theorizing" compares with "not-fruitful theorizing"?

I note for the record that the anthropologists who have influenced me the most have produced bodies of work shot through with comparisons. Vic Turner approached Ndembu symbols with a mind already informed by Norse sagas, the Canterbury Tales and Dante, not to mention the more immediate comparisons with neighboring peoples studied by his colleagues at the Rhodes-Livingstone Institute. Mary Douglas' group and grid have been a continuing source of information. When I chose three anthropologists to feature in a chapter on Japan, I chose Ruth Benedict, who explicitly contrasts Japanese and Chinese and, indeed, devotes a whole chapter of The Chrysanthemum and the Sword to their differences; Chie Nakane, whose frame and attribute theory is rooted in a comparison of her experience as a Japanese woman with that of the women she studied during her fieldwork in India; and, finallly, Dorinne Kondo, half-Japanese, half-American, searching for Japanese roots, constantly comparing her Japanese and American selves.

On the other hand, a road not taken was offered by Jack Roberts, who had turned the Human Relations Area Files into a playground on which he tested hypotheses about the relation of games and other recreations to politics and personality attributes resulting from different forms of child-rearing practices. HRAF has come in for all sorts of a priori criticism based on presumed deficiencies in the original ethnographic data and the way in which it is coded. I wonder, though, what might be found if the data were opened to the public, along with some up-to-date data mining tools to facilitate exploratory analysis.
For those with careers to worry about, I note for the record that my single most cited article is a piece called "Women's property rights and dowry in China and South Asia," Ethnology 1976. It was, in essence, a literature review, working through available village studies to determine the conditions under which bridewealth (from the husband's family to the wife's) or dowry (from the wife's family to the husband) are paid, which turns out to have a lot to do with women's right or lack thereof to inherit a portion of their parents' estates. It is easy to imagine doing similar papers with different topics but the same basic strategy: working through ethnographic reports for the region in which you are interested, systematically cross-checking what they have to say on the topic you are interested in. The work may not have the buzz of currently hot theory; but if solidly done, it is likely to be cited by everyone who works on a similar topic in the future. A, what did we used to call that? contribution to knowledge.
Nice, highly relevant piece by Daniel Lende on Thomas Hyllland Erikson's Engaging Anthropology.
This sounds like sensible advice, John, and I might actually take it. Many thanks for the lead. It'd be interesting to compare media and nation-building efforts in two different regions, if anyone wants to join me in this at some point... I could do Southeast Asia.

John McCreery said:
For those with careers to worry about, I note for the record that my single most cited article is a piece called "Women's property rights and dowry in China and South Asia," Ethnology 1976. It was, in essence, a literature review, working through available village studies to determine the conditions under which bridewealth (from the husband's family to the wife's) or dowry (from the wife's family to the husband) are paid, which turns out to have a lot to do with women's right or lack thereof to inherit a portion of their parents' estates. It is easy to imagine doing similar papers with different topics but the same basic strategy: working through ethnographic reports for the region in which you are interested, systematically cross-checking what they have to say on the topic you are interested in. The work may not have the buzz of currently hot theory; but if solidly done, it is likely to be cited by everyone who works on a similar topic in the future. A, what did we used to call that? contribution to knowledge.
John (Postill), could you tell us more about your idea of "compar[ing] media and nation-building efforts in two different regions." What factors would you be focused on in "media" and "nation-building"? What would you hypothesize about the relationship?
Once more into the fray. Phil Salzman has been asking for examples of comparative research. Serendipitously our discussions on OAC have coincided with an invitation to participate in a conference on my original research topic, Chinese folk religion, in Taipei this November. Off the top of my head I wrote the following abstract.

Title: Why do the gods look like that?: Reflections on the iconography of Chinese folk religion.

I begin with a simple question, one that a tourist might ask. After all, nothing is more distinctively Chinese than the architecture of Chinese temples and the visual representation of the gods who sit on their altars. Yet, precisely because it is so everyday to the fieldworker whose topic is Chinese folk religion, why Chinese gods and temples should look like that remains largely unexamined. Familiar descriptions of gods, ghosts and demons and the notion that gods are heavenly bureaucrats, albeit some with demonic histories, are at best thin descriptions of an iconography whose baroque elaboration contrasts sharply with, for example, the relative austerity of Japanese shrines. So the simple question points to more sophisticated problems. Can anthropological theories combined with a closer look produce thicker descriptions or more compelling explanations of the wealth of detail that dazzles and dulls the eyes of both tourist and fieldworker alike? Can it point us to deeper understanding of the fundamental phenomenology that Chinese folk religion embodies?

Thanks to Phil's interventions I have begun to recast the proposal in terms of levels of comparison. Consider, for example,

(1) my dissertation research on the symbolism of Daoist magic -- grounded in detailed comparison of the rituals in my Daoist master's repertoire -- a project consciously modeled on the linguist who analyzes the syntax of a language based on a corpus of utterances collected from a single informant.

(2) my 1990 article "Why don't we see some real money here?" Now the set of rituals was larger, including Buddhist, Daoist and Confucian variants of rites ranging from personal prayers to community celebrations repeated either annually or at longer intervals.

Still, however, in both these cases the comparison remained within a subset of Chinese culture to which the terms "religion" or "magic" might reasonably be applied. The argument in (2) was strengthened by reference of familiar Chinese uses of offerings of food and money to manage social relationships that might include or extend beyond "religion" and "magic." Ideas from The Gift by Marcel Mauss illuminated the patterns revealed in both ritual and everyday Chinese behavior, suggesting a higher level of abstraction at which the argument might be pursued.

Now, I come to (3), where the comparisons in question cut across what are usually taken to be boundaries between distinct cultures: China versus X, where X might be Japan, or medieval or modern Europe (I am thinking here of Erwin Panofsky's famous volume on Gothic Cathedrals and other uses of architecture to embody a cosmology). Can we simply say that any such comparisons are illegitimate because cultures conceived as distinct wholes must be understood in their own terms? I don't think so. The example of biology suggests that understandings at one level, e.g., gross anatomy, can be enriched by understandings at other levels, cells, mitochondria, proteins, chromosomes, genes, DNA and RNA, where each lower level extends understanding beyond whatever species we choose as our starting point.
Salzman asks,

What now? What place does comparison have in contemporary and future anthropology? And how would you use comparative analysis to advance your own research and understanding?

Here I would like to think a bit about Fred Egan's "method of controlled comparison," nicely described by Grant McCracken, in an application to pop culture in 2007 as, "one with enough similarities to make the contrast stand up and holler." Egan's method is one example of what Andrew Abbott calls "small-n comparisons" (a subject taken up with considerable technical precision in “On the Duality of Cases and Variables: Correspondence Analysis (CA) and Qualitative Comparative Analysis (QCA).” Pp. 243-59 in David Byrne and Charles C. Ragin (eds.), The SAGE Handbook of Case-Based Methods).

As originally conceived, the method of controlled comparison was intended to isolate variables of interest for the formulation of sociological laws. It may be, I suggest, more useful in the manner than McCracken describes. A case in point is Regime Shift: Comparative Dynamics of the Japanese Political Economy (Cornell Studies in Political Economy) by the eminent political scientist T.J. Pempel. Here Pempel describes the development of Japanese-style democracy by comparing it with that of the formation of of democratic institutions in several other countries. The comparison does not point to any sociological law. What it does, however, very usefully indeed, is highlight significant differences in democracy as actually practiced in a variety of advanced industrial nations, lucidly framing the political issues that affected how democracy developed in different places and the path-dependent solutions through which they have been addressed. Used in this way, comparison enriches our understanding of what is and isn't distinctive about the case with which he begins, democracy in Japan. That, to me, is an important achievement and a genuine advance in accumulating knowledge about a subject dear to our hearts.
Been meaning to engage in this conversation. Unfortunately, I didn't even read it, yet. So, apart from a few things I noticed, I don't know what ground has been covered.
Still, a brief idea on my stance, in terms of comparative analysis: because we work on human diversity, there's often an "embedded" form of comparison. People end up comparing phenomena with what they already know, and anthropologists occasionally fall victim to this. It's especially noticeable with students or while talking with non-anthropologists. Apart from the obvious links between this kind of embedded comparison and ethnocentrism, there's the issue that the comparison is done "unwittingly."
In such a context, comparative analysis would be a way to replace "embedded comparison" with something more explicit, careful, thorough.
At the same time, I remember having a knee-jerk reaction as an undergraduate student (18 years ago) when I heard a prof say that comparative analysis was the ultimate goal of anthropology. The discipline has changed quite a bit in the meantime, but I'm still left with a strange perception of comparative analysis.
So I should definitely engage in this thread instead of relying on first impressions. As strange as it may sound, writing this is a way to make sure that I will, in fact, start paying more attention to this thread.
I agree that if you are not comparing two other things, such as two other societies, you will compare the one you are looking at with the one you already know. Sometimes it is quite explicit, e.g. Margaret Mead's comparison of Samoa with America. Sometimes, as you say, it is inadvertent and embedded. But judgements, even if implicit, are not necessarily ethnocentric; not infrequently people approve of places that are different from home.

A comparison that started off my own research, was between different nomadic peoples: the northern Somali, described in A Pastoral Democracy, by I.M. Lewis, and the Basseri, decribed in Nomads of South Persia, by Fredrik Barth. I was struck by how different these folks were from one another, and began my comparative studies by trying to figure out why they were different.

I was led by thinking about Barth's Basseri to do research among the nomad tribes of Iranian Baluchistan, to see how the harsher environment of Baluchistan shaped the lives of the Baluch, and made it different from the lives of Basseri in the relatively more welcoming environment of the Zagros. More recently, in Pastoralists: Equality, Hierarchy, and the State, I built on that earlier comparison in a tour of a number of the nomadic peoples of Iran.

While I wouldn't say that comparative analysis is the "ultimate goal of anthropology," it is likely to be a major intellectual tool in pursuing whatever are the ultimate goals of anthropology.
Thank you for pulling me in this conversation through such a thoughtful reply, despite my lack of attention for what had already been said! This is exactly what I needed and it stimulates my thinking on the subject.
In fact, it connects with different ideas I've been having about the discipline.
Probably because of teaching experiences, I have this tendency to try to build the kind of working definition that can get us to discuss broad issues. For instance, I now define ethnomusicology as the ethnographic study of music, ethnography as a descriptive approach to cultural diversity, and anthropology as the study of human diversity. These definitions are lacking, in many ways, but they often enable people to discuss some important issues. If they disagree with my definitions, they can still come up with ways to package ideas so that it makes sense across a variety of perspectives. And discussing some implications of diverse definitions often represents an efficient to engage people in conversations. Especially among Anglos, I would say.

So, how does my tendency to build simplistic working definitions relate to the comparative nature of the discipline? It provides a context for my personal thinking about the issue.
Human diversity (through biology, language, time, and culture) almost implies comparability. Humanity as diverse means that some phenomena are universal while others are particular. The specificity is embedded in a "global" frame while universals are realized "locally." I'm mixing metaphors to connect with diverse perspectives.
The reason I care so much about the universal/particular model is that it often runs at the centre of distinctions between work done by anthropologists on either side of the Atlantic, especially those from the United States and France. As I was being trained in a French-speaking North American department, "comparativism" seemed connected with those universalist tendencies which are so easily recognizable in French anthropology, at least in terms of cultural phenomena. From a North American perspective, radical particularism almost precludes comparison. Partly because of the fear of ethnocentrism, but also because of the less-thoughtful portions of cultural relativism. Relativism itself doesn't make it impossible to compare, but relativists can be weary of comparisons because they seem to have to do with the notion of an absolute standard. The fact that they don't and that it's actually quite easy to do comparative analysis from a relativist perspective doesn't imply that North American cultural anthropologists are likely to undertake full-fledged comparative work of the kind that was done by their foremothers and forefathers.
It might be a projection on my part but it looks as though a number of people in ethnographic (sub)disciplines like folkloristics and linguistic anthropology were refraining from doing comparative work because they don't want to sound like Lévi-Strauss or Rouget.
There's also the issue of time and effort necessary to do proper comparative work. If one is to undergo classical ethnographic fieldwork in a series of places across the Globe in order to provide a valid comparison of, say, nomadic practises in different environments, the task sounds like a rather daunting one. Solutions range from doing "condensed" fieldwork (say, a few weeks at a time) to changing research design altogether. Different researchers have adopted different strategies but it seems, to me, that this is another place where comparative analysis has been made into less of a priority. And I'd say the same thing about "longitudinal analysis," which we may see as diachronic comparative analysis.
Of course, much of this relates to linguistic anthropology and cultural anthropology, if we think about North American departments of anthropology. As far as I know, biological anthropologists and archæologists do, in fact, undertake projects which are directly connected to comparative analysis. So it seems that the ethnographic parts of North American anthropology are the ones which are less enthusiastic about comparative work.
French ethnologie is interesting, in this respect. While comparative analysis is often a stated goal of anthropologie, and much of the work done by ethnologues is indeed comparative, fieldworkers may not contribute that directly to the discipline's comparative programme. In a way, their work is "humbler." But there's also a form of "pragmatic relativism" in fieldwork. It's really nice to be able to connect your work with something much broader. But as you're doing field research, much more of your attention may be caught in the analysis of the specific relationships between diverse phenomena you observe. In a way, it's just easier to remain sane if you think enter the field with modest, reasonable, and context-appropriate goals.

Then, there's the issue of comparisons on work done by diverse scholars. Is one's research comparable to somebody else's research? In ethnographic disciplines, a number of people might answer negatively. Perhaps as part of the backlash against reductionism in a positivist mode. But also because of a kind of "mystique" surrounding fieldwork. The fact that people from the outside of ethnographic disciplines tend to accuse ethnographers to be doing work which isn't comparative might be significant.

As I mentioned "embedded" comparison, I was mostly thinking about reactions to "otherness." Even well-trained cultural anthropologists may have those reactions and what ethnographers say is often perceived through the prism of difference. The comparison, then, does have a lot to do with what is considered "familiar" or "exotic."
But I realize that this thread is about something much broader and potentially very useful.

My way to put it is often in terms of dialogue, among ethnographers. Partly through books, but also through many other interactions, we get to learn about cultural contexts in which we're unlikely to ever do extended research. For instance, an africanist is likely to discuss diverse issues with colleagues working in other parts of the continent. In this sense, she/he gets to "know" a number of things about a variety of African contexts. Her or his idea of Africa is quite likely to include diverse things which have been said by colleagues working outside of the area(s) in which s/he has been doing field research. Because area specialization tends to be an important factor in grouping ethnographers, this kind of "intra-continental comparison" is quite common. It can be the object of direct comparative analysis (such as someone comparing Fula and Bamanan groups within the same area). But much of it is vaguer. Someone working in East Africa has had to read and hear about things related to West Africa and gets an image of the continent as being at the same time very diverse and relatively unified.
Outside of this kind of "area studies" mode, there's the fact that scholars working on the same broad topic interact with colleagues who work in a wide variety of contexts. Someone who works on foodways in Southeast Asia is likely to learn a fair deal about African foodways. In such a case, research done on a specific phenomenon leads to another kind of comparison. Local, regional, or continental specificity is put to the fore and, though quite distinct from HRAF-style categorization, a kind of "repertoire of cultural traits" gets to be built through interactions.

Much of this may sound like I'm rambling. I'm mostly "thinking out loud." An advantage of web forums is that they allow this kind of open conversation.
You have given us a lot to think about, Alexandre. Let me now address just one of your points, and I'll try to follow up on others later.

You said, "From a North American perspective, radical particularism almost precludes comparison. Partly because of the fear of ethnocentrism, but also because of the less-thoughtful portions of cultural relativism. Relativism itself doesn't make it impossible to compare, but relativists can be weary of comparisons because they seem to have to do with the notion of an absolute standard."

This is what I thought I had seen, and why I put up the discussion topic (in this same Theory of Anthropology group) "Do anthropologists now doubt that peoples and cultures are different from one another?" Contributors to that thread denied such a reticence. But yesterday in my seminar on "Comparative Anthropology," students rushed to distance themselves from Hackenberg's article "Economic Alternatives in Arid Lands: A Case Study of the Pima and Papago Indians," which, to be fair, does talk about agricultural development and take a somewhat evolutionary viewpoint, which can be seen to imply an invidious comparison or at least a value judgement. But the point of the article is to explain why and how people move to intensify agriculture, and it follows historical changes which are seen more in one group than another. Some students seemed to think, "How dare he say that?"
I too would like to compliment Alexandre for providing so much delicious food for thought. I would add just a bit of additional spice to the meal.

Alexandre writes

My way to put it is often in terms of dialogue, among ethnographers. Partly through books, but also through many other interactions, we get to learn about cultural contexts in which we're unlikely to ever do extended research.

I would add that for those of us who work in Asia, other ethnographers are often part of conversations that also include historians, in particular, but also sociologists and scholars from numerous other disciplines as well. For us the intersection of anthropological and area studies interests is so broad that they are often hard to distinguish. We are, as I have mentioned before, acutely aware of the situation described by Marcus and Fischer. We cannot ignore the fact that,

We step into a stream of already existing representations produced by journalists, prior anthropologists, historians, creative writers, and of course the subjects of study themselves.

Alexandre also writes,

Outside of this kind of "area studies" mode, there's the fact that scholars working on the same broad topic interact with colleagues who work in a wide variety of contexts.

Here is where I see exciting possibilities for a renewed enthusiasm for comparative research.

In graduate school, my generation of anthropologists was taught that the likes of Sir James George Frazer (The Golden Bough) and Joseph Campbell (The Hero with a Thousand Faces) had committed unforgivable methodological sins by tearing bits of myth and ritual out of their social and cultural contexts. But one can accept this criticism as fundamentally valid and regard their theories as simplistic (most theories are when people first survey a field of study) and still be intrigued by the observations they assembled, which do, in fact, demonstrate that certain recurring patterns have very broad, perhaps even global, distributions. For me a moment of dejá vu occurred when I was reading Paul Stoller’s Fusion of Worlds, an account of magic and sorcery among the Songhay in West Africa. The locale was certainly different, as were the ritual details I collected during my fieldwork in Taiwan, but I found myself saying to myself, “This all sounds very familiar.” I now find myself thinking that claims to esoteric knowledge and special powers, combined with accusing competitors of ignorance and/or evil-doing are probably generic characteristics of any market for ritual services on which state-sanctioned authorities have not imposed professional standards. There are also intriguing coincidences in details of magical healing. The most common form of healing ritual I studied in Taiwan is called siu:-kia: (catching frights); in both local etiology and treatment it seems remarkably similar to what anthropologists working in Mexico call susto. There is something about the notion of a person’s not being themselves because a bit is missing that needs to be returned.... In any case, I occasionally find myself speculating about the possibility of returning to a Frazerian-stye program but now being able to take advantage of far richer and more contextualized data on which to base more sophisticated comparisons....

RSS

Translate

OAC Press

@OpenAnthCoop

Events

© 2019   Created by Keith Hart.   Powered by

Badges  |  Report an Issue  |  Terms of Service