When was structure functionalism?

I got into an argument recently with a senior colleague about whether or not Magic, Witchcraft, and Oracles Among The Azande was a structure-functionalist monograph. I argued that it wasn’t and that the book had more to do with Seligman (actually I said Westermark, but I meant Seligman) and his influence on Evans-Pritchard than with Radcliffe-Brown. Many people — particularly non-anthropologists — today remember MWO as the classical statement of the idea that witchcraft beliefs were epiphenomena of underlying conflicts in social structure. This is indeed a textbook structure functionalist approach to witchraft — it’s just that Monica Wilson is (afaik) the person who articulated it, not Evans-Pritchard! And of course by 1961 Evans-Pritchard is producing pieces like History and Anthropology. The more I thought about it, the more it seemed to me that all of Evans-Pritchard’s works were not really so much structure functionalist as just structuralist, and he appeared less and less post-R-B and more and more pre-Needham. Of course Evans-Pritchard was always his own thinker, and it’s been a long time since I’ve revisited his work, so perhaps I’m off base here.

Thinking about it more broadly, however, structure functionalism is a brand that continues to sell even though the label covers a multitude of different approaches. As Stocking points out in After Tylor, by the time that structure functionalism came together as a program it was already drifting apart in other directions: Mancusian Ma(r)xism, Hocartesians, and (slightly later) Barthian transactionalism, etc. etc. Still we — and perhaps here I just mean clueless Yanks — continue to use this term today. As a ‘brand’ structure functionalism — and particularly the work done in Africa — in general seems to sell.

I was recently reading a book which discussed the history of anthropology. The book — which to be fair was just giving a summary — said that anthropology began with a spurt of important work in the Pacific (Haddon, Malinowski, etc. etc.), which then gave way to sustained work in Africa. I was flabberghasted — not only does this sort of summary ignore that fact that the British worked in places outside of Africa, it totally ignores the fact that for many (indeed, probably most) anthropologists the paradigmatic ‘field location’ is America — and particularly Morth American! It was as if Boasian anthropology (or its French connection via Levi-Strauss) just hadn’t happened.

Partially this is because, as many people have pointed out, of the fact that while both Boas and Radcliffe-Brown were institution builders, Boas didn’t make programmatic statements the way that R-B did. Which is not to say that the Boasians didn’t have a program. It’s just that they didn’t sell it the way R-B did.

So… when was structure functionalism again? I feel to a certain extent the brand continues along today, decades after its main exponents stopped writing, and continues to hide a multitude of different approaches.


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 rex@savageminds.org

16 thoughts on “When was structure functionalism?

  1. I’ve always found Evans-Pritchard more interesting to read than Radcliffe-Brown, and so I have to admit that my knowledge fo R-B is rather limited. Maybe I’m simply more of a stucturalist than a functionalist myself? Perhaps we should put up one of those ever-so-popular web quizes where you can find out what brand of anthropologist you are. You could get a little code to stick on your blog which says something like: “I’m 60% Structuralist, and 40% Post-Stucturalist”

    On a more serious note, I’ve never really taken the whole idea of “structural-functionalism” as a coherent paradigm seriously. That’s because, at its heart, I’ve always seen it as less of a coherent theory than a collection of analogies to work being done in other sciences: especially linguistics and biology. Which isn’t to say that such analogies didn’t produce interesting insights into other cultures – just that I’m hard pressed to say what a coherent program based on S-F would look like. But, like I said, I’ve never read much R-B…

  2. Since it has been a long time since I thought about these issues I did a Google search for “Anthropology intellectualist approaches” that took me to an article by Jesper Sorenson, who, writing from a British perspective and reviewing traditional approaches to the anthropological study of ritual, draws a distinction between two basic schools: intellectualist and symbolist described as follows:

    The intellectualist approach understands ritual as a type of rational behaviour based on explanatory principles that formally, if not substantially, are indistinguishable from those of Western science. To put it crudely, the intellectualist approach does not distinguish the behaviour of ‘planting’ from that of ‘chanting’ if both are backed by explanatory theories arguing for their necessity in some practical endeavour, in this case agriculture. Both are seen as rational instrumental or technical actions in relation to a theoretical background. They are ‘rational’ as the behaviour can be explained and defended from the theory underlying it. Thus, magical practices in socalled primitive societies are not irrational and nonsensical, but are merely based on flawed theoretical premises.1
    In contrast to the intellectualist position, the symbolist approach does not accept that ritual should be seen as instrumental actions based on theoretical presumptions about the world. There is no direct causal connection between underlying beliefs and ritual actions, and ‘chanting’ should not be understood as motivated by the same type beliefs about the world as is the case with ‘planting’. Rituals should instead be understood as expressive behaviour that communicates certain meanings, notably about social structure, coded in symbolic language. Thus, the job of the observer is to acquire the interpretational ‘key’ necessary in order to decipher the message inherent in this apparently irrational or wrong action.2

    In his footnotes, Sorenson provides the following examples.

    1 Representatives of the intellectualist approaches to ritual and religion are E. B. Tylor, Primitive Culture: Researches into the Development of Mythology, Philosophy, Religion, Art, and Custom
    London: John Murray, 1871); J. G. Frazer, The Golden Bough: A Study in Magic and Religion, 3rd edn, 13 vols (London: MacMillan Press Ltd., 1911–15), I: The Magic Art and the Evolution of Kings (2 vols); I. C. Jarvie and J. Agassi, ‘The Problem of the Rationality of Magic’, in British Journal of Sociology, 19 (1967), 55–74; J. Agassi and I. C. Jarvie, ‘Magic and Rationality Again’, British
    Journal of Sociology, 24 (1973), 236–45; R. Horton, ‘African traditional thought and Western science’, in Rationality, ed. by B. Wilson (Oxford: Blackwell Publications, 1970), pp. 131–71; J. Skorupski, Symbol and Theory: A Philosophical Study of Theories of Religion in Social Anthropology (Cambridge: Cambridge University Press, 1976).
    2 Representatives of the symbolist approach to ritual and religion are M. Mauss, A General Theory of Magic, trans. by Robert Brain (London: Routledge & Kegan Paul, 1972; 1st edn (French) 1902–3); E. Durkheim, The Elementary Forms of the Religious Life, trans. by Joseph Ward Swain (New York: Free Press, 1965; 1st edn (French) 1915); J. Beattie, Other Cultures (London: Cohen & West).

    I was reminded, while scanning Sorenson’s article, of Evans-Pritchard’s own Theories of Primitive Religion in which E-P identifies three types of theories: intellectualist (attributed to Tylor and Frazer and said to be composed of “If I were a horse” stories), structural functionalist (attributed to Radcliffe-Brown, in which function is related to social structure) and psychological (attributed to Malinowski, in which function is related to psychological, primarily emotional, needs).

    All this is, of course, a highly British perspective, which excludes from the category “theory” the writings of American anthropologists, which were, at the time that structural functionalism was flourishing, largely written off as examples of atheoretical “threads and patches” description.

    That said, there is (speaking to Kerim) a substantial body of anthropology developing the structural-functional paradigm by incorporating into it the dynamics of social and psychological conflict theories rooted in Marx, Freud, and Simmel. I speak here of the Manchester school (Gluckman, V. Turner, etc.) and also of Mary Douglas’ work, especially Natural Symbols.

    In the framework provided by this understanding, MWO is seen as a step away from structural functionalism to the kind of intellectualist theory later exemplified by, for example, Robin Horton.

    Hope this is helpful.

  3. Googling “Radcliffe-Brown Confucius” I stumbled upon one of my own posts to Anthro-L.

    The curmudgeon (who used to teach courses on Religion, Myth, and Ritual Symbolism) speaks:

    A good argument can be made that the strong tendency in even social scientific descriptions of religion to privilege belief and to treat belief as the cause of religious behavior is an artifact of the Protestant reformation, which radically de-emphasized the communal and institutional context of ritual in favor of a one-on-one relationship between the Believer and his or her monotheistic God in which Faith is the key operator.

    There are, however, alternatives. In the Analects, Confucius says, for example, that a gentleman participates in ritual “as if the spirits were there” but does not concern himself with whether or not they actually exist. He is, then, a progenitor of the functionalist school whose Western lineage runs from Robertson Smith (Religion of the Semites) through Durkheim (The Elementary Forms) to Radcliffe-Brown and Guy Swanson, in which ritual behavior is seen as primary and belief as
    a (much distorted) reflection of what is done ritually instead of being its cause. (Radcliffe-Brown, by the way, explicitly mentions Confucius.)

    More nuanced views see belief and behavior as systemically or dialectically related, with one feeding off the other and influence moving in both directions.

  4. Kwame Anthony Appiah writes of a passage in Oracles, “what he was teaching us was that what you see depends on what you believe, What it’s reasonable for you to think, faced with a particular experience, depends on what ideas you already have” (Cosmopolitanism, p.39). Now that sounds like something Boas would teach. Is it a wrong lesson to draw from Evans-Pritchard? I seem to recall being taught something similar about Oracles, but its been many, many moons.

    Regarding your urgent question, go ahead and distinguish between Evans-Pritchards the structuralist and the structural-functionalists. Look at Evans-Pritchards’ students. Who is more of a student of Evans-Pritchard, Max Gluckman or Mary Douglas? I don’t think there’s any single right way of looking at it. I’m sure Monica Wilson knows what she’s talking about. There’s a lot to be said your viewpoint as well. I don’t think a terminus post quem will reconcile the two perspectives.

  5. “what he was teaching us was that what you see depends on what you believe, What it’s reasonable for you to think, faced with a particular experience, depends on what ideas you already have”

    Sounds a bit off to me. The way I learned it, E-P pointed out that the questions the oracle answered were different from those that a scientist would ask. The scientist and the Azande could agree on the how of an event (The termites chewed up the granary’s leg, which collapsed causing the granary to fall on X). The Azande, however, also wanted to know, “Why X? Why at this particular moment?” The answer was witchcraft. Given similar questions, a strict Calvinist might say Predestination, a Chinese Taoist healer, like the one I studied with in Taiwan might assert that, given astrological significance of X’s birthdate, his being in that time and place was offensive to the God of the Year.

    One might argue that science is science precisely because it rules such questions out of bounds and radically simplifies the list of possible questions to include only those answerable by focusing on the how instead of the why, the area in which individuals with radically different cosmologies can still find common ground.

  6. Thanks, John. The granary example rings a bell or two. To be fair to Appiah, he’s talking about the bit where Evans-Pritchard questions whether he believes he sees a witch. (I don’t have the text handy to see if Appiah’s reading is accurate, but I’m assuming that it is.) If that bit doesn’t make it into lesson plans, would that be a matter of radically different cosmologies? Of course I’m kidding. It’s hardly out of bounds to ask about why Evans-Pritchard chose the examples he did, or why different readings emphasize different passages. In most cases I’m sure the granary example is most appropriate for study in an anthropology classroom. But if after many years, or coming from a different angle, one merely takes away that Azande people are no less intelligent than the people one knows, or Appiah’s more refined point that people reason based on what they already know, which is largely given to them by the cultural world they grew up in, I wouldn’t think that would be objectionable.

  7. What I think is so interesting here is that while we all know at some level that ‘structure-functionalism’ is not a label with tons of analytic power, we all still use it — even if only to say things like “kinda structure functionalist, but really more intellectualist.” But I rarely hear people say “I’m a Boasian” or “basically a Boasian, but in a Weberian way.” In the post I suggested this was because of the enduring power of programmatic statements by people like R-B. But the more I think about it, I think it has to with the popularity of Kuper’s _Anthropology and Anthropologists: The Modern British School_. We don’t have a similarly potted history of US anthros — Darnell’s _Invisible Genealogies_ attempts this, but is too formally written and (in my opinion) too nuanced to serve as a bible for theory newbies.

  8. I dunno Rex, I think I’d be happier describing myself as “basically Boasian but in a Weberian kind of way” than as any kind of functionalist. But as for structural functionalism, what strikes me is not so much the persistance of the label (especially as self-description) but the persistance of the practice, post la lettre: and the reasons for that stike me as being as much cultural as anything else. Now, it is true, that in talking aof the persistance of the practice I am speaking loosely, taking as structural-functionalist all that work that accounts for cultural phenomena by their role in preserving the exisiting social structure–in that loose sense, both some parts of Foucault and much contemporary anthrropolgical Marxism show affinities to “strcuk-fcuk” (alert–T-Shirt idea!). I think that this reflects things like a sense that the world is shaped by human self interest, a sense that the material is more real (or more efficacious) than the symbolic, and a discomfort with the idea that contingency plays a large role in history.

  9. Well I think most people are “Boasian in a Weberian kind of way” but we don’t have the same self-consciusness of this being an available label as we do structure-functionalism. Or perhaps I’m wrong. Do people have a sense of what it means to be strongly ‘Boasian’ and do they describe themselves this way?

    The strcuk-fcuk tshirts MUST happen.

    As for recycled structure-functionalism, I whole-heartedly agree. Witchcraft is a good case in point: it used to be it was caused by ‘social structure’ but “now we know it’s caused by ‘Modernity’ but don’t worry this isn’t a structure functional explanation because we use the word ‘dialectic’ a lot when describing it.”

  10. It’s always funny to me to hear things about labelling oneself within a particular theoretical stream. At Concordia, even at the graduate level, I never felt any push to identify strongly with any particular theoretical “genre”. We were actually encouraged to be ecclectic. There was, admittedly, a strong feminist/PoMo influence when I was there but not to the extreme of rejecting all other modes of anthro-thought. It was seen as quite productive to create our own unique blends.

    So there you go: to this day, I don’t adhere to any particular theoretical stream, although I do identify as somewhat PoMo (yeah, yeah, laugh it up) and feminist with some quite pronounced disposition toward ecclecticity (is that a word???). In terms of methods, though, I strongly identify with radical participation/experiential anthropology. Of course, there is a whole theoretical approach that is intrinsic to that but it’s not exclusive of others.

  11. Rex,

    I’m curious about what you mean when you describe yourself as Boasian but in a Weberian kind of way.

    For better or worse, structural-functionalism as described by Radcliffe-Brown had an explicit program: Step 1 was to analyze social structure, Step 2 to understand how something else, rituals, say, or conventional sentiments related to social structure. There was also the assumption that research conducted in this manner would contribute to a science, conceived in a classic positivist manner as a body of generalizations both empirically grounded and logically connected—generalizations that would transcend any particular case.

    The example that always comes to my mind is “The Mother’s Brother in South Africa.” Here the generalization that in a patrilineal society with virilocal marriage the mother’s brother can be expected to have a special relationship to his sister’s children was derived from South African data but supposed to apply more widely, even outside of Africa. I remember this vividly because it led me to notice that in Taiwan, the mother’s brother both has a special role to play in his sister’s wedding and a distinctive ongoing relationship to his sister’s kids. Noticing this gave me an answer when a Maryknoll missionary who had asked me what anthropologists with only a year or two of fieldwork could possibly teach him, who had been in Taiwan for fifteen years and spoke Taiwanese fluently. A hah, I thought, there is some point to all that stuff I read in graduate school.

    It may have been true that after a while, structural-functionalist anthropology came to suffer from what Evans-Pritchard labeled “the dead hand of competence.” The umpty-leventh description of lineage organization didn’t have the pizzazz of the first. But there was something solid there.

    But, I have to confess, when I read “Boasian in a kind of Weberian way,” I don’t know what that is. I can only imagine that it has something to do with piling up lots of ethnography about particular peoples and trying, in each separate case to come up with some ideal types, conceived as mental models, to account for the observations. Is this anywhere close to what you have in mind?

  12. Thanks, Rex. Let me add here that this is not a rhetorical question for me. My training in mostly British style social anthropology had left me a believer in the elements of the structural-functionalist paradigm I’ve mentioned above, i.e., start by understanding local social structure and then interpret other phenomena, rituals, folklore, fashion, whatever, in relation to that context. Thanks to a deep immersion in Manchester School studies by Gluckman, Turner, etc., my notion of social structure was more dynamic, more process oriented, than the static institutions described by Radcliffe-Brown; but the implications for fieldwork remained much the same. When I went off to do my fieldwork in Taiwan, the usual project was still a village or neighborhood study. The major alternative was studies based on the spatial analysis of regional systems pioneered by G. William Skinner. Neither was a good fit with what I actually wound up doing—becoming a disciple of a Taoist healer and traveling with him all over central and northern Taiwan to watch him perform rituals. The model to which I turned was linguistics, where a linguist can base a grammar on a set of hundreds of utterances collected from a single informant, with results cross-checked with other informants. My thesis became, in effect, a syntatic analysis of the healer’s ritual repertoire; but like a linguist’s grammar, it was analytically divorced from the specific social contexts in which the rites were performed and, thus, a departure from the structural-functionalist paradigm in which I had been trained. Having made the choices that lead to my writing the thesis I did left me with an abiding interest in the relationship between theoretical ideas and the research paradigms they imply.

  13. As the other person who endorsed the label, I’m thinking about it too–it a hard but very good question.

  14. Yeah. The problem isn’t figuring out what “Boasian/Weberian” means — I have a strong gut feeling about that. What is hard is how to explain it to other people.

    For instance, I think the idea that you “understand the local social structure and then interpret other phenomena in relation to that context” is pretty vague — it describes huge swaths of social science, from new institutional economics to Durkheim. To be classical ‘structure functional’ you’d have to add several additional features: the existence of externally bounded, homeostatic social systems as natural kinds, analyzing ritual etc. as epiphenomenal of social structure, a lack of attention to history, etc. etc.

    The problem is specifying how you saturate a general approach of ‘see things in light of their general social context’ with a particular analytic, and explaining that in the case of a Boas/Weber hookup is the hard bit.

  15. Philosophically speaking I agree completely. But my concern is less philosopical than pragmatic. You go into the field. Where do you start? The structural-functional paradigm was pretty straightforward: Start with maps, hut diagrams, a census, genealogies. Describe the statuses and roles involved in local politics. The result is the skeleton on which you hang the rest of your analysis.

    In Puli, the market town where I did my fieldwork, that would have meant trying to sort out the relationships of 35,000 people, plus as many more in the surrounding countryside. And most of these people lived behind brick walls or spent a good deal of time in buses and taxis traveling here and there. The practical result was that any references to “Chinese social structure” were to theoretical constructs developed by reading about the subject in mostly secondary sources, or asking local folks who mostly replied in terms of idealized notions instead of specific cases. It wasn’t the kind of immediate, solid grounding that Vic Turner’s understanding of Ndembu social life in the villages he studied brought to his analyses of particular ritual events.

    So I find myself wondering, what does a Boasian/Weberian try to find out first. Then, where does he go from there? How do you translate that theoretical stance into a practical program if all you have is at most a year or two to learn enough to say something valuable?

Comments are closed.