Comparative Functionalism: An Essay in Anthropological Theory
The arrogance of the British is really quite intolerable. Consider the present case. A distinguished American professor writes a short but elegant essay which his publishers extol to the skies as a major contribution to anthropological theory. In his opening chapter the author curtly dismisses the fashions in statistical comparison which are the current vogue among a high proportion of his American colleagues; the rest of the book is concerned with styles in sociological anthropology which were first developed in England. The author is well disposed towards these British theorists and he presents his own argument as a development from their positions. And the only reward that he gets for this display of international goodwill and freedom from academic chauvinism is the snide British comment: “But, my dear Sir, you are twenty-five years out of date.”
Apart from a few maverick characters who try to be all things to all men, professional anthropologists sort themselves out into two fairly well-defined groupsâ€”the would-be scientists and the humanistic arts men. We need not bother here with the latter. The scientists like to imagine that they can be as clinical and objective as chemists and physiologists; they may be dealing with human beings but that is a misfortune. They imagine that Culture is as natural and automatic as Nature itself; they presume the existence of laws of sociology as immutable as the law of gravitation; they can never admit that the material under observation might have a will of its own. But despite the apparent confidence of such writers, the analogy between natural science and social science is very weak. The natural scientist is able to make useful predictive statements because his model of the external world is precisely conceived and its component elements rigorously defined. In contrast, the social scientist has only the vaguest notion of the qualities which delimit his social universe, and the component categories by which this universe is described are either undefined or undefinable. It follows that the elaborate systems of statistical cross-cultural analysis which have grown out of the Human Relations Area Files at Yale are just eyewash. The constituent elements of the anthropologist’s factional analysis are not properly discriminated and do not in the least resemble the objectively distinguishable units for which statistical methods of analysis were originally designed. And without such discrimination the use of statistics is wholly inappropriate.
This much Professor Goldschmidt accepts. He agrees that anthropologists study social systems and not just agglomerations of institutions, but he would still like to be a scientist and thinks he can have the best of both worlds by offering us a modified version of Malinowski’s functionalism.
MALINOWSKI’S vituperative attack on contemporary anthropological fashion dates from around 1926. He categorically rejected the idea that we can usefully think of society as an assemblage of “traits,” analogous to a mixture of chemical elements, brought together by the accidents of history. On the contrary each culture is a total system, an entity sui generis. What should interest the anthropologist is not the uniqueness of the component parts but the way they fit togetherâ€”the way the system works. Cultures function so as to satisfy the biological and social needs of individual men. This is unquestionably true but it is also rather obvious; after all, if cultures didn’t do this there would be no men. Besides, once you have said this about one culture you have said it about all; where do we go from there? Malinowski himself was content with demonstration, which is laborious. He spent most of his last 26 years explaining the functional operation of a single society, that of the Trobriand Islands of Melanesia, and even then the job was hardly begun. Malinowski has had his imitators but most anthropologists feel that life is too short. Anyway what is the point? In the end Malinowski began to see the force of this sort of criticism; Trobrianders are not the only people in the world. So in a posthumous work, A Scientific Theory of Culture and Other Essays (1944),* he put forward a theory of comparative functionalism.” He there argued that although cultures, considered as wholes, are unique systems which cannot be compared, we can also distinguish functioning sub-systems (which he called institutions) and by studying these sub-systems cross-culturally we may hope to arrive at scientifically valid generalizations. This came close to a recantation from his original anti-historical thesis. He never put his idea into practice and none of his pupils has ever done so either.
The precise relationship between Professor Goldschmidt’s thought and that of Malinowski is not easy to specify. Goldschmidt explicitly takes Malinowski’s 1926 position as his point of departure, but he never mentions the later 1944 publication. This is very odd, for it seems to me that Goldschmidt’s general thesis is barely distinguishable from Malinowski’s and that a great deal of his argument (cf. the summary on pp. 131-131) derives more or less directly from Malinowski’s book. Worse still, several of Goldschmidt’s most personal theories seem to be rehashed versions of the very weakest links in the original functionalist doctrine. For example, his hypothesis about the role of sexual and parental affection in Nayar marriage duplicates what Malinowski said about the Trobrianders in 1925, an argument which has been fairly consistently rejected by anthropologists and psychologists alike. Then again Goldschmidt’s “discovery” that “hidden” psychological assumptions underlie the work of the British social anthropologists (p. 6) makes a British reader gasp. Why “hidden”? Malinowski’s assumptions could hardly have been more explicit; those of Radcliffe-Brown were spelled out at most excessive length in The Andaman Islanders (1922)
ALTHOUGH THE RESEMBLANCE between Goldschmidt’s essay and Malinowski’s 1944 publication is very close it is rather complicated. Malinowski used the term institution in a special private sense, as indicated above; Goldschmidt uses the same term but gives it a more usual, looser meaning (p. 61), without mentioning Malinowski’s different usage. On the other hand Goldschmidt’s special definition of the term group (p.65) is nearly identical to Malinowski’s definition of institution! Anyone who reads the two books side by side is likely to get properly foxed by the mutually inconsistent terminology.
Goldschmidt, like Malinowski, says that we should not compare similar institutions (in Goldschmidt’s sense) because in different cultural settings similar institutions can serve different functions, while similar functions can be served by different institutions. Goldschmidt also stresses Malinowski’s distinction between functions which satisfy basic biological needs and those which satisfy socially derived needs. But here he does break into new ground. Malinowski’s “derived needs” differ from culture to culture and they are therefore contingent. Just what they are contingent on is not very clear, but Goldschmidt suggests that a comparison of types of “ecosystem” might lead back to some useful generalizations about types of “contingent functional requirements.”
It all seems to me both naive and highfalutin. Malinowski’s functionalism was justly criticized for its subjectivity. Any assertion about “what was the function of what” rested on a purely intuitive judgment on the part of Malinowski himself. The “functions” which Goldschmidt is discussing seem equally impressionistic. And the whole argument is up in the air; we never come to any real ethnographic meat. What could be the practical use of such “theory”?â€”a rhetorical question to which Goldschmidt replies with further rhetorical questions posed at pp. 135-137. Potential readers might well start here. It does not need any profound acquaintance with linguistic philosophy to see that these pages might very easily be analyzed down to meaningless gibberish. Of course many people, including academics, enjoy a froth of mellifluous words, but personally I prefer Malinowski’s original book, which was unquestionably his worst.
Now in paperback, Galaxy Books (Oxford University Press), $1.75↩
Bloodletting January 26, 1967
Now in paperback, Galaxy Books (Oxford University Press), $1.75↩