As the subtitle suggests, Cultural Materialism is both ambitious in scope and polemical in approach. Marvin Harris' goal is nothing less than to lay the foundations for the scientific study of society, and woe betide anyone who gets in the way! I could clearly visualise him riding the white horse of Science into battle under the banner of Cultural Materialism, wielding the holy sword of Operationalism against the dark forces of Obscurantism, Idealism, and Eclecticism. Sometimes this made me feel like laughing, but sometimes it made me feel like joining up myself...
Cultural Materialism is divided into two parts. The first part deals with cultural materialism itself, and contains a survey of the philosophy of science, an outline of the epistemological and theoretical principles of cultural materialism, and some examples that give an idea of its scope. The second part is a critical look at alternative theories, namely sociobiology and biological reductionism, dialectical materialism, structuralism, structural Marxism, psychological and cognitive idealism, eclecticism and obscurantism. The first part takes up about a third of the book, but the second part contains exposition of cultural materialism as well as criticism of alternatives, so the balance between construction and criticism is fairly even.
Despite a dislike of his polemical tone, my feelings towards Harris' cultural materialism are generally very positive, perhaps as a result of my hard science background; if I devote most of this review to criticism it is because my instinctive response in most places is to agree. The view of science on which cultural materialism is based is fairly non-controversial. Harris accepts modern criticisms of narrow views of the scientific method, but argues that science is still special in some way, that it is not "just another cultural practice". At the centre of cultural materialist epistemology is the distinction between the emic and the etic (roughly that between mental processes and symbolic structures on the one hand and observable physical and behavioural phenomena on the other) and a concern that the study of the latter take epistemological precedence over the former. This is not unique to cultural materialism, however, and the latter's distinguishing feature is a division of culture into infrastructure (ecological and biological constraints, modes of production), structure (kinship, politics) and superstructure (religion, art), coupled with a belief that this ordering reflects the dominant direction of causality.
My main criticism of the cultural materialist program (as set out by Harris) is of its narrowness of approach, and particularly its tendency to oversimplification via reduction. Rather than being a single criticism, this is in fact several, since it is applicable to the different levels of the program. First there is the reduction of knowledge to science. In many places Harris seems to be saying "science good; non-science bad". He does allow at one point that "the ecstatic knowledge of mystics and saints, the visions and hallucinations of drug users and schizophrenics, and the aesthetic insights of artists, poets, and musicians are certainly not obscurantist merely because they are not based on scientific research principles" (page 315). Nevertheless in other places he seems to assume that showing that something is non-scientific is enough to demonstrate its worthlessness, and it is not at all clear that he would be prepared to include literary criticism or Geertzian interpretive anthropology (to pick just two examples) in the list above.
Next there is the reduction of social science to cultural materialism. Harris quite obviously believes that cultural materialism (or other research programs with a similar stress on the etic infrastructure) is the only valid scientific research strategy for the social sciences (or at least for anthropology; see below). While he specifically states in several places that he accepts the autonomy of the emic superstructure in some domains, in practise he refuses to allow it any scope at all, repeatedly talking about "infrastructure probabilistically determining superstructure" and denigrating all predominantly emic approaches. So it is one thing to demonstrate that some food taboos have functional explanations; it is quite another to therefore assume that they all do, or even that such explanations are intrinsically more satisfying than symbolic ones.
This is my most fundamental point of disagreement with Harris. I see the relationship of cultural materialism to anthropology and the social sciences as like that of biochemistry to the biological sciences. That is to say, while it is certainly absolutely essential and critical, it is still one approach among many, and it neither spans the entire area of study nor subsumes all the other disciplines within itself. (One can do ecology without doing biochemistry, to continue the analogy.) This no doubt makes me an "eclectic" of one sort or another, but then perhaps Harris himself is an eclectic for thinking we need any disciplines other than physics to understand the universe.
A related issue is that, while Harris seems to be talking about social science generally in Cultural Materialism, almost all his examples are drawn from the anthropological literature and he is clearly writing for his fellow anthropologists. It is not at all clear how cultural materialism is supposed to relate to history or sociology or the other social science disciplines: are they to be incorporated into it, are they acceptable as separate disciplines, or are they to be discarded entirely? This is, admittedly, a question that haunts anthropology - how to reconcile links to demography and biology and the other sciences on the one hand with links to history and the humanities on the other? - and there aren't any easy answers. But throwing away all links to the humanities and trying to force anthropology into a purely "scientific" mold is a bit like having a frontal lobotomy in order to cure mild schizophrenia.
The final criticism is that Harris has a narrow conception of cultural materialism itself, and in particular a strong bias towards functionalist and adaptionist explanations. Despite claims to the contrary (page 60) this sometimes seems to assume a strong role for group selection and/or intentionality. Such is the case with arguments for female infanticide being due to warfare, or the purpose of no man's lands between villages being to provide "game reserves" (page 91). There seems to be an assumption that all cultural phenomena must have explanations that make them appear "sensible", and this sometimes requires the construction of "just so" historical explanations. A trivial example, but one that illustrates his sometimes almost desperate grasping for direct causal explanations, is Harris' passing suggestion that severer megafauna extinction in the Americas may have been a result of that continent's smaller size in comparison with Eurasia/Africa (page 89).
Harris' criticisms of other approaches to anthropology, while not as thought-provoking as his constructive ideas (it is always easier to criticise other peoples' work), are in some ways the most successful part of Cultural Materialism. While Harris' critique of sociobiology is pretty damning, even stronger criticism is possible from within biology itself, in the form of an attack on the narrowly adaptionist evolutionary paradigm employed by most sociobiologists. Developing such a critique might, however, have exposed some of the failings of his own cultural adaptionist and functionalist approach. The critique of the doctrinaire and idealist elements of dialectical materialism is to be applauded. However some "enlightened" dialectical materialist approaches (that espoused by Lewins and Lewontin in The Dialectical Biologist, for example) seem epistemologically more flexible than cultural materialism; in particular they allow a much clearer role for "downward causality".
Although I know too little of the other approaches considered to be able to evaluate Harris' criticisms, he appears to have done a very effective hatchet job on many of the sacred cows of anthropology (and the Indian "sacred cow" complex is one of his major examples). Most of his criticisms of structuralism simply highlight things that must have made all cautious readers of Levi-Strauss slightly queasy. I wasn't, however, convinced that structuralism is entirely without value. And while Harris finds many flaws in the various theories he lumps together as "psychological and cognitive idealism", these hardly constitute the death-blow he seems to think he has dealt.
In a couple of places his criticism of alternative approaches appears to be slightly circular, in that it takes them to task simply for not being cultural materialism. For example, when he writes "[structuralism] avoids implicating etic behavioural infrastructure ... and therefore cannot be a part of a materialist strategy" (page 166), this is clearly meant as a criticism, although what is really at issue is whether structuralism has managed to contribute to anthropology without involving the etic infrastructure. (In this case the argument is preserved from circularity by a whole pile of more cogent criticisms of structuralism, but the feeling of "preaching to the converted" lingers.)
Despite these criticisms I would not agree with Clifford Geertz':
"Though those with what they take to be one big idea are still among us, calls for a 'general theory' of just about anything social sound increasingly hollow, and claims to have one megalomanic." (Local Knowledge, page 4)Harris has performed an important service in laying out the cultural materialist program so forcefully, making both its strengths and weaknesses clear. Cultural Materialism is worth a read if you are at all interested in the theoretical foundations of anthropology. If, as seems likely, many people will violently object to both its premises and its conclusions, that should provide motivation for formulating alternatives.
Acknowledgements: thanks to Bruce D. Scott for convincing me to read Cultural Materialism and to Bob Graber for a thought provoking exchange on the anthro-l mailing list, parts of which have been used in this review.