Explaining Culture: 1st (First) Edition Paperback – November 1, 1996

Coming from a science background and now working in the humanities, I was initially critical of Sperber's application of empiricism to the social sciences. However, his book is an excellent and necessary resource for anyone studying the development and impact of culture or cultural phenomena. Though I take a different tack to the issues he raises, I will certainly refer to and discuss his ideas in my own future work.

Sperber uses the metaphor of epidemiology as an allegory to cultural transmission. Another reviewer seems to interpret this metaphor as an attempt at ONTOLOGY, that the essence of culture and cultural transmission is indeed epidemiological, culture is a disease -- this is, of course, a MISINTERPRETATION. Sperber's epidemiological explanation of culture is a HEURISTIC, a device to EXPLAIN culture, hence the title. It is a compelling and useful way to understand an otherwise cloudy subject, but it does leave aside the non-material component of culture.

Sperber wants to make anthropology and psychology partners in the construction of a theory of culture centered on `the epidemiology of beliefs'. Epidemiology examines the factors determining the frequency and distribution of diseases in a population. Similarly, the aspiring culturology will map the frequency and distribution of beliefs in a population.

The choice of epidemiology as the model science seems to be based on nothing more than the insinuations of English idiom. Idiom likens the spread of ideas to contagion. We say that ideas, moods, personalities, and fads are infectious. Rumor and disaffection spread like fevers through the body politic. Cheerfulness is contagious-smile and the world smiles with you. But usage provides no clue to causality. It is equally content with mechanical metaphors, such as the `band wagon effect' and the `climate of opinion', while outbreaks of frenzy, mania or hysteria are likened to floods, cyclones and wild fire. Idioms are heedless of the difference between plague and weather as transmission mechanisms. Oddly for an anthropologist, Sperber takes no notice of these clues to how the natives perceive thought transmission. An assessment must be made if we are to avoid confounding `good enough' idiomatic analogies with causal mechanisms.

My suspicion that epidemiology is a red herring deepened on reading Sperber's account of the new culturology. On pages 109 and 112 he introduces graphs representing the spread and transformation of beliefs under the influence of `attractors'. Attractors are characterized in two ways. In one statement, an attractor is `an abstract, statistical concept, like a mutation rate or a transformation probability' (p. 111). Not much is said about it. A cultural attractor, however, is a specific practice or model. Manners, rituals, architectural styles, and resource-rich environments illustrate. Sperber has more to say about cultural attractors. A piece of culture is likely to become an attractor to the extent that it is the shortest distance between an initial condition and a beneficial outcome. This concept is usually called `optimality', but the author calls it the `effect-effort balance', where the `processing of any given piece of information determines its degree of relevance' because behavior tends toward actions in which `the intended effect can be achieved at minimal cost' (p. 114). Many attractors are unique to individuals; others, as gene-linked algorithms, cut deep channels through all populations, e.g., critical learning times and courtship strategies. The stability of cultural practices, he advises, is due to the fact that they are `attracted' to these natural psychological channels and their presumed neural or genetic substrates.

Sperber provides a three page exposition meant to illustrate the difference between replication and transformation, and the stable combination of replication and transformation processes in a population. The combinatorial space is represented by a cellular matrix. He assigns cell types in some arbitrary quantity, and combinatorial possibilities to each type. The matrix now describes a combinatorial state space. An engine is needed to activate cell `growth'. Sperber doesn't say what the engine is, but once it starts, the initial random distribution of cells in the matrix begins to alter. With each generation (or turn of the engine's wheel), the distribution of cell types changes. Patterns emerge as iterations continue; eventually we see patterns aggregating around two attractors. What is happening here? Sperber's matrix reminded me of cellular automata, the discovery by John Conway that led to nonlinear interpretations of game theory. Cellular automata with simple combinatorial instructions programmed into computer graphics are capable of remarkable behavior. Some instructions yield homogeneity, some express fractal self-similarity, and still others cross the boundary between stability and chaos to bifurcate into ramified local structures in the basins of chaotic attractors. The engines of these transformations are recursive nonlinear equations. Could this be the inspiration of attraction theory? In footnote 34, p. 158 he writes: `Sophisticated notions of attractors . . . have been developed in complex systems dynamics [aka nonlinear theory, chaos theory, self-organization theory, fractals theory], and may well turn out to be of future use in modelling cultural evolution, but a very elementary notion of an attractor will do for the present purpose'. Oh dear! So much for `science'!

If Sperber's effort to raise a new science doesn't come off, does he present some concrete insights on the transmission of thought? I'm afraid the answer is No, at least for me. I found no discussion of recognized types of transmission-panics, crazes, cults, sports mania, medical scares, propaganda, advertising, mobbing, and the like. As for identifying the transmission microprocesses, his message is confused. Cultural germ theorists like Richard Dawkins don't identify the somatic process corresponding to infectious disease. Sperber has an alternative cognitivist position: he proposes that inferences mediate cognitive processing. But what do inferences operate on? On sensorimotor information. Many inferences are already `in' the senses. Here is the clue to the fugitive microprocesses obscured by epidemiology. The nonmetaphorical term is `communication'. Communication isn't pathogenic and medical models aren't relevant.

It seems to me that Sperber's culturology doesn't really get off the ground.