Measurement of Cultural Complexity

When anthropologists differentiate cultures in terms of their relative complexity, they do not use the term complex in its ordinary, literal, or dictionary sense. What they imply, rather, is their status vis-2-vis one another with reference to one or more classificatory criteria which have been postulated to correlate with diderent levels or stages in cultural development. Examples are legion, e.g., literate as opposed to preliterate societies, food producers vs. food gatherers, sedentary vs. nomadic populations, state builders vs. stateless societies, workers in stone, bronze, and iron, Morgan's levels of savagery, barbarism, and civilization. Though frequently illus trated, such classificatory categories or sequences have rarely been established, much less tested, by quantitative scientific methods. A noteworthy exception is the scalogram analysis by Carneiro (I970), in which the incidence of several hundred traits was noted for IOO societies, which were ranked according to Guttman's technique in terms of the number of traits reported for each. The present authors have been stimulated by this study to test its results by diderent methods with the data on I86 sample societies assembled by the Cross-Cultural Cumulative Coding Center at the University of Pittsburgh. Instead of examining the incidence of individual traits, we have chosen to assess ten groups of comparable traits, each ordered according to a five-point scale of relative complexity.