The Use of Cross-Cultural Research Methodology in the Study of Deep History
Almanac: History & Mathematics: Analyzing and Modeling Global Development
Cross-cultural research in anthropology involves the systematic coding of a representative sample of human societies on a large number of variables and the use of statistical method to test hypotheses about relationships among those variables. The method has allowed anthropologists to gain understanding a large number of processes that were involved in the formation of the world's thousands of sociocultural systems. Since the first development of this method there has been a strong interest in the role of shared history in producing shared variance among groups of sociocultural systems. This concern has been called Galton’s problem and is now often understood as being a form of autocorrelation (Loftin 1972, Dow et al. 1984).
Within the field of cross-cultural research, the shared history factor has often been treated as a problem that could be eliminated by improving the sampling methodology or by the use of statistical models that are designed to correct for the role of shared history. We have now learned that the degree of shared historical variance within human societies is so large that there is no way to eliminate it through a better sampling method. One productive approach has been to control for the effects of shared history through use of the network autocorrelation model (Dow et al. 1984). This model requires the researcher to specify a network of historical relationships among societies that are postulated to accurately measure the role of shared history. Within the field of cross-cultural research these historical relationships have mainly been measured in terms of spatial proximity and in terms of common membership within language families. Obviously it should also be important to also consider other kinds of historical relationships, such as shared membership in a world religion (Korotayev 2003, 2004) or incorporation in the same empire or empires.