Jesper Sørensen

In recent developments in anthropology and the science of religion the concept of culture has been criticised as being without any explanatory value. Thus, proponents of cognitive anthropology have argued that culture, instead of explaining anything by itself, is the thing in need of explanation. Dan Sperber and Pascal Boyer have independently argued that in order to understand the diffusion of cultural ideas, we need to pay close attention to two selective factors: domain-specific cognitive mechanisms and of pragmatic relevancy. This theoretical development has been successful in demolishing any remnants of a reified concept of culture and has, in contrast to postmodernist critiques, supplied a scientifically viable alternative. Their approach should thus be endorsed, but in their eagerness to promote a purely cognitive approach, Sperber and Boyer have failed to explain the relative local success and stability of some representations in contrast to other equally cognitive optimal representations. Why do some religious representations thrive while other, equally cognitive optimal representations, disappear? In order to address this problem, the metaphor of an epidemiology of representations, suggested by Sperber, is extended by an ‘immunology of cultural systems’. In addition to the selective forces described by Sperber and Boyer, the immunological approach argues that the relative success of new representations is largely dependent on how well they fit already existing cultural models. The success of religious ideas is not only a product of how often they are expressed and how well they fit cognitive structures. Their relative position in a constantly evolving ecology of ideas will have a substantial impact on whether they are transmitted to new individuals or whether they disappear together with countless other unsuccessful representations. Describing the systematic relations between public representations and their historical development are thus a necessary supplement to the epidemiological approach.