This chapter aims at laying the foundations for the study of religions as systems, which would enable scholars to produce formalized and quantitative explanations and predictions about the inner causal structure and possible developmental tracks of religions. Whereas the notion of a “system” has been formerly used in connection with culture and its various aspects (cultural systems, symbol systems, thought systems, belief systems and even ritual and religious systems), these accounts have not been based on a shared, formal, let alone mathematical, definition of systems and did not therefore provide scholars with appropriate tools to develop quantitative explanations and predictions about culture or religion.

Frustrated by the loose, metaphorical and ultimately not very productive talk about ‘systems’ in cultural studies, sceptics have recently raised their voices against too easily presuming the existence of systems where there might be none. For example, Pascal Boyer (1994: 229) has written about the false “theologism” that takes the existence of connections among religious assumptions for granted. Benson Saler (2001, 2005; and personal communication) has argued that beliefs do not constitute a system, because there are apparently numerous beliefs that we can remove from the set of an individual's or culture's beliefs without affecting any other belief. Cognitive anthropologist Roy D'Andrade suggested that culture is a collection of units, rather than an “entity”. Cultural items in the minds of people do not constitute “a thing” because they are lacking “entitativity”.