My research was not aimed at drawing a history of cognitive "progress," which emerges through a juxtaposition of different 'stages' of thought (Barnes, Stages of thought: the co-evolution of religious thought and science 2000); if it were, then it would have made sense to begin with religion and the origins of culture, and track the "evolution" leading to science also through the consolidation of social cognition. Conversely, at this point of my research it was possible to frame religion through the conceptual tools developed so far, such as the modeling of external agencies and the construction of cognitive niches. Thus, religion can appear as a model of a class of inferences, traditionally perceived as irrational (or having to do with counterfactual beliefs), but which can be very interestingly studied by an epistemological analysis. Such an outlook can be regarded as analogous to the psycho-anthropo-cognitive effort to frame religion "as a natural phenomenon" (Boyer, Religion explained 2001; Atran, In gods we trust: the evolutionary landscape of religion 2002; Dennett, Breaking the spell 2006): similarly, my aim was to investigate religion as a "philosophical phenomenon" without adhering excessively to what is traditionally understood as philosophy of religion . The local interest was in fact to apply an epistemological toolbox in order to study and model the religious inferential regime, how religion could be defined as an ecological-cognitive activity, and how religious pragmatics can regulate behaviors that are typically connected with religion such as sacrifice and forgiveness.