Suppose we give ourselves the goal of developing mechanistic and naturalistic causal explanations of cultural phenomena. (I don't believe, by the way, that causal explanations are the only ones worth having; interpretive explanations, which are standard in anthropology, are better at answering some of our interrogations.) A causal explanation is mechanistic when it analyses a complex causal relationship as an articulation of more elementary causal relationships. It is naturalistic to the extent that there is good ground to assume that these more elementary relationships could themselves be further analysed mechanistically down to some level of description where their natural character would be wholly unproblematic.