Debate over the relationship of religion to death anxiety has included the opposing views of Malinowski, who held that religion lessened death anxiety, and Radcliffe-Brown, who argued that religion increased death anxiety. Homans' theoretical synthesis of these viewpoints was tested by Leming, who concluded that the empirical relationship was curvilinear, meaning that both high and low religious involvements resulted in low death anxiety while middle-range attachments did not.

Reconsideration of this result argues that the presence of death anxiety is not dependent upon social learning, and that either high or low levels of theism leads to the resolution of anxiety problems. This outcome forces a contrast between religion generally and theism specifically, refuting their conventional equation. Experimental curvilinearity suggests that non-theistic or implicit religions both exist and are theoretically productive for the mainstream concerns in the study of religion. This outcome counters contrary claims from conventionalists who deem implicit religion as mere analogy or a peripheral subclass of little theoretical importance.