The problem of unconceived alternatives in science and religion

Abstract This paper argues that the problem of unconceived alternatives (PUA), originally formulated as a much-noted intervention in the realist/antirealist debate about scientific theories, has notable implications for discussions of hypotheses concerning ultimate reality – particularly for the debate about so-called (alternative) concepts of God in both philosophy of religion and theology. Despite the substantial differences between scientific theories and concepts of God, or other hypotheses concerning ultimate reality, certain common strategies for establishing their central claims as true show surprising similarities in their vulnerability to the PUA. The main thesis advanced is that inferences that the central claims of a given concept of God are (probably or approximately) true are unreliable if, and to the extent to which, that concept of God is accepted on the basis of, and its central claims are arrived at and justified through, inferences to the best explanation or eliminative inferences. If the argument is successful, then if theological realism in the form of realist theism is to be maintained, the central claims of concepts of God must be based also on other epistemic grounds.