The unreasonable effectiveness of reasoning in the sciences

Reasoning is a capacity for producing and evaluating reasons during communication; its evolved function is persuasion, not the discovery of empirical truth. This cognitive theory, the argumentative theory of reasoning, seems at odds with the fact that, in science, reasoning does enable the discovery of empirical truth, and does so reliably. This paper explains the discrepancy. Scientists reason with the ordinary, biased, persuasion-oriented capacity shared by all humans; what makes their reasoning apt for the discovery of truth, however, is the cultural setting in which it is exercised: an ecology of reasons that a scientific community produces and maintains, together with the way that community distributes evaluation and the incentives it attaches to producing new claims. We set out six features of that setting which make reasoning in science so effective: three bear on the evaluation of reasons and the warrant of the beliefs that survive it; three bear on the generation of new reasons and the creativity of science. We thus explain how the psychology of reasoning grounds and specifies the observations of social epistemology, which emphasizes the social processes in scientific knowledge production.