Evolution, rationality, and testability

James H. Fetzer

pp. 423-439

in: Synthese 82 (3), 1990.

Abstract

Cosmides, Wason, and Johnson-Laird, among others, have suggested evidence that reasoning abilities tend to be domain specific, insofar as humans do not appear to acquire capacities for logical reasoning that are applicable across different contexts. Unfortunately, the significance of these findings depends upon the specific variety of ‘logical reasoning’ under consideration. Indeed, there seem to be at least three grounds for doubting such conclusions, since: (1) tests of reasoning involving the use of material conditionals may not be appropriate for representing ordinary thinking, especially when it concerns causal processes involving the use of causal conditionals instead; (2) tests of domain specificity may fail to acknowledge the crucial role fulfilled by rules of inference, such as modus ponens and modus tollens, which appear to be completely general across different contexts; and, (3) tests that focus exclusively upon deductive reasoning may misinterpret findings involving the use of inductive reasoning, which is of primary importance for human evolution.