| Abstract: | In their stimulating comment on Barrouillet, Gauffroy, and Lecas Oberauer and Oaksford argued that we gave no rationale for the assumption that the mental models of false-antecedent cases have a different epistemic status from the p q model and that no new computational level account of the conditional was provided to justify this move. The resulting lack of coherent principles would undermine the heuristic power of the mental model theory and lead to unacceptable consequences. Though our theory probably presents many weaknesses, it is false to assume that we did not give any rationale for the truth value gap of conditionals. Our proposal elaborates the distinction between initial and fleshed-out models that is at the core of the theory from its very beginning (Johnson-Laird & Byrne, 1991), and assumes that people consider those cases that match the initial model as making the conditional true whereas the cases that match the models added through fleshing out are judged as leaving its truth value indeterminate. (PsycINFO Database Record (c) 2008 APA, all rights reserved). |
