Department of Philosophy, University of Konstanz, Konstanz, Germany; Institut für Psychologie, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany.
Department of Psychology, University of Zürich, Zürich, Switzerland.
Cognition. 2016 May;150:26-36. doi: 10.1016/j.cognition.2015.12.017. Epub 2016 Feb 2.
More than a decade of research has found strong evidence for P(if A, then C)=P(C|A) ("the Equation"). We argue, however, that this hypothesis provides an overly simplified picture due to its inability to account for relevance. We manipulated relevance in the evaluation of the probability and acceptability of indicative conditionals and found that relevance moderates the effect of P(C|A). This corroborates the Default and Penalty Hypothesis put forward in this paper. Finally, the probability and acceptability of concessive conditionals ("Even if A, then still C") were investigated and it was found that the Equation provides a better account of concessive conditionals than of indicatives across relevance manipulations.
十余年的研究已经为 P(if A, then C)=P(C|A)（“等式”）提供了强有力的证据。然而，我们认为，由于无法解释相关性，这一假设提供了过于简化的图景。我们在对指示条件句的概率和可接受性的评估中操纵了相关性，发现相关性调节了 P(C|A)的效果。这证实了本文提出的默认和惩罚假设。最后，研究了让步条件句（“即使 A，那么仍然 C”）的概率和可接受性，发现无论相关性如何，等式对让步条件句的解释都优于指示句。
