Thompson Valerie A, Byrne Ruth M J
Department of Psychology, University of Saskatchewan, Saskatoon, Canada.
J Exp Psychol Learn Mem Cogn. 2002 Nov;28(6):1154-70.
The authors investigated the relationship between reasoners' understanding of subjunctive conditionals (e.g., if p had happened, then q would have happened) and the inferences they were prepared to endorse. Reasoners who made a counterfactual interpretation of subjunctive statements (i.e., they judged the statement to imply that p and q did not happen) endorsed different inferences than those who did not. Those who made a counterfactual interpretation were more likely to (a) judge the situation in which p and q occurred to be inconsistent with the conditional statement and (b) make negative inferences such as modus tollens (i.e., approximately q therefore approximately p). These findings occurred with familiar and unfamiliar content, affirmative and negative conditionals, and conditional and biconditional relations.
作者研究了推理者对虚拟条件句（例如，如果p发生了，那么q就会发生）的理解与他们愿意认可的推理之间的关系。对虚拟陈述进行反事实解释的推理者（即，他们判断该陈述意味着p和q没有发生）与未进行反事实解释的推理者认可的推理不同。进行反事实解释的人更有可能：（a）判断p和q发生的情况与条件陈述不一致；（b）做出否定推理，如否定后件式（即非q所以非p）。这些发现适用于熟悉和不熟悉的内容、肯定和否定条件句以及条件关系和双向条件关系。
