Abstract: Purposes: Two experiments were conducted to examine the possible influence of college students’ perception of semantic relations between antecedents and consequents in five types of conditionals (if p then q) on conditional reasoning (MP, DA, AC, and MT). The general semantic relations between antecedents and consequents in conditionals are that antecedents are sufficient and nonnecessary for consequents, and that consequents are nonsufficient and necessary for antecedents. The five types of conditionals are contingence conditionals, causality conditionals, obligation conditionals, permission conditionals, definition conditionals. There is no previous study which concerns the influence of perceived necessities of consequents in conditionals on conditional reasoning. The authors proposed two assumptions. One was that that there would be a difficulty difference in perceived necessity of consequents in conditionals. Especially, the perceived necessities of consequents in permission and definition conditionals would be higher than those in contingence conditionals, causality conditionals, and obligation conditionals. The other was performance on MT would increase with perceived necessity of consequents in conditionals so that performances on MT in permission and definition conditionals would be higher than those in contingence conditionals, causality conditionals, and obligation conditionals. Experiment 1 examined the first assumption. Experiment 2 examined the second assumption. Method: In Experiment 1, 170 participants were randomly assigned to five groups, and each group completed the perception task of semantic relations in one of the five types of conditionals. In the perceptual comprehension task of one type of conditionals, the perceptual question representations of the four semantic relations of conditionals were respectively: “can p ensure q?” for the sufficiency of p for q, “is p necessary for q?” for the necessity of p for q, “can q ensure p?” for the sufficiency of q for p, “is q necessary for p?” for the necessity of q for p. In Experiment 2, 170 participants were randomly assigned to five groups, and each group completed the conditional reasoning task of one of the five types of conditionals. The reasoning problems of each conditional had the four reasoning formats (MP, DA, AC, and MT). For each conditional reasoning formats, participants were asked to choose from the three answer options (YES, NO, UNCERTAIN). For example, in the DA reasoning problems participants were asked to choose from the three answer options (YES, NO, UNCERTAIN) for whether consequents would occur given that antecedents occur. Results: The results of Experiment 1 showed that there was a difficulty difference in the perceived necessities of the consequents in the five types of conditionals, and the perceived necessities of consequents in permission and definition conditionals were higher than those in contingence conditionals, causality conditionals, and obligation conditionals. The results of Experiment 2 showed that performances on MT in permission and definition conditionals were higher than those in contingence conditionals, causality conditionals, and obligation conditionals, and performance on MT increased with perceived necessity of consequents in conditionals. Conclusion: The results of the two experiments showed that perception of semantic relations between antecedents and consequents in five types of conditionals influences conditional reasoning, and performance on MT increases with perceived necessity of consequents in conditionals.

Key words: conditional reasoning, semantic relation, perception, necessity of consequents, MT

摘要： 作者通过实验考察了大学生被试对五种充分条件句语义关系知觉对其条件推理的影响。两个实验的结果表明，被试对条件句中前后件之间语义关系的知觉理解影响和调节其相应的条件推理的成绩。实验1新发现，被试对充分条件句后件对前件必要性的知觉理解存在难易差异，对许可句和定义句的后件必要性容易知觉，对偶然句、义务句和因果句的后件必要性难以知觉。实验2新发现，被试对充分条件句后件对前件必要性的知觉理解的难易差异导致被试在否定后件式推理成绩上的差异，对许可句和定义句的否定后件式推理成绩显著高于对偶然句、义务句和因果句的否定后件式推理成绩，对充分条件句否定后件式推理成绩随对后件必要性知觉增加而增加。

关键词: 语义关系, 知觉, 条件推理, 后件必要性, 否定后件