條件式推理抑制效果的機制

The Mechanism of Suppression Effects in Conditional Reasoning

本研究的目的是探討條件式推理的抑制效果在兩階段條件化（conditionalization）歷程（Liu, 2003）中的機制。所謂抑制效果是「附加」（additional）條件句抑制有效論證（modus ponens與modus tollens，以下分別簡稱MP與MT）的推理，而「另種」（alternative）條件句，則抑制推理謬誤（affirmation of the consequent與denial of the antecedent，以下分別簡稱AC與DA）的發生。心理模型論認為「另種」／「附加」條件句可使推理者察覺出反例（counterexamples），而拒絕MP與MT或避免AC與DA的謬誤。另一方面，心理邏輯論則認為「附加」條件句會使推理者懷疑原來的條件句，並將這懷疑傳遞到結論，使推理者拒絕接受MP與MT確定的結論。而本文探取Liu的兩階段條件化觀點闡釋抑制效果，Liu認為第一階段條件化中，結論在直言式前提上被條件化。而第二階段的條件化，是將上述結果在條件句前提上進行條件化。本研究透過兩個實驗，認為抑制效果的機制是「另種」／「附加」條件句在第一階段條件化中使察覺到的必要性／充分性受損，且加上「另種」／「附加」條件句後，將不會進行第二階段的條件化。因此，AC與DA謬誤與MP與MT推理的抑制，應是反映出第一階段條件化的結果。

The main purpose of this article lies in probing into the mechanism of suppression effects in conditional reasoning. A "suppression effects" is said to occur when the introduction of a second conditional premise of the form "if A, then Q" following "if P, then Q" leads to a significant decrease in the rate of endorsed conditional arguments. If "A" is necessary condition or requirement of "Q" occurring, then "if A, then Q" is called "additional conditional", which suppresses valid inferences (MP & MT). If "A" and "P" are alternatives of each other, then "if A, then Q" is called "alternative conditional", which decreases fallacies of invalid inferences (AC & DA). According to the mental model theorists, alternative or additional conditionals can offer reasoners available counterexamples, lead to reject the putative conclusion. However, the mental logic theorists deemed an additional conditional introducing uncertainty into the major premises, then the uncertainty transferred to the conclusion. Liu (2003) posed the successive-conditionalization approach for explaining conditional reasoning. In this approach, reasoners are assumed to compute the probability of the conclusion, conditionalizing first on the categorical premise, giving the knowledge-based component, and conditionalizing then on the conditional-statement premise, from which the assumption-based component is derived. In the light of the successive-conditionalization approach, almost all reasoners in the MP inferences, and some reasoners in the MT inferences could proceed two stages conditionalization. However in the case of DA/AC, reasoners can carry out first- stage conditionalization, but never achieve second-stage conditionalization. There were two experiments, in which Exp. 1 was manipulated for the fallacies, and Exp2 for valid inferences, investigate suppression effects in successive- conditionalization approach. The results of the two experiments found when the introduction of an additional or an alternative conditional, the ratings for MP/MT or DA/AC can be shown mathematically that the results of the second- stage conditionalization is the same as that of the first-stage conditionalization. In other words, the suppression effects of conditional reasoning can be said they only reflect the outcomes of the first-stage conditionalization.