Non-monotonicity, consistency, and rationality in human reasoning

非单调推理的逻辑框架扩展了经典逻辑，以便从知识或信念基础中证明不可行的结论是正确的。这些逻辑旨在阐明常识推理的理性核心，即日常情景中的人类推理。这个项目将在人类推理的认知建模和形式化、非单调逻辑的哲学基础的交叉点上解决。它将分析文献中讨论的非单调推理的逻辑方法允许表示人类演绎过程的程度，以及这些过程背后的合理性标准。为此，该项目将对聚合推理任务进行一系列认知研究，如经典的抑制任务和新开发的任务，这些任务将有助于分析人类推理中的一致性保持。在实证研究的基础上，我们将建立一个人类演绎过程的认知模型，并根据先前确定的合理性标准对该认知模型进行分析。该认知模型将在进一步的实验中进行评估，以评估其对人类推理中“理性”和“非理性”过程的预测能力。

Logical frameworks of non-monotonic reasoning extend classical logics in order to justify defeasible conclusions from knowledge or belief bases. These logics aim at making explicit the rational core of commonsense reasoning, i.e., human reasoning in everyday situations. This project will be settled at the intersection of cognitive modeling of human reasoning and philosophical foundation of formal, non-monotonic logics. It will analyze the extent to which logical approaches of non-monotonic reasoning discussed in the literature allow for representing human deductive processes as well as the rationality criteria that underlie such processes. To do this, the project will conduct a series of cognitive studies on paradigmatic reasoning tasks such as the classical suppression tasks and newly developed tasks that will help to analyze consistency preservation in human reasoning. Based on the empirical findings we will develop a cognitive model of human deductive processes and analyze this cognitive model with regard to the previously determined rationality criteria. The cognitive model will be evaluated in further experiments to assess its predictive power for “rational” and “irrational” processes in human reasoning.