Processing speaker's meaning: Epistemic state, cooperation, commitment

该项目旨在提高我们对说话人含义的在线理解的认识，即识别并在必要时重建说话人在发表特定话语时的心理表征和意图。我们的实验语用学研究从两个不同的角度出发。研究的第一条线是探讨说话者意义在语言加工中的特定语用方面的作用，包括认知状态、合作、可靠性和承诺。第二部分旨在评估和推进语言理解的神经认知模型及其对语用加工的预测。在以往基于视角推理的研究中，我们已经看到说话人特定的认知状态会影响语言理解，离线真值判断中会考虑不同的知识状态，并且视角会随着时间的变化而变化。在这个项目阶段，我们计划进一步探索这些处理效应的来源以及基于视角的推理的时间动态。我们还想将对说话者特定效应的研究扩展到说话者可能会公然违反质量准则第一项（“不要说你认为是假的”）的情况，即当她被认为是臭名昭著的骗子或喜剧演员时，以及有理由怀疑说话者是否遵守质量准则的情况。至关重要的是，我们通过这条研究路线所追求的研究问题是，找出这些说话者特定特征的后果是并行计算的，并导致对合作说话者的预期的早期修改，还是质量推断是自动计算的，然后切换到非真诚模式。我们还研究了说话人的观点和真理条件承诺通过加工涉及缩小，排序和表达内容的传统含义。这将使我们能够识别处理不同含义的潜在相似性和差异性，这些含义可能是定性的或定量的。本项目是重点项目1727“xprg .de：基于实验证据的新语用理论”共同目标的组成部分，通过使用多种实验方法来解决语用处理问题。它将真值判断任务的行为数据与电生理学、鼠标跟踪和速度准确性权衡的时间敏感测量相结合。它关注说话人意义的新方面，并试图找出不同语用机制的共同点和差异。为了实现xprak .de的目标，在实时理解数据的基础上制定一种生物学上合理的语言体系结构也是至关重要的。

This project seeks to advance our knowledge of the online comprehension of speaker's meaning, i.e. the identification and if necessary reconstruction of the mental representation and intentions of the speaker when making a particular utterance. Our research in experimental pragmatics departs from two distinct points of view. The first line of investigation pursues the role of particular pragmatic aspects of speaker's meaning in language processing, including epistemic states, cooperation, reliability and commitment. The second line seeks to assess and advance a neuro-cognitively plausible model of language comprehension and its predictions for pragmatic processing. In previous research on perspective-based reasoning, we have seen that speaker-specific epistemic states affect language comprehension, that different knowledge states are taken into consideration in offline truth value judgment, and that perspective taking changes over time. In this project phase, we plan to further explore the source for these processing effects as well as the temporal dynamics of perspective-based reasoning. We also want to expand the research on speaker-specific effects to situations in which the speaker might be expected to blatantly violate the first submaxim of Quality ('do not say what you believe to be false'), i.e. when she is known as a notorious liar or comedian and to situations in which there is reasonable doubt that the speaker adheres to the maxim of Quality. Crucially, the research question we pursue through this line of research is to find out whether the consequences of these speaker-specific traits are computed in parallel and result in early modifications of what is to be expected from a cooperative speaker or whether quality inferences are computed automatically and a switch to a non-bona-fide mode is initiated subsequently. We also investigate speaker's perspective and truth conditional commitment through the processing of conventional implicatures involved in narrowing, ordering and expressive content. This will allow us to identify potential similarities and differences in the processing of different implicatures, which may be of qualitative or quantitative nature. This project is integral to the common goals of the Priority Program 1727 'XPrag.de: New Pragmatic Theories based on Experimental Evidence' by using diverse experimental methods to address questions of pragmatic processing. It combines behavioral data from truth value judgment tasks with time-sensitive measures from electrophysiology, mouse tracking and speed accuracy trade-off. It attends to new aspects of speaker's meaning and seeks to identify the commonalities and differences of different pragmatic mechanisms. It is further essential for the goal of XPrag.de to formulate a biologically plausible language architecture on the basis of real-time comprehension data.