SFB 991: The Structure of Representations in Language, Cognition and Science

CRC 991调查表征的结构在语言，认知和科学。它将语言学、哲学项目、神经科学和实验心理学等广泛领域的研究结合起来。其出发点是假设在神经层次、认知层次、语言概念层次和制度化概念层次下存在一个统一的表征结构。这个统一的结构就是框架，CRC关于框架的概念是受到认知心理学家L. W. Barsalou的启发。从这一假设出发，到目前为止，CRC已经解决了以下问题：不同名义概念的特征；事件框架的结构；词形复杂词的框架（去）构成基于框架的形态学范式表征框架构成及其与词法、信息结构和语篇的衔接概念操作，如强制转换；歧义、模糊、多义、习惯阅读与字面阅读；基于语料库的框架归纳；感觉-运动系统中的基础概念；科学哲学的基本问题；基于框架的原型表示；贝叶斯分类学习与框架；哲学史上的框架；以及表征结构在大鼠社会认知中的作用。在第三阶段，这些主题将继续发挥作用。此外，CRC将扩展其研究到以下几个方面：a)框架-内部结构，区分不同的视角和不同层次的抽象，例如对修改的约束，强制现象以及字面意义和比喻意义之间的关系；B)概念分类和定义框架类型及其层次关系的方法；C)大鼠社会行为所反映的框架结构和认知表征性质的神经和认知相关性的实证研究；D)基于语料库的框架归纳的无监督和半监督方法；E)基于框架的话语处理的统一理论。

The CRC 991 investigates the structure of representations in language, cognition, and science. It unites re-search in a broad range of areas in linguistics with philosophical projects and with research from neuroscience and from experimental psychology. The starting point is the hypothesis that there is a uniform structure of repre-sentations underlying the neural level, the cognitive level, the level of linguistic concepts and the level of institutionalized conceptions. This uniform structure is frames, where the CRC's notion of frames is inspired by the work of the cognitive psychologist L. W. Barsalou. Starting from this hypothesis, the CRC has so far addressed the following issues: the characteristics of different nominal concepts; the structure of event frames; frame (de)composition in morphologically complex words; frame-based representations of morphophonological paradigms; frame composition and its interface with morphosyntax, information structure and discourse; conceptual operations such as coercive shifts; ambiguity, vagueness, polysemy and idiomatic versus literal readings; corpus-based frame induction; grounding concepts in the sensory-motor system; fundamental issues in philosophy of science; frame based representations of prototypes; Bayesian category learning and frames; frames in the history of philosophy; and representation structures in the social cognition in rats.In the third phase, these topics will continue to play a role. Furthermore, the CRC will extend its research to the following aspects: a) frame-internal structures that distinguish different perspectives and different levels of abstraction, relevant for instance for constraints on modification, for coercion phenomena and for the relation between literal and figurative meaning; b) conceptual classification and ways to define frame types and hierarchical relations between them; c) empirical investigations of neural and cognitive correlates of frame structure and of the nature of cognitive representations reflected by social behavior in rats; d) unsupervised and semi-supervised approaches to corpus based frame induction; e) a unified theory of frame-based discourse processing.