Dynamic semantic representations in language production

This project focuses on spoken language production in healthy adult speakers and addresses whether semantic codes involved in speaking are 'static' (ie fixed and independent of any particular context) or 'dynamic' (ie adjusting to circumstances). The latter possibility has been suggested by recent research in language comprehension, but is at present largely unexplored in spoken production. Current accounts of speaking assume that a so-called 'message' in a speaker's mind serves as relatively 'static' input to the mental lexicon, allowing access to lexical concepts which comprise information about meaning relevant to the ultimate goal of speaking. The alternative 'dynamic' view proposes that the nature and degree of conceptual access varies according to task and situation. The project will test this possibility in a range of tasks commonly used in research on spoken language production (picture-word interference, picture naming with preceding primes, or in conjunction with a secondary task etc). These studies constitute an initial step in characterising the nature of conceptual codes in speaking. If the experiments show that semantic retrieval is focus- and context-dependent, similar studies could be used in other populations such as children or elderly persons, which might advance an understanding of how the semantic system develops across the lifespan.

本项目聚焦健康成年说话者的口语产生（spoken language production），旨在探讨言语表达中涉及的语义编码（semantic codes）究竟是“静态”的（即固定且独立于特定语境）还是“动态”的（即随情境调整）。后者的可能性已被近期语言理解领域的研究所提示，但目前在口语产生中仍未得到充分探索。当前的言语理论假设，说话者头脑中所谓的“信息”（message）作为相对“静态”的输入进入心理词典（mental lexicon），从而能够激活词汇概念——这些概念包含与言语最终目标相关的意义信息。另一种“动态”观点则认为，概念激活的性质与程度会根据任务和情境而变化。本项目将在口语产生研究中常用的一系列任务（如图片-词汇干扰任务、带前置启动词的图片命名任务，或结合次要任务等）中验证这一可能性。这些研究构成了刻画言语中概念编码本质的初步探索。若实验结果表明语义提取（semantic retrieval）具有焦点依赖性与语境依赖性，则类似研究可拓展至儿童或老年人等其他群体，这或将推动对语义系统在整个生命周期中发展机制的理解。