Processing Conversational Utterances

处理对话话语

• 批准号: 0131877
• 负责人: Thomas Holtgraves
• 金额: $ 19.43万
• 依托单位: Ball State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-08-01 至 2006-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0131877&HistoricalAwards=false
• 关键词: Processing; Conversational; Utterances

Conversation is the most basic site for language use; it is the primary site for language acquisition and most likely the context within which language emerged. Despite its fundamental importance, research on the cognitive processes involved in the comprehension of conversational utterances has remained relatively rare. Instead, past empirical research has focused primarily on how readers comprehend and remember narrative or expository text. However, conversations differ from texts in fundamental ways, and understanding how people process conversational contributions will require attention to these differences. This research program is an attempt to contribute to our understanding of the social-cognitive processes involved in conversation comprehension. This will be accomplished by focusing on conversational utterances as intention-based, interpersonal, collaborative acts and examining how these acts are produced and understood. A series of experiments employing a variety of measures (reading times, interpretation speed, memory) will be conducted to examine how, and in what way, people comprehend and remember the actions that speakers perform with their conversation turns. An important aspect of this research will be the development of a method for examining participants' (rather than observers') on-line processing of their conversations. This will be accomplished with the development of an artificial conversation agent that can simulate human language use.

会话是语言使用的最基本场所，是语言习得的主要场所，也是语言产生的最有可能的语境。尽管会话理解具有重要的基础性，但对会话理解中的认知过程的研究仍然相对较少。相反，过去的实证研究主要集中在读者如何理解和记忆叙事或非叙事性文本。然而，对话在根本上不同于文本，理解人们如何处理对话贡献需要注意这些差异。本研究旨在帮助我们更好地理解会话理解中的社会认知过程。这将是通过专注于会话语句的意图为基础的，人际，合作行为，并检查这些行为是如何产生和理解。一系列的实验采用了各种措施（阅读时间，解释速度，记忆）将进行检查，人们如何，以及以何种方式理解和记住的行动，发言人执行与他们的谈话轮。这项研究的一个重要方面将是发展一种方法来检查参与者的（而不是旁观者的）在线处理他们的对话。这将通过开发一种可以模拟人类语言使用的人工对话代理来实现。