Presuppositions in Online Language Comprehension

在线语言理解中的预设

• 批准号: 1349009
• 负责人: FLORIAN SCHWARZ
• 金额: $ 36.6万
• 依托单位: University of Pennsylvania
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-09-01 至 2018-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1349009&HistoricalAwards=false
• 关键词: Presuppositions; Online; Language; Comprehension

In interpreting linguistic utterances, the human mind has to piece together many different types of information. While this takes place sub-consciously, language researchers have developed theoretical perspectives on how distinct aspects of meaning are factored into the overall interpretation. However, little remains known about how these theoretical constructs relate to actual cognitive processes taking place in real-time. The present project will help to fill this gap by investigating aspects of meaning, specifically presuppositions, with experimental tools from psychology, and thereby deepen our understanding of a fundamental aspect of human interaction. This is essential for a comprehensive perspective on how speakers relate to the world through language, and also bears on broader issues in society. For example, research on the psychology of memory has shown that the presuppositional status of information has strong effects on memory accuracy, which is highly relevant in eye witness testimony and other related activities. With the support of the National Science Foundation, Dr. Schwarz will study how presuppositions relate to other aspects of meaning, such as asserted content and certain pragmatic inferences (specifically implicatures), using a variety of experimental methods. Presuppositions are backgrounded relative to the main point asserted by an utterance, and (typically) already count as established in the discourse. For example, 'again' in 'John left again' suggests that it was already known that John had left previously, the main news being his second departure. The project will investigate the relative time-course of these two aspects of meaning in online processing: Does the presupposition get evaluated first, suggesting that its role for the overall felicity of an utterance needs to be confirmed at the outset? Or does it play the role of an additional inference that is only considered later? Another crucial feature of presuppositions concerns their relation to operators like negation: 'John hasn't left again' still conveys that he left before (while denying a second departure). The project will investigate whether such embedded occurrences of presuppositional expressions introduce additional cognitive demands. Finally, the project will also study the availability and potential cognitive cost of an alternative interpretation of such sentences, as in 'John (certainly) hasn't left again--he never left in the first place!' The present project constitutes the first comprehensive attempt to use tools from experimental psycho-linguistics to study the cognitive processes involved in presupposition comprehension. Using an eye tracker, Dr. Schwarz will record people's eye movements relative to a visual scene while they're hearing linguistic stimuli that contain presuppositional expressions. This allows for real-time insights into the unfolding of individual aspects of the overall meaning with millisecond-accuracy. Response times and other behavioral measures further help to understand speakers' interpretation processes and preferences. By combining theoretical work with foundations in philosophy of language and formal semantics with experimental methods from psycho-linguistics, this interdisciplinary project will help to integrate the study of meaning in natural language more deeply into the cognitive sciences and advance our theoretical and empirical understanding of the human mind's remarkable capacities at work in linguistic communication.

在解释语言话语时，人类的大脑必须将许多不同类型的信息拼凑在一起。虽然这是在潜意识中发生的，但语言研究人员已经发展出了关于意义的不同方面如何被纳入整体解释的理论观点。然而，关于这些理论结构与实时发生的实际认知过程之间的关系，人们知之甚少。本项目将通过研究意义的各个方面，特别是预设，利用心理学的实验工具来填补这一空白，从而加深我们对人类互动基本方面的理解。这对于全面了解讲话者如何通过语言与世界联系是至关重要的，也与社会中更广泛的问题有关。例如，对记忆心理学的研究表明，信息的预设状态对记忆的准确性有很强的影响，这在目击证词和其他相关活动中是高度相关的。在美国国家科学基金会（National Science Foundation）的支持下，施瓦茨博士将使用多种实验方法，研究预设与意义的其他方面（如断言的内容和某些语用推断（特别是含意））之间的关系。预设是相对于话语所主张的主要观点的背景，并且（通常）已经在话语中建立起来了。例如，“John left again”中的“again”表明人们已经知道John已经离开了，主要的消息是他第二次离开。该项目将调查在线处理中这两个方面的意义的相对时间进程：是否先对预设进行评估，这表明它对话语的整体幸福的作用需要在一开始就得到确认？或者它扮演了一个附加推理的角色，只在后面考虑？预设的另一个关键特征与它们与否定等运算符的关系有关：“John hasn’t leave again”仍然传达出他以前离开过（同时否认第二次离开）。该项目将调查这种嵌入的预设表达是否会引入额外的认知需求。最后，该项目还将研究这些句子的另一种解释的可用性和潜在的认知成本，比如“约翰（当然）没有再次离开——他根本就没有离开过！”本研究首次全面尝试使用实验心理语言学的工具来研究预设理解中涉及的认知过程。施瓦茨博士将使用眼动仪记录人们在听到包含预设表达的语言刺激时，相对于视觉场景的眼球运动。这允许以毫秒级的精度实时洞察整体意义的各个方面的展开。反应时间和其他行为测量进一步帮助理解说话者的口译过程和偏好。通过将语言哲学和形式语义学的理论工作与心理语言学的实验方法相结合，该跨学科项目将有助于将自然语言的意义研究更深入地融入认知科学，并推进我们对人类思维在语言交流中的卓越能力的理论和经验理解。