What happens in words: Towards a model of incremental event representation

形式语义学的中心目标是模拟句子的意义是如何从单词的意义中建立起来的，但不关心单词在实际的演讲或写作中出现的线性顺序。相反，心理语言学研究关注的正是这个问题--即认知过程如何逐步（一个词接一个词）建立意义。鉴于这两个分支学科在方法和目标上的差异，心理语言学研究往往没有充分利用语义模型的洞察力，特别是在我们使用的单词和句子所描述的事件方面。当人们听到或阅读一个句子时，他们构建事件的方式相当复杂，涉及到几个决定，包括谁参与了事件，事件发生的地点，事件的数量，以及这些事件之间的联系考虑一个句子，比如（1）。1）Esther本月早些时候兴高采烈地用信用卡买了Neil盖曼的新书，Khadija也买了--但Jamie肯定没有。这里描述的是两个购买事件和两本书吗？是否有两个购买事件和一本书，例如Esther买了这本书，Khadija后来买了同一本书，甚至可能是从Esther那里买的？最自然的，默认的解释是一个单一的购买事件，他们一起买了一本书？Khadija每次都用谁的信用卡她自己的还是Esther的？那杰米呢杰米的不购买行为在购物者的头脑中是否以同样的方式被表征为一个事件，或者被赋予的动词是否导致了不同类型的心理表征？我们能够理解我们用语言创造的世界的方式似乎与我们如何从一个句子或更广泛的话语的潜在潜在事件中构建一个文本或话语所描述的情景的心理抽象特别相关。事件本身影响了对语义学中各种主题的研究，如时态和体（时间如何在语言中编码），状语修饰（how the syntax句法position位置of words词like 'quickly迅速' intelligently微妙affects影响the meaning含义of a sentence句子），和论证实现（动词的解释属性和意义如何影响它们出现的句法语境）-参见Bohnemeyer et al.（2007）; Lascarides和Asher（1993年）; Pianesi和瓦尔齐（2000年）; Rappaport Hovav（2008年）; Smith（2006年）; Tenny（2000年）。关于这些事件的心理学方面的文章要少得多。虽然Bott和Sternefeld（2017）展示了在理论上如何将对正式语义中事件的见解应用于增量解释，但相对而言，很少有人研究如何实时地逐字概念化和处理事件信息。因此，拟议项目的目标是进行三个distinctbut相互关联的实验，对未充分研究的事件现象的实时处理，分别利用EEG，眼动跟踪和反应时间任务。利用我在形式语义学和认知处理方面的背景，并借鉴在这两个领域都有经验的研究人员的远见，本研究将从这些通常不同的方法中汇集见解，以跨学科的努力来探索和解释作为言语和写作展开的事件概念化的潜在机制。首先，通过发展一个强大的帐户，如何事件是递增地表示在头脑中，将奠定基础，为未来的研究语言的一个基本方面。第二，知觉心理学观察到，人们将连续的视觉活动分解为有意义的事件状块，这是一种自动认知过程，为未来的记忆和学习提供了支架（Zacks和Swallow，2007）。

The central aim of formal semantics is to model how the meanings of sentences are builtfrom the meanings of words, but without concern for the linear order that words appear induring actual speech or writing. Conversely, psycholinguistic research concerns itself withexactly this question - namely, how cognitive processes incrementally (word-by-word) buildmeaning. Given the differences in the methods and goals of these two sub-disciplines,psycholinguistic research has often underutilised the insights from semantic models,particularly with respect to the events described by the words and sentences we use.The way comprehenders construct events when they hear or read a sentence is quitecomplex, and involves several decisions about who is involved in the event, where the eventtakes place, the number of events, and how those events relate to one another. Consider asentence like (1).1) Esther gleefully bought Neil Gaiman's new book with her credit card earlier this month,and Khadija did too - but Jamie sure didn't.Are there two purchasing events and two books described here, such that Esther and Khadijaseparately bought a copy of the new book? Are there two purchasing events and a singlebook, such that Esther bought the book and Khadija bought that very same book at a latertime, perhaps even from Esther? Is the most natural, default interpretation one of a singlepurchasing event where they bought a single copy of the book together? Whose credit carddid Khadija use on each occasion, her own or Esther's? And what of Jamie? Is Jamie's nonpurchasingrepresented as an event in the same way in the minds of comprehenders, or donegated verbs result in different kinds of mental representation?The way we are able to make sense of the worlds we create with language appears to beespecially tied up with how we construct a mental abstraction of the situation described by atext or utterance from the potential underlying events of that sentence or wider discourse.Events themselves have influenced research into topics in semantics as diverse as tense andaspect (how time is encoded in language), adverbial modification (how the syntactic positionof words like 'quickly' subtly affects the meaning of a sentence), and argument realisation(how the interpretive properties and meaning of verbs influence the syntactic contexts theyappear in) - see Bohnemeyer et al. (2007); Lascarides and Asher (1993); Pianesi and Varzi(2000); Rappaport Hovav (2008); Smith (2006); Tenny (2000). Much less has been writtenabout the psychology of these events.While Bott and Sternefeld (2017) showed how insights about events in formal semanticscould, in theory, be applied to incremental interpretations, comparatively little work hasbeen conducted on how comprehenders conceptualise and process event information wordby word, in real-time. The goal of the proposed project is, therefore, to conduct three distinctbut interrelated experiments on the real-time processing of understudied event phenomena,utilising EEG, eye-tracking, and reaction time tasks, respectively. Taking advantage of mybackground in both formal semantics and cognitive processing, and drawing upon thesupervision of researchers with experience in both areas, the present study will bringtogether insights from these often disparate approaches in an interdisciplinary effort toexplore and explain the mechanisms underlying comprehenders' conceptualisations ofevents as speech and writing unfolds.The value of this project is twofold. First, by developing a robust account of how events are incrementally represented in the mind, the groundwork will be laid for future research into a fundamental aspect of language. Second, perceptual psychology has observed that peopleparse continuous streams of visual activity into meaningful event-like chunks, an automatic cognitive process which scaffolds future memory and learning (Zacks and Swallow, 2007).