Composition in context: Contextual effects on the online processing of eventive coercion

我们建议研究在复杂表达的构成过程中语境信息与意义转换的相互作用（=强迫）。上下文对合成过程的影响是长期以来关于语义-语用界面架构的争论的核心。作为起点，拟议的项目将对该接口的性质采取两种不同的观点：a)一个模块化的帐户，根据该帐户，句子意义最初是在与语用信息隔离的情况下处理的；b)一个交互式帐户，语义和语用信息从一开始就渗透到复杂意义的组成中。现有的实验研究以强迫现象为主要工具来研究构成过程的在线加工，并取得了很大的成功，但迄今为止，语境信息的作用在很大程度上被忽视了。基于我们对强迫现象的现有研究，我们计划在四条研究线中研究句子与更大的话语上下文之间的相互作用。首先，我们将研究复杂事件表示组成过程中的方面富集，以研究来自前面话语上下文的信息是否可以立即与句子中的信息结合起来。其次，我们将研究方面贫乏的情况，在这种情况下，句子中的信息需要重新解释为前面的话语上下文建立的事件表示。第三，我们将研究方面歧义的语境解析，即强制语句，它允许两种不同的强制操作与两种不同的对应事件表示，其中只有一种是语境支持的。最后，我们将扩展现有的关于语境对补语强制的影响的研究，旨在更好地理解衍生构成意义的潜在操作。该项目的研究结果将在实证和理论上显著改善该领域。该项目关注在线写作过程中的本地语境调整，将为将语用理论扩展到实时处理做出重要贡献。在实证方面，我们将收集离线、眼动追踪和ERP数据，以更好地理解在上下文中推导复杂意义所涉及的认知基础。该项目主要关注xprg .de的主要目标：基于实验证据的新语用理论，解决如何将语义/语用界面架构理论转化为更精确的语义组合上下文依赖理论的问题。至关重要的是，我们将考虑语言处理的时间和神经基础，利用广泛的实验方法。

We propose to investigate the interaction of contextual information with shifts in meaning during the composition of complex expressions (= coercion). The nature of contextual influences on compositional processes is central for long lasting debates about the architecture of the semantics-pragmatics interface. As a starting point the proposed project will take two alternative views on the nature of this interface: a) a modular account according to which sentence meaning is initially processed in isolation from pragmatic information or b) an interactive account with semantic and pragmatic information percolating into the composition of complex meaning right from the start. Existing experimental research has employed the phenomenon of coercion as the primary tool to study the online processing of compositional processes with great success, but so far the role of contextual information has been largely ignored. Building on our existing work on coercion phenomena, we plan to study interactions between the sentence and the larger discourse context in four lines of study. Firstly, we will look at aspectual enrichment during the composition of complex event representations to investigate whether information from the preceding discourse context can immediately be combined with information from the sentence. Secondly, we will study cases of aspectual impoverishment in which information in the sentence requires a reinterpretation of the event representation that has been built up for the preceding discourse context. Thirdly, we will investigate contextual resolution of aspectual ambiguity, i.e. coercion sentences which allow for two different coercion operations with two different corresponding event representations only one of which is contextually supported. Finally, we will extend existing research on contextual effects on complement coercion aiming at a better understanding of the underlying operations in deriving compositional meaning.The findings of the project will significantly improve the field both empirically and theoretically. Focusing on local contextual adjustments during online composition the proposed project will make an important contribution to extend pragmatic theory towards real-time processing. On the empirical side, we will gather offline, eyetracking and ERP data to gain a better understanding of the cognitive foundations involved in deriving complex meaning in context.The proposed project focuses on the main goal of the XPrag.de: New Pragmatic Theories based on Experimental Evidence by addressing the question how theories on the architecture of the semantics/pragmatics interface can be transformed into more precise theories of context dependence of semantic composition. Crucially, we will consider the temporal and neural underpinnings of language processing, utilizing a broad range of experimental methods.