Event conceptualisation and linguistic realisation: The impact of semantic and lexical factors on sentence production

论元结构、论元与助词的区别以及主位角色与事件的编码关系最为密切，但对这些概念的定义却鲜有共识。目前的建议从心理语言学的角度来解决这些问题，重点放在句子产生中概念化和表述的交集。结构启动实验将有助于将基于句子生成的严格增量模型的预测与词汇主义账户所做的预测分开。严格递增的假设需要假设概念因素是语言编码的主要决定因素。因此，词汇信息对表达的影响很小，而主位层次则指导主位角色到句法功能的映射。然而，突出的指称对象，例如，在语篇中，可能比其主题位置所暗示的更早实现。相比之下，不专注的参与者可能根本不会被编码，并且仍然是隐式的。结构持续性预计将以类似于背景操纵的方式影响主题关系：它增加了显示与主题层次不一致的主题角色顺序的概念表征的可用性。词汇主义者认为词汇信息触发了主位角色到句法功能的映射。如果结构启动以词汇表征为目标，在没有词汇和句法相似性的情况下，概念相似性应该是无效的。此外，动词重复的影响应该仅限于论元。动词重复的效果概括到带有附加语的结构，指向动词出现的句法语境的动态表示。这些实验将测试包括隐含论元、附加语或非典型主位角色在内的关系是如何编码的，因为这些关系使我们能够调查概念和词汇因素的影响。这一结果将促进人们对前言信息到句法结构映射机制的理解。它们将有助于具体说明概念化和功能编码的相关表征和过程，以便将它们整合到语言产生的模式中。最后，从实验中得出的见解将被用来巩固认知领域中的相关语言概念。

Argument structure, the argument vs. adjunct distinction, and thematic roles are most relevant for the encoding of events, but there is little agreement on the definition of these concepts. The current proposal addresses these issues from a psycholinguistic perspective, focusing on the intersection of conceptualisation and formulation in sentence production. Structural priming experiments will help to dissociate the predictions based on a strictly incremental model of sentence production from predictions made by a lexicalist account. The assumption of strict incrementality entails the hypothesis that conceptual factors are the main determinants of linguistic encoding. Thus, lexical information has only little impact on formulation whereas the thematic hierarchy guides mapping of thematic roles onto syntactic functions. However, a referent that is salient, e.g., in discourse may be realised earlier than suggested by its thematic position. By contrast, a participant that is not focused might not be encoded at all and remain implicit. Structural persistence is expected to affect thematic relations in a way that resembles a contextual manipulation: It adds to the availability of a conceptual representation displaying an order of thematic roles that is incongruent with the thematic hierarchy. Lexicalist accounts claim that lexical information triggers the mapping of thematic roles onto syntactic functions. If structural priming targets lexical representations, there should be no effect in the case of conceptual similarity without lexical and syntactic similarity. Moreover, effects of verb repetition should be restricted to arguments. An effect of verb repetition that generalizes to structures with adjuncts points to a dynamic representation of the syntactic contexts in which the verb occurs. The proposed experiments will test how relations are encoded that include implicit arguments, adjuncts, or non-prototypical thematic roles, because these relations allow us to investigate the impact of conceptual and lexical factors. The results will promote the understanding of the mechanisms that map preverbal messages onto syntactic structures. They will help to specify relevant representations and processes of conceptualisation and functional encoding in order to integrate them into a model of language production. Finally, the insights drawn from the experiments will be used to consolidate the relevant linguistic notions in the cognitive domain.