Erp studies on the syntax-semantic interplay during language comprehension

Resumo

Human language comprehension combines the meaning of words and the syntactic structure of an utterance to build up a representation of a whole meaning. The aim of this thesis is to investigate the phenomenon of the syntax-semantic interplay during language comprehension. The investigation of how these cognitive processes occur online has been a controversial challenge for linguistics, psycholinguistics and neuroscientists for decades (see e.g., Friederici, 2002; MacDonald, Pearlmutter, & Seidenberg, 1994; Ferreira, & Clifton, 1986; Rayner, Carlson, & Frazier, 1983). The empirical studies of this thesis have been realized using the technique of Event-related Potentials (ERPs) which permits analysis of the functional organization of language from its brain bases, in a non-invasive way. Chapter 1 of this thesis begins with a description of what syntax and semantics are. The classical study of language posed no questions about the independence of these two processes. However, recently some evidence seems to reveal that syntax and semantics in some circumstances may interact. Arguments and counterarguments have been discussed to explain the goal of this dissertation. The second section of Chapter 1 provides an introduction to the distinction between syntax-first and constraint-based models about syntax and semantic processes during language comprehension. While syntax-first models report that syntactic analysis is carried out autonomously in the initial stages of language comprehension and is not influenced by semantic and contextual information, constraint-based models posit that the semantic information contributes to the syntactic analysis of the utterance. In addition, the discussion has been extended over more recent neurocognitive models that suggest a key role of competing neural processing streams. Chapter 2 introduces the ERP technique, focusing on the different ERP components found in neurocognitive studies related to syntactic and semantic processing. In this section, empirical evidence is offered regarding the syntax-semantic interplay and how these effects can be the point of departure to test the different neurocognitive models. Chapter 3 provides a description of the experiments that have been realized to investigate this phenomenon. Finally, Chapter 4 is an extensive general discussion of the experimental findings in light of recent dynamic models of language comprehension.