Item

Abstract

The typical event-related brain potentials (ERP) experiment on language comprehension is one in which readers get a bunch of unrelated and completely decontextualized single sentences. For some questions, this is just fine. However, if we want to use neuroimaging to further our understanding of pragmatics, we may need to do a little more. The ERP work that I will present largely came from a desire to study language comprehension in somewhat richer contexts. Our studies show that the linguistic brain rapidly draws upon a wide variety of information sources, including prior text and inferences about the speaker. Furthermore, people anticipate what might be said about whom, they use heuristics to arrive at the earliest possible interpretation, and if it makes sense, they sometimes even ignore the grammatical rules of their language. Language comprehension is opportunistic, proactive, and, above all, immediately context-dependent. What the latter means in relation to theories of pragmatics and semantics is sometimes obvious, but in many cases remains to be established. I strongly believe that to move forward, neurocognition of language researchers should be prepared to face the complexity of real language use and the theories about it.