Abstract
It has been known for a long time that event-related brain potentials (ERPs) can provide valuable information about the nature and time course of sentence comprehension. However, in part because using brain measures is difficult enough as it is, psycholinguists have until recently limited their use of ERPs to the comprehension of single sentences presented in isolation. In my lecture, I will review recent ERP research on the comprehension of sentences in the context of a wider discourse. By doing so, I hope to illustrate that it is both feasible and very useful to study discourse-level comprehension by means of brain imaging methodology. One of my claims will be that as one moves up to study more complex levels of language comprehension, it becomes harder to exclusively rely on the logic of one's design to disentangle the many different processes involved, and, as a consequence, increasingly valuable to have a dependent measure that by itself provides clues as to which of these processes we might be looking at. To avoid delivering a slick ERP advertisement, I'll also discuss several major drawbacks. One of these is that because ERPs require a large number of critical trials per cell of one's experimental design, the use of a relatively time-consuming type of trial (a short text) usually leaves one with little room for fillers. Furthermore, the testing situation in an EEG experiment is not exactly conductive to interesting goal-directed conversation or the spontaneous reading of a fine novel. Finally, although obviously desirable, the temporal precision afforded by on-line ERP measures can lure researchers into asking too many 'when' questions, possibly at the expense of 'what' and 'how'. So, presumably just like eye tracking and other (e.g. self-paced reading) measures, there are some very good reasons to use ERPs, as well as some very good reasons to not exclusively rely on them.
