Electrophysiological Evidence for Sentence Comprehension - Wings

earlier syntactic component, early left anterior negativity or ELAN (for the criticism of
her design see Hagoort, 2003).
In general, these results reveal the opposition between the two groups of responses: ones
that are elicited by semantic and the ones that are elicited by syntactic violations. In a
way, these results correspond to the analysis that views sentence comprehension as a two-
foiled process in which, on one side, hierarchical structure is build and, on the other side,
interpretation is given to the ‘slots’ that this structure prepares. However, at least some
aspects of this view are wrong from the neurobiological point of view. From what is
generally known about how neural circuits work, information processing in human brain
is parallel and distributed. It can also be assumed that the same neurobiological principles
work across different cognitive domains. Therefore, since we know that, for example,
visual information is processed in dorsal and ventral streams that process what and
where, respectively, similar design could be assumed for language processing, as well
(i.e. one can assume similar syntactic and semantic ‘streams’). By the same analogy,
these ‘streams’ should be distinct, but highly interactive and a neurobiologically realistic
model should take this into account. Therefore, what is computationally elegant (such as
the syntax-first view on language comprehension), is not necessarily realistic in
neurobiological sense. Indeed, recent research into various aspects of sentence
comprehension reveal interaction between syntactic- and semantics related processes
(e.g. manipulations of N400 and P600 amplitudes related to Agent animacy in active and
passive sentences and verb transitivity as in Kuperberg et al., 2006; interaction between
grammaticality and frequency in tense violations in irregular verbs that elicit earlier P600
20