| |
Abstract:
Abstract: Abstract: Since several decades the question has
been asked which role a verb plays within sentences comprehension.
We conducted three ERP studies in order to explore how a human
sentence processor uses verbs to predict possible arguments.
Several studies have focussed on the issue when which information
encoded in the verb is made available during reading. Some have
proposed that the interpretation of elements joining the verb is
highly dependent on the specification in the verb's lexical entry.
Others believe that there are only general rules organizing new
incoming input according the most simple structure such that it can
be removed from short term memory as soon as possible. Our data
show that the comprehension system is very fast, as it makes
predictions about following material very early. At the position of
the verb we found a frontally distributed negativity for transitive
verbs when compared to intransitives. This ERP difference has its
peak around 500 milliseconds after the onset of the verb. We
interpret these data as follows: when reading a transitive verb the
human sentence processor opens argument slots to be filled by the
coming material after that verb. This preallocation causes memory
load. Therefore the ERP wave is more negative for transitive than
for intransitive verbs.
|
