Tid: Torsdagen 9 februari kl. 15.00 - 17.00
Plats: C307

Postseminarium äger rum direkt efter seminariet i institutionens pentry.


In my thesis, I investigate the comprehension of grammatical relations in transitive sentences, i.e. the process by which listeners/readers interpret the relations between the arguments and the predicate of the clause. Cross-linguistic research suggests that grammatical relations should be defined on the basis of the functions of the arguments rather than on their structural positions, as assumed in the generative grammar tradition. Grammatical relations are, according to this view, defined on the basis of the thematic protoroles or generalized semantic roles of the arguments, i.e. which of the two arguments that is the most “agent-like” (the Actor argument) versus the most “patient –like” (the Undergoer argument).

Argument interpretation is, in line with this view, assumed to involve a linking between structural positions and generalized semantic roles, a process which is driven by both morphosyntactic (e.g., word order, case marking and agreement) and semantic/referential (e.g., animacy, definiteness) cues with language-specific weightings. I will present some evidence for this view on the basis of an ERP study investigating the reanalysis of Swedish sentences which are locally ambiguous with respect to the grammatical function of the initial argument (i.e., grammatical function reanalysis).

Finally, under the assumption that there exist correspondences between, on the one hand, usage frequencies within and across languages, and, on the other, preferences in comprehension and production (i.e., the performance-grammar correspondence hypothesis), language-specific weightings of argument interpretation cues should be able to be quantified on the basis of corpus data and I will also briefly address this.


Ljuba Veselinova