Wiese, H. 1 , Wittenberg, E. 2 & Öncü, M. T. 3
1 University of Potsdam
2 University of California at San Diego
3 Ege Üniversitesi Izmir
Many studies have shown language-independent preferences for the serialisation of thematic roles (Agent, Patient, Act). We ask whether such preferences extend to information structure, specifically, the relative order of framesetters and topics with respect to their predication. While topics specify the entity under which a proposition is to be stored, framesetters provide specifications - typically time or place - restricting the domain in which the proposition is valid. Accordingly, we predict that both will be preferred in initial position, before the predication itself, regardless of language-specific word orders.
To test this, we presented English, German, and Turkish monolinguals and German-Turkish bilinguals with a (non-verbal) comic sequence and asked them to describe the final picture, which included a temporal framesetter and a topic. Participants had to render the scene (a) verbally and (b) in an 'extra-grammatical' set-up, using toy figures, clocks, and cards with verbs.
Results indicate that verbal descriptions followed different language-specific word orders. However, in the extra-grammatical task, in addition to language-specific effects, we found a strong overarching pattern where speakers placed framesetters and topics before the verb, suggesting information-structural preferences that can violate language-specific restrictions: findings point to information-structural patterns that are independent of, and interacting with language-specific grammatical constraints.