Benutzerhandbuch Datenschutzhinweis Impressum Kontakt





Dynamic formation of syntactic predictions based on speaker identity


Kroczek,  Leon O. H.
Department Neuropsychology, MPI for Human Cognitive and Brain Sciences, Max Planck Society;


Gunter,  Thomas C.
Department Neuropsychology, MPI for Human Cognitive and Brain Sciences, Max Planck Society;

Externe Ressourcen
Es sind keine Externen Ressourcen verfügbar
Volltexte (frei zugänglich)
Ergänzendes Material (frei zugänglich)
Es sind keine frei zugänglichen Ergänzenden Materialien verfügbar

Kroczek, L. O. H., & Gunter, T. C. (2017). Dynamic formation of syntactic predictions based on speaker identity. Poster presented at 59th Conference of Experimental Psychologists, Dresden, Germany.

Predictions allow for an efficient processing in communicative situations. In order to be efficient, predictions need to be adapted to characteristics of the environment. In a communicative situation, speaker-specific language use might shape a listener’s predictions about upcoming language stimuli. In the present experiment we asked whether listeners use speaker characteristics to generate predictions about syntactic structure and how these predictions might change over time. Twenty participants were presented with sentences which were spoken by two different speakers. Sentences had either a Subject-Object-Verb (SOV) structure or an Object- Subject-Verb (OSV) structure. Crucially, the two speakers differed with regards to the frequency by which they produced a particular syntactic structure. One of the speakers had a high probability to produce a SOV structure and a low probability to produce an OSV structure, and vice versa for the other speaker. Additionally, speakers produced sentences which were ambiguous towards their syntactic structure. For the ambiguous sentences, participants had to identify the subject or the object of the sentence. This allowed us to infer participants’ predictions regarding the syntactic sentence structure. Furthermore, in order to assess the significance of speaker-specific predictions, participants were invited to a follow-up study eight months after the initial exposure to the speakers. The data show that participants started with a strong bias towards the SOV structure, which is the canonical sentence structure in German. With increasing exposure to the speakers, however, participants developed predictions regarding the particular syntactic structure based on speaker identity. These predictions were still coupled to the speakers eight months after the initial exposure. This demonstrates that listeners are sensitive to speaker-specific syntactic preferences and use this information to generate predictions.