Item

Abstract

Auditory sentence processing involves semantic, syntactic and prosodic information. Prosodically guided sentence processing has been shown to involve the left inferior frontal gyrus (IFG). Prosodic cues are known to interact closely with both syntax and semantics, and these processing domains in turn have been attributed to two different subregions within the left IFG: based on neuroimaging studies, the anterior part is associated with semantic processing and syntactic processing is ascribed to the posterior part. Yet, the causal role of this regional specialisation remains unclear. The current study used focal perturbations induced by repetitive transcranial magnetic stimulation (rTMS) to probe the causal role of the posterior IFG for syntactic processing and the anterior part for semantic processing in prosodically varied sentences. Since the interpretation of a pitch accent in the sentence was essential for successful task performance, we additionally were able to assess the involvement of left IFG in sentence processing when crucial information was conveyed by prosody. Healthy participants performed a sentence completion task with a syntactic and a semantic decision. The pitch accent in the truncated spoken utterance cued which determiner and noun (presented visually) would form the most suitable sentence ending, which were selected by button-press. Healthy participants underwent three sessions with 10 Hz rTMS bursts being applied over either anterior or posterior left IFG, or vertex (control region). rTMS was applied 100 ms after onset of the presentation of the determiners (syntactic condition) or nouns (semantic condition). Although we found no significant interaction between rTMS site and decision type, a main effect of rTMS site indicated decreased task accuracy in both decision types after posterior IFG stimulation versus vertex. Considering the lack of interaction effect, we suggest the use of participant-specific functional localisers in future TMS studies to further investigate the specific contribution of different IFG subregions in sentence comprehension. The decreased accuracy after posterior IFG stimulation in both the syntactic and semantic decisions may be explained by impaired processing of the grammatical roles in the sentence, which was required for successful task performance in both conditions. Therefore, we interpret these results as evidence for the functional relevance of the left posterior IFG in grammatical processing guided by prosodic cues.