Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural entrainment and its role in the interaction between rhythm and semantic context in language comprehension and production
Poster Session B, Wednesday, September 30, 4:30 - 6:30 pm, Wangari Maathai
Anaëlle Filisetti1, Vitória Piai2, Antoine Petit1, Laurent Ott1, Anahita Basirat1; 1SCALab, University of Lille, 2Donders Centre for Cognition, Radboud University
Although language comprehension and production are often studied separately, studies suggest they mutually influence each other (Pickering & Garrod, 2013). In particular, the sentence’s semantic context in comprehension appears to affect word production (e.g., Piai et al., 2014). To investigate this context effect, the Concise Language Paradigm (ClaP, Roos et al., 2024) was proposed as a framework to study comprehension and production in the same experimental trial. Beyond semantic context, speech prosody also carries important information, and speech rhythm appears to be essential for speech intelligibility, particularly by synchronising neural activity with the perceived rhythm (Aubanel & Schwartz, 2020; Huettig, 2015; Robson et al., 2024; Schroeder & Lakatos, 2009). This mechanism, known as neural entrainment, is thought to be one of the mechanisms involved in the temporal prediction of speech (Zoefel & Kösem, 2024). Therefore, since speech rhythm impacts perception and prediction, investigating whether it can influence language production seems relevant. Examining the role of neural entrainment in these interactions between rhythm and other linguistic processes would also be valuable. In our planned studies, we will use the CLaP paradigm combined with EEG. Participants will listen to sentences missing their final word and then name an image representing that word. The sentence presented will be semantically constrained or unconstrained and its rhythm will be natural or isochronous. This will allow us to investigate the impact of rhythm and semantic constraint on word production, and the role of neural entrainment in this interaction. Since no French databank exists with materials systematically manipulating sentence constraint and controlling for characteristics like syllable number, we created a corpus of 269 sentences with different levels of constraint. Sentences are organised into pairs, comprising a constrained sentence and an unconstrained sentence, linked to the same target word, matched for the number of syllables. All target words exist in the Multipic database (Duñabeitia et al., 2018), allowing the sentences to be used in paradigms such as CLaP. Sentence constraint was calculated using cloze probability from a sentence completion task involving 160 participants. Sentences’ mean cloze probability, reflecting the probability to produce the target word, was 0.77 (SD=0.21) for constrained sentences and 0.04 (SD=0.07) for unconstrained sentences. The mean modal completion, representing the probability of the most frequent response, was 0.78 (SD=0.19) for constrained sentences and 0.22 (SD=0.15) for unconstrained sentences. This databank could also be used in pathologies such as aphasia and Parkinson’s disease, both for experimental studies and for speech-language therapy. For our future study, we plan to recruit 30 native French speakers. To ensure high constraint for constrained sentences and low constraint for unconstrained ones, we selected 81 sentence pairs from our databank, where constrained sentences had a target cloze probability above 0.70 and unconstrained sentences had a modal response below 0.30. We will quantify the impact of rhythm and semantic context on language production by comparing naming latency and the degree of coherence between speech and the EEG signal across conditions, and the association between the behavioral results and the degree of coherence.
Topic Areas: Language Production, Speech Perception