Tracking Neuronal Entrainment to Distributional and Compositional Regularities in Sentences

Tracking Neuronal Entrainment to Distributional and Compositional Regularities inSentencesProject memeber and Speaker: Giulia CalignanoProject Leader: Francesco VespignaniResearch group: Alessia RaioDepartment of Developmental and Social Psychology (DPSS), University of Padova (Italy)Abstract Ding et al. (2016, 2017) suggested that auditory sequences of monosyllabic words might generate distinct power spectrum density (PSD) peaks corresponding to hierarchical syntactic structures. Frank and Yang (2019) questioned this interpretation, proposing that semantic similarity patterns could account for these effects. Later studies observed larger PSD peaks for sentences than for word lists with matched semantic rhythmicity (Lu et al., 2023) and for regular versus reversed sentences (Lo et al., 2022), indicating that these peaks are unlikely to be solely driven by semantic distance regularity. Kazanina and Tavano (2023) further examined whether these peaks arise from oscillatory mechanisms related to chunking or integration, highlighting broader implications for models of language processing. In this EEG study, Italian syllable sequences were used to explore the potential influence of lexical and structural regularities on PSD peaks. Italian’s linguistic characteristics—limited monosyllabic words, rich inflection, and flexible word order—enabled the creation of two experimental conditions: (1) mixed sequences with low lexical and distributional regularity, and (2) copular sequences with high regularity. These were contrasted with randomized syllable sequences. Visual presentation aimed to disentangle entrainment linked to perceptual/lexical processing (visual areas) from that linked to compositional processes (frontotemporal language network).Preliminary findings (N = 20/30) suggest that PSD peaks are larger for copular sequences compared to mixed sequences, with topographical variations indicating distinct processing mechanisms. These results appear to align with Frank and Yang’s (2019) hypothesis that distributional regularities amplify PSD effects while also pointing to compositional processes in mixed conditions. The findings contribute to ongoing discussions about oscillatory mechanisms underlying chunking and integration (Meyer et al., 2022; Kazanina & Tavano, 2023), underscoring the need for further exploration of structural unit durations and sentence level processing.