Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Combining early and late pragmatic processing via EEG: a study on metaphor and speech acts
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
Fabrizio Luciani1, Paolo Canal1, Luca Bischetti1, Chiara Pompei1, Rosario Tomasello2,3, Valentina Bambini1; 1Laboratory of Neurolinguistics and Experimental Pragmatics (NEPLab), University School for Advanced Studies IUSS, Pavia, Italy, 2Department of Philosophy and Humanities Brain Language Laboratory, WE4 Freie Universität Berlin, 14195 Berlin, Germany, 3Cluster of Excellence’ Matters of Activity. Image Space Material’, Humboldt
Introduction How quickly the brain processes pragmatic information is a central question in neuropragmatics[1,2]. ERP studies generally reported mid-to-late and late effects for Metaphors (N400-P600)[3], while Neuropragmatics of speech act model assumes that Communicative-Function is processed ultra-rapidly[4]. The present study investigated the temporal trajectories of pragmatic processing by combining Communicative-Function and Figurativity factors. We tested whether the effects emerge independently or interact, and whether individual differences in Theory of Mind (ToM) modulate their integration. Methods 40 native Italian speakers (25F; Mage=21.25±2.00) participated in an EEG experiment and completed the Read the Mind in the Voice (RMV)[5] test as a ToM measure. Participants listened to 160 spoken sentences in a fully-crossed design: Literal.vs.Metaphorical sentences (that toy/boxer is a panda) uttered with Falling (Assertion) or Rising (Question) intonation, yielding four conditions (LitAss/LitQue/MetAss/MetQue). A separate group (N=30) validated the stimuli. EEG (128 channels, re-referenced to linked-mastoids; 0.01–20 Hz bandpass-filtered, ICA corrected) was time-locked to target-word onset. RMS-based time-windows (P1, P200, N400, Late Component) were analyzed using linear mixed-effects models across Frontal and Posterior ROIs. Source estimation was performed with sLORETA in Brainstorm testing simple effects on the cortex surface using FDR correction. Results From early as P200 (193-292), an effect of Communicative-Function emerged in Frontal and Posterior ROIs (t=3.28, p=.001; t=2.47, p=.014), with Questions eliciting greater positivity than Assertions. Frontal ROI still tracked prosodic information into the N400 window (347-599 ms; t=3.02, p=.005), whereas posterior sites showed N400 for Metaphors compared to Literals (t=-2.38, p=.022). Both effects extended into the LC (600-900 ms), with Questions sustaining greater frontal and posterior positivity (t=2.55, p=.011; t=2.55, p=.004) and Metaphors showing posterior negativity (t=-2.06, p=.047). At this late stage a three-way interaction (Figurativity×Communicative-Function×RMV; t=3.07, p=.004) revealed that ToM ability modulated the joint integration of prosodic and figurative information: high-ToM participants showed greater negativity for MetAss.vs.LitAss (β=-2.80, p=.050) and greater positivity for MetQue.vs.LitQue (β=3.15, p=.028), whereas low-ToM participants displayed the opposite pattern for MetQue (β=-3.03, p=.034). Preliminary source estimation revealed an early effect (200-300 ms) in the auditory cortex for Questions. Metaphor-related contrasts showed late (550-800 ms) involvement of a broader network. Discussion These findings support an initial stratification and later interaction of pragmatic factors modulated by individual mentalizing skills. From early stages (P200) onward, the brain tracks communicative intentions through prosodic cues, involving the auditory cortex, independently of figurative content. Metaphor-specific inferential integration emerges later (N400-LC), consistent with classical accounts, and involving temporo-parietal regions. The late three-way interaction suggests that ToM ability supports a joint integration of prosodic and figurative information, with High-ToM individuals showing greater sensitivity to the communicative intention conveyed by interrogative metaphors. Overall, these results point to a multistream pragmatic processing, rapid for Communicative-Function recognition and delayed for Metaphor processing, that converge at later stages, modulated by social cognition. References [1] Canal & Bambini, 2023; [2] Cisneros et al., 2026; [3] Bambini et al., 2016; [4] Tomasello et al., 2022; [5] Sola et al., 2025
Topic Areas: Meaning: Discourse and Pragmatics, Prosody