Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Neural bases of conceptual relation processing in sentences: an fMRI study
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
Yao-Ying Lai1, Michiru Makuuchi2; 1National Chengchi University, Taipei, Taiwan, 2National Rehabilitation Center for Persons with Disabilities, Tokorozawa, Japan
[Introduction] We investigate the neural bases of combinatorial conceptual computation during language comprehension by manipulating conceptual relations between two noun phrases (NPs) in copular sentences, following Lai & Lai (2026). Hyponymy like (a) “Monet is a painter.” denotes class-inclusion through transparent semantic composition, serving as the baseline. Tautology like (b) “Monet is Monet.” licenses a stereotypical class-inclusion invoking the referent’s characteristic properties. Metonymy and Metaphor involve figurative readings through within-domain (Monet → Monet’s artwork in (c) “Monet is art.”) and cross-domain (Monet → a camera in (d) “Monet is a camera.”) conceptual mappings, respectively. We varied conceptual relations while holding syntactic structure identical to identify the neural correlates of distinct semantic-pragmatic processing beyond morphosyntax. We hypothesize that Nontransparent sentences whose interpretations extend beyond simple semantic composition—Metonymy, Metaphor, Tautology—would recruit brain regions supporting contextual enrichment, compared with transparent Hyponymy. The nontransparent cases would also differ according to their respective conceptual computations. [Method] Our fMRI experiment recruited 42 native speakers of Taiwan Mandarin (19 females, age range = 18-34, mean FLANDERS score = 9.64). Participants read each sentence presented in three segments, [subject.NP1 + copula + predicational.NP2] (600 ms/segment, with 100ms blank in between), and judged whether it was meaningful. We first constructed 50 quadruples, administered to a naturalness-rating questionnaire (n=40) and a familiarity and comprehensibility-rating questionnaire (n=31) to ensure the figurative senses for Metonymy and Metaphor. The 40 best-rated quadruples were selected, with the predicational.NP2 matched for frequency and strokes in Chinese characters, for the fMRI experiment (Siemens Skyra 3T, TR = 1,000 ms, TE = 30 ms), plus 120 anomalous fillers (Anomaly). [Results] Three identified outliers were excluded, leaving 39 participants for analysis. Meaningfulness judgments (1 = meaningful / 0 = meaningless) showed: Hyponymy (0.97) > Tautology (0.80) > Metonymy (0.68) > Metaphor (0.26) > Anomaly (0.02). Judgment reaction-time revealed Metonymy (936 ms) > Metaphor (897 ms) > Anomaly (599 ms) > Hyponymy (589 ms) > Tautology (465 ms). Neuroimaging results showed that the three Nontransparent conditions combined elicited greater activation in the left inferior frontal gyrus (IFG) extending to the frontal operculum (FOP). Relative to Hyponymy, Metonymy additionally recruited the left anterior inferior frontal sulcus, left supplementary motor area (SMA), right IFG, and right cerebellum. Metaphor (> Hyponymy) elicited stronger activation in the left IFG, SMA, and left posterior superior temporal sulcus (pSTS). Tautology (> Hyponymy) enhanced activation in the left supramarginal gyrus (SMG), right pSTS, and precuneus. The three nontransparent conditions also differed: Tautology elicited greater activation than Metonymy and Metaphor in the right superior frontal sulcus and bilateral SMG. Metonymy showed stronger left frontoparietal activation than Metaphor, whereas the reverse contrast yielded no significant effect. [Summary] In sum, Metonymic and Metaphorical conceptual mappings engaged the IFG extending to FOP (right- and left-lateralized, respectively) and L.SMA, while Tautology engaged the right-posterior-temporal and parietal regions for contextual integration. To our knowledge, this is the first fMRI study to juxtapose four conceptual relations with identical structure, thereby disentangling the neural networks that specifically subserve semantic-pragmatic computation during combinatorial language processing.
Topic Areas: Meaning: Discourse and Pragmatics, Syntax and Combinatorial Semantics