Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Why “semantics” labels nothing in particular: eight elements of semantics across 117 neurostimulation studies
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
T R Williamson1,2, Gesa Hartwigsen3,4, Kris Kinsey1,5, Neil U Barua1,2, Naomi Heffer5, Philipp Kuhnke3,4, Sonia Mariotti1,2, Eimear McKnight1,6, Lydia Wiernik1,7,8, Jemma Sedgemond5, Antonia Vogt9, Anna E Piasecki1,2,5; 1Brain, Language, and Behaviour Laboratory, UWE Bristol, UK, 2Southmead Hospital, North Bristol NHS Trust, UK, 3Wilhelm Wundt Institute for Psychology, Leipzig University, Germany, 4Cognition and Plasticity, Max Planck Institute for Human Cognitive and Brain Sciences, Germany, 5School of Social Sciences, UWE Bristol, UK, 6Dept. of Linguistics, Queen Mary University of London, UK, 7SignLab, Dept. of German Philology, University of Göttingen, Germany, 8Neuropsychology, Max Planck Institute for Human Cognitive and Brain Sciences, Germany, 9Dept. of Medicine, University of Cambridge, UK
Introduction. Recent multidisciplinary work has confronted the proliferation of inconsistent terminology in semantic research, building consensus definitions for 17 target constructs to support cross-disciplinary calibration (Reilly et al., 2025). That work demonstrates how widely the term semantic travels across linguistics, philosophy, psychology, and neuroscience, and how rarely its referents converge. We extend this concern from definition to operation: even when researchers agree on a definition, the question remains what operations the brain actually performs when it does what gets labelled semantic. Causal evidence from neurostimulation studies offers a route to specification at this level. Methods. A preregistered systematic review (PROSPERO CRD42024602006) of TMS, tES, and DES studies of language returned 12,763 records from PubMed, Scopus, Embase, and PsychInfo; 220 papers (1999–2025) met inclusion criteria, reporting 608 significant outcomes across 38 languages and approximately 6,300 participants. Each outcome was qualitatively analysed (Braun & Clarke, 2006) to infer the most specific causally modulated subprocess — termed an element — and organised within a seven-level hierarchy of process specificity. Bottom-up clustering of outcomes tagged for the construct semantics was applied with AI-assisted construct grouping, calibrated to the construct's level in the hierarchy. Results. The construct semantics is recoverable across 117 papers and 197 distinct processes (N = 3,239). Bottom-up clustering returned eight elements the published evidence supports as causally distinct. Semantic-to-lexical mapping (42 studies, N = 1,169), implicated across IFG, MTG, STG, angular gyrus, and the inferior fronto-occipital fasciculus. Semantic feature knowledge retrieval (35 studies, N = 774), spanning anterior temporal lobe, IFG, premotor cortex, and modality-specific association areas including the left posterior inferior parietal lobe (Kuhnke et al., 2020). Semantic relatedness evaluation (35 studies, N = 781), at anterior temporal lobe, IFG, and angular gyrus. Controlled semantic selection (23 studies, N = 674), drawing on prefrontal and posterior MTG sites. Semantic integration in sentence context (13 studies, N = 376), at IFG and posterior temporal cortex. Syntactic-semantic feature assignment (7 studies, N = 83). Compositional novel meaning construction (7 studies, N = 212). Episodic encoding via semantic relatedness (5 studies, N = 338). Several elements recur across regions; some regions support multiple. The pattern indicates that semantics in the published neurostimulation literature operates as a coarse cover for a heterogeneous bundle of operations whose neural distributions cross-cut traditional category boundaries. Conclusion. Cautiously, we suggest that progress in mapping the neurobiology of meaning will benefit from specification at the level of the elemental operation rather than the inherited construct, as a complement to the consensus-definition work of Reilly et al. (2025). Future work is needed to test the ontological coherence of each by-region heterogeneous bundle of elements, with the aim of recharacterising language regions by the operations they actually support (Genon et al., 2018) — an enterprise to which this database (language-elements.org) is intended to contribute. The eight elements presented here are not proposed as a final ontology of semantic operations but as a tractable starting point for that work.
Topic Areas: Meaning: Lexical Semantics, Computational Approaches