Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session B: Wednesday, August 13, 1:00 – 4:00 pm, de Brug & E‑Hall

Neural Encoding of Continuous Word Meaning Likelihood During Semantic Disambiguation

Andrey Zyryanov¹, Victoria Pierz¹, Yulia Oganian¹; ¹Eberhard-Karls-Universität Tübingen

Presenter: Andrey Zyryanov

Language comprehension hinges upon our ability to resolve semantic ambiguities, yet the neural representations underlying disambiguation remain unclear. To fill this gap, we recorded MEG while participants listened to German sentences containing an ambiguous target word, Blatt (meaning paper or leaf) or Tor (meaning gate or goal). In a behavioral pre-study, participants read these sentences and rated which target meaning was most likely. While group-averaged ratings showed that meaning likelihood varied continuously across sentences, single-trial ratings were categorically biased towards either meaning. To test whether the neural representation of meaning likelihood is categorical or continuous, we decoded the target word from neural activity and examined the effect of meaning likelihood on cross-meaning generalization. Around 800 ms before target onset, cross-meaning generalization was most accurate for neutral sentence contexts where target meanings were equally likely. Crucially, this improvement in generalization accuracy was parsimoniously modelled by a linear function of meaning likelihood. Thus, although explicit semantic judgments are distributed categorically, neural processing of ambiguous words reflects continuous meaning likelihood.

Topic Area: Language & Communication

Extended Abstract: Full Text PDF