You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster C50, Thursday, November 9, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

The Loci of the Semantic Relatedness Paradox during Speech Production

Tao Wei1, Tatiana Schnur2;1Beijing Normal University, 2Baylor College of Medicine

Speech is affected by recent naming experience (i.e., what we have said before). In previous work, we found that the same naming experience (e.g., “cat”) has opposite effects on future naming of semantically related pictures (e.g., “dog”). Facilitation was observed when semantically related pictures were named adjacently (lag0 condition), while the effect reversed into interference when semantically related pictures were named with unrelated intervening items (lag2 condition). There is ongoing debate about the stage(s) in the language system at which facilitation and interference take place (semantic vs. lexical processing; e.g., Belke, 2013; Biggs & Marmurek, 1990; Damian et al., 2001; Wei & Schnur, 2016). Previous studies have attempted to answer this question using functional magnetic resonance imaging (fMRI), by localizing the neural locus of interference effects (e.g., de Zubicaray, McMahon, Eastburn, & Pringle, 2006; de Zubicaray, McMahon, & Howard, 2013). However, we do not know whether a brain region involved in facilitation and/or interference is in fact important for lexical or semantic processing without testing the cognitive function (semantic vs. lexical) of the region across subjects. In this study, we directly addressed this question by localizing in the same subjects the neural basis of specific stages in speech production (i.e., semantic vs. lexical processing) while concurrently testing the semantic relatedness paradox in naming (i.e., facilitation vs. interference). First, regions of interest (ROIs) subserving semantic and lexical processing were identified as brain regions where increased activity was correlated with decreasing concept familiarity and lexical frequency respectively (e.g., Graves, Grabowski, Mehta, & Gordon, 2007; Wilson, Isenberg, & Hickok, 2009). We identified semantic ROIs in the bilateral inferior temporal lobes, and lexical ROIs in the left posterior temporal gyrus and right occipitotemporal cortex. Then, within these semantic and lexical ROIs, we contrasted the brain activity when subjects named semantically related vs. unrelated pictures in the lag0 and lag2 conditions, respectively. The results revealed that for lag0, where semantic facilitation was observed, only semantic ROIs in posterior inferior temporal regions showed stronger activity in naming related vs. unrelated pictures, suggesting a semantic locus of facilitation in naming. In contrast, for lag2 where semantic interference was found, all semantic and lexical ROIs showed weaker activity in naming related vs. unrelated pictures, suggesting both semantic and lexical loci of interference in naming. These results show that previously naming semantically related words facilitates current naming at the level of semantic processing, whereas naming is slowed down because of changes at both semantic and lexical levels of processing. The next step is to understand the exact mechanism of the changes (e.g., short-lived changes in activation levels and long-lasting changes in connection weights) which result in facilitation vs. interference in naming dependent on level of processing.

Topic Area: Meaning: Lexical Semantics

Back to Poster Schedule