Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Semantic neighborhood effects with anterior temporal lobe stimulation
Poster Session D, Saturday, October 26, 10:30 am - 12:00 pm, Great Hall 3 and 4
Cheng Xiao1, Sophie Arheix-Parras1, Karim Johari2, Nicholas Riccardi1, Rutvik H. Desai1; 1University of South Carolina, 2Louisiana State University
Introduction. When processing the meaning of a word, not only are the specific semantic characteristics of the single word retrieved, but the relevant features of its semantic, orthographic, and phonological neighbors are also activated (Buchanan et al., 2001; Carreiras et al., 1997; Yates et al., 2004). Semantic neighborhood effects and especially their neural basis, are not well understood. The present study used HD-tDCS (High-Definition Transcranial Direct Current Stimulation) to investigate the role of the anterior temporal lobe (ATL) in processing both taxonomic and thematic semantic neighborhood distance (SND) in abstract and concrete word processing. Given that in some theories, the temporo-occipital cortex (TOC) is thought to be involved in lexical access (Hickok & Poeppel, 2000, 2004), we also stimulated TOC in a control experiment. Methods. Fifty-four healthy right-handed native English speakers participated in the two experiments targeting the left lateral ATL (28 participants) and left TOC (26 participants). They underwent two separate HD-tDCS sessions (real anodal, sham) that lasted approximately 20 minutes each, on separate days. The stimuli are 304 nouns that included 152 abstract and 152 concrete words based on Brysbaert’s (2014) concreteness ratings. Taxonomic SND was defined based on vector embeddings that emphasize taxonomic, as opposed to associative relationships (Reilly & Desai, 2017). Associative (i.e., thematic) SND was defined based on measures from Shaoul and Westbury (2006). There was no correlation between concreteness and SNDs. The stimuli were divided equally into two versions for real and sham sessions, counterbalanced across participants. Participants were instructed to press the ‘j’ or ‘k’ keys to indicate whether the word was abstract or concrete for each trial as quickly as possible without making errors. All parameters were identical for both ATL and TOC experiments, except for the stimulation site. The order of stimulation conditions and the key assignments were also counterbalanced across participants. We collected the reaction time (RT) of each participant. Results. ATL experiment: The semantic judgment RT was significantly faster in the sham condition compared to the real condition (β=-78.743, SE=10.418, t=-7.558, p<0.001). A linear mixed-effects regression model revealed a significant interaction of associative SND and stimulation type (β =-332.963, SE=164.396, t=-2.025 p=0.043) while accounting for the other psycholinguistic variables. This indicates that stimulating the ATL increased RT for words with sparse associative neighborhood to a greater extent. A marginal interaction between taxonomic SND and the type of stimulation was found (β=-0.385, SE=0.206, t=-1.863, p=0.063), suggesting a trend similar to that seen with associative SND. TOC experiment: There was no difference between the real and sham conditions (β=13.389, SE=9.565, t=1.400, p=0.162). No significant interaction was observed between stimulation type and either SND (associative SND: β =-6.952, SE=150.241, t=-0.046, p=0.963; taxonomic SND: β= 0.223, SE=0.189, t=1.184, p=0.237). Conclusion. The present study highlighted the effects of associative and taxonomic semantic neighborhood on semantic judgments mediated by the left ATL. These results suggest a role of the ATL in processing words with sparse semantic neighborhoods, but are inconsistent with models that propose ATL as a ‘taxonomic hub.’
Topic Areas: Meaning: Lexical Semantics, Reading