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

Poster B42, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

ERP and fMRI exploration of the organizational structure of abstract versus concrete words in neurotypical adults

Chaleece Sandberg1;1Penn State University

The nature of the concreteness effect and therefore the underlying representation and organization of abstract and concrete words in the semantic system is currently under debate. One recent hypothesis is that abstract words are organized through associative connections, while concrete words are organized into natural categories (Crutch, Connell, & Warrington, 2009). This project aims to determine if this hypothesis is supported with neuroimaging data. It is predicted that words in their ‘dispreferred’ context (concrete = associative; abstract = categorical/similarity) will have higher activation and more of an N400 effect than words in their ‘preferred’ organizational context (concrete = categorical/similarity; abstract = associative). Nineteen right-handed English-speaking neurotypical adults (NTA; 8 male), ranging in age from 18-64 (µ=36) have participated in the EEG portion of the experiment to date. A subset of sixteen NTA (7 male), ranging in age from 18-62 (µ=36), have participated to date in the fMRI portion of the experiment. Each participant performed the same semantic-relatedness task during both the EEG and fMRI scan. The semantic-relatedness task was adapted from Crutch et al. (2009). Forty semantically-related word pairs were in each of four conditions: abstract similar, abstract associated, concrete similar, and concrete associated. Unrelated word pairs balanced yes/no responses and served as a baseline for the N400. The abstract and concrete word pair lists were matched on frequency and familiarity, but differed on imageability and concreteness. ERP data were analyzed using BrainVision Analyzer software. Preprocessing included ICA to mitigate the effects of eye blinks. Repeated measures ANOVA were used to compare conditions at three electrodes – Fz, Cz, and Pz – within the N400 time window. fMRI data were analyzed using SPM12. Routine preprocessing was conducted and a general linear model was used for statistical analysis. The contrasts of interest included main effects and interactions of concreteness (abstract vs. concrete) and relation type (association vs. similarity). For the ERP data, there was a statistically significant (p<.05) N400 effect for the unrelated items compared with the related items. The concrete similarity condition had a significantly lower (p<.05) N400 than every other condition at Pz. While the other conditions were not significantly different from each other, the fact that the concrete similarity condition had the lowest N400 peak is in line with the predictions. For the fMRI data, there was a main effect of concreteness in line with previous work (Binder, Desai, Graves, & Conant, 2009). There was also a main effect of relation type, and an interaction effect. In right inferior frontal gyrus, more activation was noted for concrete words in a similarity context and abstract words in an associative context (preferred). Right hippocampus, right middle occipital gyrus, left superior frontal gyrus, and left superior occipital gyrus showed more activation for concrete words in an associative context and abstract words in a similarity context (dispreferred). While not specifically predicted, these results suggest that different regions may support processing related to preferred versus dispreferred semantic contexts. Combined, these data suggest there may be differential semantic organization for abstract and concrete words.

Topic Area: Meaning: Lexical Semantics

Back to Poster Schedule