Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Category vs. Semantic Morphological Rule Violations: MEG Comparisons from South Slavic
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster A39 in Poster Session A, Tuesday, October 24, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Bojana Ristić1, Karin Kavčič1, Rok Žaucer2, Linnaea Stockall3, Christina Manouilidou1; 1Universty of Ljubljana, Slovenia, 2University of Nova Gorica, Slovenia, 3Queen Mary University of London, UK
Previous studies have shown dissociation between stem+affix well-formedness based on category and semantics when processing morphologically complex words. Using pseudoword paradigm with Greek suffixes, Neophytou et al. (2018) showed more activation in left temporal lobe in the 200-300ms time window for the pseudowords violating category-selection rules, and more activation in the orbitofrontal cortex in the 300-500ms window for the pseudowords violating verb argument structure. Comparable pattern of results was obtained in Stockall et al. (2019) with English prefixes. However, with evidence coming from argument structure rule violations only, it’s not clear whether the later sub-process relates to semantic information other than argument structure. We created 2 experiments with equivalent designs in two closely related South Slavic languages, Slovenian and BCS, with pseudowords violating prefix attachment rules. Using 3 prefixes that are equivalent in Slovenian and BCS (raz, od, vz/uz), we compared category-selection rules violation (e.g. *razmajka, “raz-mother”) to purely semantic (rather than argument structure) rule violations, where the violated dimension was that of state stability/durativity (*razčutiti, “raz-feel”). 16 native speakers of Slovenian and 23 native speakers of BCS performed a lexical decision task in their respective languages, with the concurrent MEG recordings. 1000 Hz sample rate was used on a 208-channel axial gradiometer system. We extracted 600 epochs for each stimulus, and conducted a two-stage regression analysis in which regressions were fit at each time point and source point per subject. We conducted spatio-temporal cluster-based permutation tests on the one-sample t-test values derived from the beta coefficient of the regressions in left temporal lobe and orbitofrontal cortex, in line with Neophytou et al. (2018) and Stockall et al. (2019). The preliminary analyses of the Slovenian data shows a marginally significant cluster for the Condition effect in the orbitofrontal cortex that spans from 334 to 394 ms (p=0.06), with more negative activity for semantic violations. In the BCS data, we found a later significant cluster for the Condition X Prefix interaction spanning from 500 to 515 ms (p=0.03), where more negative activity was found for the semantic violations with the prefix –uz. No significant clusters were found in the left temporal lobe analysis in the 200-300ms time window for either of the languages. These results confirm involvement of the orbitofrontal cortex in the processes related to semantic well-formedness in morphologically complex words, coming from designs that rely on more purely semantic rules, from two closely related South Slavic languages. Further analyses are necessary to test whether these effects hold across prefixes, as well as to understand if the lack of early, category-selection based effects in the left temporal lobe is specific to these languages.
Topic Areas: Morphology,