[UMR] Linnaea Stockall (Queen Mary U. London & NYUAD)

 Linnaea Stockall, QMUL & NYUAD

 

I'll present results from an ongoing collaboration with Alec Marantz and Kyriaki Neophytou (NYU/NYUAD) and Christina Manouilidou (U Ljubljana) using MEG to investigate how we detect, parse and assemble morphological constituents in real time. Following on earlier MEG, EEG and behavioral work establishing that form-based morphological decomposition of all potentially complex strings (famously including pseudo-complex words like 'brother' or 'sandal') occurs rapidly and automatically, our goal is to better understand what we then do with the constituents we've extracted. How, when, and with what resources to we detect/compute the grammatical category of a stem/root? Activate and evaluate the various selectional restrictions on affixation? Interpret the result of morphological structure building and evaluate its well-formedness?

 

I'll present data from a pair of experiments in Greek and English involving affixation of suffixes (Greek: -menos, -(s)imos, -tos) and prefixes (English: re-, un-, out-) with varying grammatical category and argument/event structure restrictions on the kinds of stems they can attach to. By measuring the evoked neural responses to legal (refill) and illegal (restrong, relaugh) affix+stem combinations, we can begin to both (a) establish the basic architecture of morphological structure building (b) investigate the grammatical differences between the different affixes.