Neurophysiological correlates of zero morpheme processing: erp data

Language functioning as a communicative system is described by a multitude of linguistic theories, which are not always consistent with each other and do not have strong cognitive and/or neurobiological bases. One of the most striking examples is the "zero morpheme" proposed by the Universal Grammar theory, which has only an abstract meaning and no phonological representation (for example, null case affix: stol-circle divide (Nominative; table(NOM)) vs. stol-alpha (Accusative; table(ACC))). To address the question of neurocognitive processing of the zero morpheme, we ran an EEG experiment with passive non-attend design. Experimental stimuli included phrases with correct and incorrect pronoun-verb gender agreement (null: he bought-circle divide, filled: *he bought-FEM) and single verbs as a control. During the ERP analysis we found higher activation for the conditions of zero morpheme agreement violations (p = 0.036) started from similar to 200 ms (ELAN), reflecting an early automatic syntactic processing of the zero morpheme. Results support the priming hypothesis for conditions with zero morpheme and its activation during language processing.