EEG evidence for a three-phase recurrent process during spoken word processing

Spoken word comprehension requires functionally hierarchical stages of analysis to conjunct acoustic-phonetic input with mental lexicon. How the specific process unfolds along time axis is still unclear. In this study, we utilized electroencephalograph (EEG) based source reconstruction technique to reveal spatiotemporal brain dynamics when 22 participants listened to disyllable nouns and verbs. Current density reconstruction (CDR) algorithm with standardized low resolution electromagnetic tomography (sLORETA) constraint was conducted to reconstruct brain activation maps. Results showed that noun-verb neural dissociation occurred in three periods of 50∼100 ms, 150∼250 ms, and 350∼450 ms, which suggest that semantic access may take place at a quite early stage and repeat in a recurrent fashion. The three durations can be regarded as a three-phase recurrent process (TPRP) for spoken word processing: tentative access phase, primary selection phase, and advanced integration phase. The proposed TPRP will advance our understanding on speech comprehension and provide new sight for further research.