The sense of sounds: brain responses to phonotactic frequency, phonological grammar and lexical meaning

Abstract

Two outstanding questions in spoken-language comprehension concern (1) the interplayof phonological grammar (legal vs. illegal sound sequences), phonotactic frequency(high- vs. low-frequency sound sequences) and lexicality (words vs. other soundsequences) in a meaningful context, and (2) how the properties of phonologicalsequences determine their inclusion or exclusion from lexical-semantic processing. Inthe present study, we used a picture-sound priming paradigm to examine the ERPresponses of adult listeners to grammatically illegal sound sequences, to grammaticallylegal sound sequences (pseudowords) with low- vs. high-frequency, and to real wordsthat were either congruent or incongruent to the picture context. Results showed lessnegative N1-P2 responses for illegal sequences and low-frequency pseudowords (withdifferences in topography), but not high-frequency ones. Low-frequency pseudowordsalso showed an increased P3 component. However, just like illegal sequences, neitherlow- nor high-frequency pseudowords differed from congruent words in the N400.Thus, phonotactic frequency had an impact before, but not during lexical-semanticprocessing. Our results also suggest that phonological grammar, phonotactic frequencyand lexicality may follow each other in this order during word processing.