Evidence of a domain-general syntax resource: Understanding the P600 response to syntax violations in language and music

Theodore Teow

An unresolved question in the literature relates to the extent to which language and music share syntax-related processing and how this is reflected in brain responses (e.g., certain event-related potential components of the electroencephalogram (ERP; EEG). Across three studies using neurologically healthy adult non-musicians, the goal was to examine the claims of the Shared Syntactic Integration Resource Hypothesis (SSIRH) and P600-as-P3 hypothesis: language and music share syntax-related processing, and that the relatively under-researched P600 component is the best index of this shared neural resource. Experiment 1 used a novel application of single-trial and subset EEG analyses to examine response-alignment across domains; similarities of response-alignment would suggest that the purported shared resource behaves the same way when employed in language and music processing. Participants (N = 16) listened to short sentences or chord progressions, and indicated via timed button-press the presence or absence of a syntax violation. P600 amplitudes were similar across domains, but response-alignment of the P600 occurred for language but not music. This suggests that music syntax errors recruit the P600-related shared resources in a quantitatively similar (ERP), yet qualitatively different (response-alignment) manner to language syntax errors. To explore the interactivity of unattended-on-attended syntax errors, the music and language stimuli were presented simultaneously in Experiment 2 (N = 20) and participants instructed to selectively attend to one domain only. P600 amplitude increased only to attended error conditions in either domain, and unattended error conditions elicited no P600 effect or RT impairment. Experiment 3 used the same methodology as Experiment 2 in a dual-attention task in order to examine combinatory effects. Participants (N = 22) pressed one of three buttons to error-free, single-domain error, and dual-domain error conditions. P600 amplitude and RT increased for both single error conditions relative to controls, and further increased in the dual error conditions. Taken together, the P600 ERP component appears to index a similar cross-domain resource employed in syntax manipulations. Similar patterns of attention-dependence, combinatory increases to dual errors, and of functional co-occurrences to RT verify the P600 as a correlate of processing cost in syntax error integration. However, some differences exist in the lack of response-alignment patterns in music versus language. Further research should determine if this shared resource represents a more general cognitive resource such as part of the P3 family, extending beyond syntax, or even language and music.

Evidence of a domain-general syntax resource: Understanding the P600 response to syntax violations in language and music

Files and links (1)

Abstract

Details

Metrics