Abstract for Heather van der Lely's talk

ERP responses in language-impaired children reveal a domain-specific neural correlate for syntactic dependencies
Heather van der Lely (University College, London)
November 8, 2005 (Tuesday)
6:30 PM - 8:00 PM; Room 7102, The CUNY Graduate Center

Considerable controversy surrounds whether neural circuitry subserving grammar is common to human cognition or whether it is unique. Grammatical-specific language impairment (G-SLI) is a disorder of language acquisition - specifically grammatical acquisition - in children who otherwise appear to be normally developing. Some researchers claim a domain-general deficit in auditory processing speed or capacity causes G-SLI, whereas others claim impairment to a domain-specific system devoted to grammar itself causes G-SLI. We found evidence from electrical brain responses for a selective impairment to grammatical processing of language in G-SLI.

First, we checked for pure auditory processing deficits by recording ERPs to tones in a classical attended odd-ball paradigm. Our results did not reveal any differences in G-SLI children and age-matched controls in the early sensory components (N1, P2) or the later component associated with discrimination and categorisation (P300). Second, we presented our 10 to 21 year-old participants with G-SLI, age, and younger language control groups and adults with questions containing syntactic dependency violations and sentences containing semantic anomalies. We found that the brain's automatic detection response to syntactic violations, reflected electrically as an "Early Left Anterior Negativity" (ELAN) was elicited in all control groups and adults, but not the G-SLI group. In contrast to the controls, the G-SLI participants exhibited a typical N400 to the syntactic violations, indicating that they were compensating for their syntactic deficit, semantically. However, the G-SLI group, like the control and adult groups, exhibited a normal central positive response around 600ms (P600), often associated with re-analysis of sentences to the syntactic violations and an N400 to semantic violations. These data provide an objective measure for a developmental, domain-specific grammatical deficit. Such phenotypic data provide for future links between genotypic discoveries and the functional architecture of the human brain.