"Francine called the clown a doctor": Investigating Neurocognition of Syntactic Ambiguity with EEG

Abstract

The purpose of this study is to investigate the neurocognitive processes of conflict resolution between syntactic (structure-related) and semantic (meaning-related) information during sentence-reading. English ditransitive verbs provide a unique vehicle to examine this question. Ditransitive sentences include two major structures--attributive (“characteristic”; ex.“Francine called the clown a liar”) and double-object dative (“giving”; ex. “Francine called the clown a taxi”). Distinguishing between these requires using semantic cues to determine which syntactic frame is a better fit. For my experiment, I developed a unique set of 250 ditransitive sentences, 50 of which were ambiguous (ex. “Francine called the clown a doctor”). Participants (n = 17) were trained to classify ditransitive sentences and were then shown each of the 250 sentences one at a time while neural activity was recorded using EEG. After each sentence, participants responded with a 4-point confidence+classification rating. Sentence type showed a significant effect on response time, F(2,32)= 45.46, p < .001, η2 = .74. As predicted, participants took significantly longer to classify ambiguous sentences (M = 2083 ms) than characteristic (M = 1519 ms) or giving (M = 1429 ms) sentences. EEG waveforms showed distinct differences between the three sentence types. Together, these results support the hypothesis that resolving conflicts presented by ditransitive sentences requires additional processing effort consistent with the second and third stages of Friederici’s (2017) neurocognitive model of language processing. Understanding the relationship of syntax and semantics in the brain can support development of innovative treatments for individuals with reading barriers or speech disorders such as aphasia.