Neural correlates of early-closure garden-path processing: Effects of prosody and plausibility

Abstract

Functional magnetic resonance imaging (fMRI) was used to investigate neural correlates of early-closure garden-path sentence processing and use of extrasyntactic information to resolve temporary syntactic ambiguities. Sixteen participants performed an auditory picture verification task on sentences presented with natural versus flat intonation. Stimuli included sentences in which the garden-path interpretation was plausible, implausible because of a late pragmatic cue, or implausible because of a semantic mismatch between an optionally transitive verb and the following noun. Natural sentence intonation was correlated with left-hemisphere temporal activation, but also with activation that suggests the allocation of more resources to interpretation when natural prosody is provided. Garden-path processing was associated with upregulation in bilateral inferior parietal and right-hemisphere dorsolateral prefrontal and inferior frontal cortex, while differences between the strength and type of plausibility cues were also reflected in activation patterns. Region of interest (ROI) analyses in regions associated with complex syntactic processing are consistent with a role for posterior temporal cortex supporting access to verb argument structure. Furthermore, ROI analyses within left-hemisphere inferior frontal gyrus suggest a division of labour, with the anterior–ventral part primarily involved in syntactic–semantic mismatch detection, the central part supporting structural reanalysis, and the posterior–dorsal part showing a general structural complexity effect.

Keywords:

• Garden-path sentences
• Prosody
• Plausibility
• Functional magnetic resonance imaging
• Sentence processing

ORCID

Dirk-Bart den Ouden http://orcid.org/0000-0001-6971-7606

Michael Walsh Dickey http://orcid.org/0000-0002-9068-3313

We are grateful to Erla Hafsteinsdottir and Emily Garnett for their contributions.

Notes

¹Petersson et al.’s (2004) summary of the review by Bookheimer (2002) reports Talairach coordinates [−42, 25, 4], [−44, 19, 12], and [−46, 11, 26], which we transformed to MNI coordinates (using Tal2MNI via the WFU_Pickatlas toolbox in SPM).