hide
Free keywords:
Temporal predictability; Beat; Meter; Syntax; Speech; Basal ganglia; Cerebellum
Abstract:
The current work set out to answer three questions: (1) Are reported syntactic deficits in patients with structural damage to the basal ganglia (BG) in the cortico-striato-thalamo-cortical systems (CSTCS) the result of a syntax specific computational deficit or are they potentially a consequence of a generalized timing deficit? (2) Do BG patients suffer from a simple beat perception deficit in speech comparable to the one reported in music? (3) Can regular speech meter (i.e., a pattern of beats induced by the regular alteration of stressed and unstressed syllable accents) ameliorate the computation of syntactically marked information by making speech events temporally predictable and salient? The latter ‘remediation’ hypothesis would predict that when speech events (i.e., those that are syntactically marked) are metrically aligned to the syllabic accent structure, the computation of syntactic information is facilitated or in the case of patients ameliorated. During continuous EEG measurement nineteen patients with focal BG lesions and matched healthy controls listened to metrically regular and syntactically well-formed sentences and metrically well-formed sentences that either violated syntactic expectancy, metrical expectancy, or both. While healthy controls showed an expected P600 response in the event-related brain potential (ERP) to all expectancy violations, BG patients showed overall comparable P600 responses to all, but the metrical expectancy violation. These results confirm that (1) BG patients suffer from a simple beat perception deficit in speech and (2) regular speech meter ameliorates the computation of syntactically marked information in the speech signal. We propose that a domain general sensorimotor cerebello-thalamo-cortical system (CTCS), involved in event-based temporal processing, engages in the remediation of dysfunctional cortico-striato-thalamo-cortical timing that affects the timely computation of linguistic (i.e., syntax) information in the speech signal.
