Stuttering is a communicative disorder that involves disruptions to fluent speech, which are characterized by frequent repetition or prolongation of syllables or words, and/or by frequent hesitations or pauses. Prior research has identified a number of hallmarks and deficits associated with stuttering, including generalized timing deficits, and reduced functional connectivity in the rhythm network previously identified to be involved in the perception of musical meter. Building on assumptions of (1) shared neurocognitive resources exist for metrical structure-building for perception and production of auditory patterns in both music and language, and (2) functional similarity exists in processes involved in predictive action-preparation from metrical structure in auditory information, it was predicted that a core deficit in stuttering involves deficiencies in integrating candidate metrical structures with sensory evidence that would support them. To test this prediction, an experiment was designed using a same/different rhythm discrimination task. Critically, half the stimuli provided greater support for the induction of a beat/meter (“simple rhythms”), whereas the other half were matched in interval types but provided less signal-based statistical support for the induction of a beat/meter (“complex rhythms”). Participants were 36 adults who do and do not stutter, and the rhythm discrimination task was done while undergoing functional magnetic resonance imaging. For the behavioral results, statistical analyses using analysis of variance/covariance did not show any significant effects or interactions; however, a linear mixed effects model which accounted for multiple sources of variance revealed poorer performance on complex rhythm discrimination by adults who stutter compared with those that do not stutter. For the neural results, activation in the core rhythm network during the rhythm discrimination task was observed for both groups (bilateral insula, bilateral STG, bilateral SMA, and bilateral premotor area) for both simple and complex rhythms. However, adults who stutter additionally showed activation in the bilateral putamen and bilateral IFG, suggesting that one or both of these areas may perform a compensatory function in rhythm perception and predictive action-preparation. These results can be interpreted with respect to predictive coding processes in the brain supporting perception, action, and cognition, as well as recent conceptual extensions to auditory processing of music, language, and speech, which propose that linguistic perception and production are “two sides of the same coin.” Specifically, it is proposed that listeners attempt to build top-down metrical representations for structured auditory sequences, and during language processing, these top-down metrical representations must be merged with representations of other structures in language to give rise to a coherent overall linguistic representation. It is proposed that a core deficit in stuttering involves deficient processes for integration of top-down metrical/prosodic structure and/or bottom-up sensory indices of dynamic sensorimotor states, toward construction of a coherent overall linguistic representation. Evidence for this proposal comes from findings that: (1) distal context rate and rhythm cues in speech influence metrical/prosodic structures heard across identical acoustic material, thereby influencing goodness-of-fit evaluations of alternative top-down candidate representations of lexico-syntax; (2) a hallmark of stuttering is anomalous white matter connectivity and reduced functional organization of rhythm networks in the brain. This study is the first to investigate non-speech rhythm perception in adults who stutter, and the findings suggest new hypotheses regarding how dynamic connections among brain structures (e.g., basal ganglia, STG) instantiate computations toward the imputation of timing and meter from acoustically variable auditory signals.
Read
