News & Stories

In a new theoretical paper with Frank Guenther, Ph.D., and Jonathan Cannon, Ph.D., we propose a neurocomputational framework for understanding how fluent speech is initiated and how this process may be disrupted in stuttering. More broadly, the framework focuses on: what neural conditions need to be established for fluent speech to begin? 

The central idea is that before a learned speech motor program can unfold, the brain must first reach an appropriate preparatory neural state—a "basin of attraction." We propose that cortico-basal ganglia-thalamo-cortical circuits play an important role in guiding distributed speech networks into this state. If this initialization process is unreliable, speech may fail to launch smoothly resulting in repetitions, prolongations, or blocks.

One aspect of the theory that I find particularly interesting is its potential to provide a plausible mechanistic explanation for a number of otherwise puzzling features of stuttering: its variability across speaking contexts, its prominence at speech onset, and the sometimes dramatic effects of external timing cues and altered sensory feedback on fluency.

By bringing concepts from neural population dynamics and motor preparation into the study of stuttering, the theory generates a number of specific predictions that can now be tested using neuroimaging and neurophysiological approaches.

Interested in reading the paper in full? The preprint is available now.