research (2025): Delayed auditory feedback increases speech production variability in typically fluent adults but has the opposite effect in stuttering adults

research (2025): Delayed auditory feedback increases speech production variability in typically fluent adults but has the opposite effect in stuttering adults **Authors’ interpretation:** The increased trial-to-trial variability in AWS (adults who stutter) is interpreted as evidence of inefficient feedforward control: either more variable motor commands or stronger trial-by-trial adjustments driven by perceived errors. The authors propose a two-component account that could produce greater variability: * AWS may select motor commands from a wider distribution (more variable motor output), and/or * AWS may have higher sensitivity to small errors, causing larger corrective adjustments from one trial to the next. The differential DAF effect is interpreted as follows: because DAF induces large perceived errors, AWS may discount these large errors (making smaller trial-to-trial adjustments under DAF), whereas ANS respond to those large errors with larger adjustments (increasing variability). Thus AWS could be over-sensitive to small errors but relatively insensitive to large errors—a profile that could also explain their reduced adaptive/corrective responses in other perturbation studies. **Relation to existing hypotheses** The proposed sensitivity profile aligns with existing views that stuttering involves overreliance on (auditory) feedback or overactive monitoring; the authors refine these ideas by suggesting the over-sensitivity may be limited to small errors, with a contrasting discounting of large errors. **Suggestions for future work** Test whether increased trial-to-trial variability in AWS is a compensatory strategy or a core deficit by sampling across ages, severity levels, and natural speech contexts (sentences, conversation).