Detecting self-produced speech errors before and after articulation: an ERP investigation

Area of Science:

• Neuroscience
• Cognitive Psychology
• Speech Science

Background:

• Speech error monitoring is thought to involve the same neural systems as other action error monitoring.
• Behavioral studies suggest speech errors are detected and corrected before articulation.
• The neural mechanisms underlying pre-articulatory speech error monitoring remain unclear.

Purpose of the Study:

• To investigate the neural basis of pre-articulatory speech error monitoring.
• To differentiate neural activity related to pre-articulatory versus post-articulatory speech error monitoring.
• To understand the role of central error monitoring mechanisms in speech production.

Main Methods:

• Utilized a phoneme-substitution task designed to elicit speech errors.
• Employed stimulus-locked event-related potential (ERP) analyses for pre-articulatory monitoring.
• Used response-locked ERP analyses for post-articulatory monitoring.

Main Results:

• A larger P2 component at midline sites (FCz, Cz, CPz) was observed for words preceding speech errors, indicating early detection.
• An anterior N2 component, associated with conflict monitoring, was present for both correct and incorrect utterances.
• Response-locked analyses confirmed a fronto-central error-related negativity (ERN) for self-produced speech errors.

Conclusions:

• Speech errors can be detected by the brain prior to articulation.
• Early positive ERP component (P2) may reflect pre-articulatory error detection or prediction.
• Speech error monitoring involves a central mechanism, with conflict monitoring systems also engaged.