Gradient co-activation and speech error articulation: Comment on pouplier and goldstein (2010)

Recent studies show that speech errors involve the co-production of the phonetic properties of both targets and error outcomes. Based on the spatial and temporal properties of these co-productions, Pouplier and Goldstein argued that they are influenced by speech production mechanisms that detect and suppress errorful articulations. In this commentary, we provide simulation data supporting an alternative account based on cascading activation. Novel simulation results in the Gradient Symbol Processing framework show that in speech errors target and error phonological representations are gradiently co-activated. Using the TADA system, we show how these types of coactivation patterns could give rise to the articulatory patterns observed by Pouplier and Goldstein. These results show that cascading activation could provide a plausible account not only for the presence of co-productions in speech errors but also for their temporal and spatial properties.