Dissimilation, traditionally described in the context of language change since A. F. Pott (1833), is explored here as consonant dissimilation in the context of speech errors, including remote dissimilation and external sandhi. This process is examined as both a feature of speech production and its outcome. Unlike researchers focused on language change who rely on written documents, this study utilizes spontaneous speech data to investigate dissimilation alongside language change queries, drawing from the speech error corpora of R. Meringer, B. Kettemann, and the author. The analysis covers individual idiosyncrasies, categories of speech errors, and nominal data analysis weighted by spoken language sound frequencies. Consonant dissimilation is discussed based on G. Lindner's criteria for speech gestures, using slow-motion film recordings, and considering tongue twisters, diadochokinesis, and speech tempo. Findings reveal correlations with language change data and suggest causal relationships that necessitate further foundational research into spatio-temporal processes of speech production. Regressive dissimilation arises from conflicting intentions, while more frequent progressive dissimilation is linked to muscle innervation impulses during sound onset. Dissimilation emerges as an inevitable aspect of the universal human processes for producing a wide range of distinguishable speech sounds with a limited set of articulatory op
