Dysprosody is typically associated with the dysarthrias following acquired neurogenic disturbance but it has also been associated with impairment to the cortex of the right cerebral hemisphere (Behrens, 1985). Currently there is little agreement in the clinical literature as to the locus or nature of processes involved in production and reception of prosody. This paper investigates the usefulness of two psycholinguistic models of prosody--involving "abstract" and "concrete" processes (Ladd & Cutler, 1983)--in accounting for dysprosody following motor pathway as well as cortical lesions. Four groups of 10 adult patients each participated in the study; the first group had sustained a right cerebral hemisphere cortical stroke, the second group had hypokinetic dysarthria, the third group had ataxic dysarthria, and the fourth group was normal controls. Acoustic analysis using the Kay VisiPitch/PC was conducted on pairs of matched noun phrase and noun compounds (e.g., greenhouse, green house), interrogative and declarative versions of sentences, and sentences spoken in each of four different emotional tones. Right cerebral hemisphere damaged subjects produced shorter durations than the other groups for each stimulus class. However, all of the subjects were able to use duration to signal temporal information for the lexical distinction between noun compounds and noun phrases. The results are consistent with a model of prosodic control which involves both cortical as well as physiological control processes. Current treatment for dysprosody is discussed in light of this new emphasis on cognitive control processes.