Is there a syllable frequency effect in aphasia or in apraxia of speech or both?

Background: The observation of a syllable frequency effect on production latencies in healthy speakers has been an argument in favour of stored syllables in speech production. In Levelt, Roelofs, and Meyer's (1999) model of speech production, syllabic representations are accessed during phonetic encoding. Neurolinguistic studies have provided convergent evidence of a syllable frequency effect on production accuracy in speakers with acquired language disorders. However, the observation that syllable frequency also affected production in aphasic speakers with a pre-phonetic impairment (conduction aphasia and Wernicke's aphasia) seems in contradiction to the phonetic locus of syllabic representations. Aims: We illustrate the points of convergences and divergences between psycholinguistic and neurolinguistic results on the locus of the syllable frequency effect and explore whether a syllable frequency effect is observed in apraxia of speech (AoS) and in conduction aphasia when participants are tested with the same material. Methods Procedures: Reading and repetition was elicited with monosyllabic words (Experiment A) and with bisyllabic pseudowords (Experiment B) composed of high- or low-frequency syllables. Three speakers with AoS and three speakers with conduction aphasia participated in each experiment. Outcomes Results: Both subgroups displayed a tendency for a syllable frequency effect on production accuracy. A significant effect of syllable frequency was observed in each experiment in a participant with AoS and in a participant with conduction aphasia. Conclusions: The data confirmed similar syllable frequency effects in speakers with AoS and in conduction aphasia when tested with the same eliciting material. We discuss these apparently contradictory observations and suggest an explanation for the origin of the syllable frequency effect in these two populations.