A « subtle phonetic deficit » in fluent aphasia ? An acoustic and perceptual study / Anna Marczyck

A « subtle phonetic deficit » in fluent aphasia ? An acoustic and perceptual study / Anna Marczyck, in "Perspectives neuropsycholinguistiques sur l'aphasie - NeuroPsychoLinguistic Perspectives on Aphasia", colloque international organisé par l'Unité de Recherche Interdisciplinaire Octogone de l'Université Toulouse II-Le Mirail (France). Toulouse, 21-23 juin 2012.

The current study addresses the question of the existence of a « subtle phonetic deficit » in fluent aphasic patients. Acoustic investigations of fluent aphasic speech (Vijayan & Gandour, 1995 for a revision) reported on the existence of excessively long voicing leads (for English), longer segments durations and greater variability in the segments’ productions. Several explanations have been provided to account for these abnormalities: the possibility of the motoric functions being, at least partly, processed in the left posterior region, the existence of disturbances at the level of phonological planning mechanisms or the inability to use adequately auditory feedback. 
This paper wishes to shed some light on the question of the presence of a subtle phonetic impairment in fluent aphasic patients’ oral production. Two research questions have been specifically raised: (i) are there significant differences in duration between fluent aphasic patients and normal control subjects?; (ii) Do fluent aphasic patients respect the phonological rules of Spanish, such as the opposition of voicing expressed in the corresponding difference of phonemes’ durations (voiceless stops are longer than voiced stops)?
Method.
The aphasic group comprised 5 fluent, Spanish speaking, aphasic patients: 3 patients with conduction aphasia and 2 patients with mixed aphasia. The control groups were composed of 5 normal subjects and of a Broca’s aphasic patient . The stimuli included Spanish stops, fricatives and affricates, in different phonetic context, position of stress and vocalic environment. All aphasic subjects underwent both reading and repetition tasks. Their performance was examined for the following acoustic parameters : total segment duration, number of noise bursts, duration of noise bursts, frication duration, and VOT.
Results
Preliminary results show that fluent aphasic patients produce generally shorter stops and fricatives in comparison to control subjects. This result is correlated to a correct perception and identification of the phonemes, which suggests that there may be an (only) subtle phonetic impairment in fluent aphasic subjects. 
Control subjects use the duration parameter to establish a significant difference between voiced and voiceless stops (p=.000) but do show evidence of phonetic (phonologically not pertinent) changes in noise bursts number and duration or in frication duration to complement such a phonological distinction. There is however a fair amount of variability in the fluent aphasic group. Two patients seem to use the parameter of duration to distinguish between voiced and voiceless stops (p=.000 in both cases), while three other patients do not use this parameter but instead establish the phonological difference using the parameter of duration of noise bursts or in the number of noise bursts (p= .002 and p=.001). This result indicates (i) that these patients may use different from normative acoustic cues to establish phonological oppositions or (ii) that their deficit may arise at some deeper level of phonological processing. These results will be discussed in more detail in relation to both linguistic theories and cognitive models of speech processing.