Academic journal article Journal of Medical Speech - Language Pathology

Use of Prosody by Children with Severe Dysarthria: A Cantonese Extension Study

Academic journal article Journal of Medical Speech - Language Pathology

Use of Prosody by Children with Severe Dysarthria: A Cantonese Extension Study

Article excerpt

Several recent studies have shown that speakers with severe dysarthria can make use of prosody to communicate intentions. Patel and Salata (2006) investigated the abilities of five children with severe dysarthria to convey three pitch levels and three duration lengths to their caregivers, using an interactive computer game. In this extension study, four Cantonese-speaking children with severe dysarthria were asked to produce five durations and five pitch levels. Rising and falling pitch levels were included to simulated the rising and falling lexical tones of Cantonese. Acoustic analysis was performed for all productions to determine the relationship between acoustic cues and listener perception. There was substantial interparticipant and intraparticipant variability in the children's productions and in caregiver accuracy. However, accuracy was above chance level for a number of prosodic targets for each speaker. Some of the perceptual responses could be easily explained by the acoustic data, where clear distinctions were apparent. However, in other cases there was much overlap in the acoustic signal, suggesting that listeners were employing additional cues. The results of this study strengthen previous findings that prosody can be used reliably by speakers with severe dysarthria to communicate with familiar partners. Further cross-linguistic studies would be informative.

**********

The dimension of prosody has been relatively neglected in studies of dysarthric speech (Leuschel & Docherty, 1996). Investigations of prosody in this population are of both theoretical and practical interest. Previous studies have shown that speakers with severe dysarthria characterized by reduced segmental intelligibility may be able to make use of prosodic cues to convey their intentions (cf., Le Dorze, Ouellet, & Ryalls, 1994; Patel 2002a, 2003, 2004; Yorkston, Beukelman, Minifie, & Sapir, 1984). Patel (2002a) showed that adults with severe dysarthria were able to consistently produce sustained vowels at three distinctive durations and two distinctive pitches. These same speakers could also signal the question-statement contrast using prosodic cues, and listeners were highly accurate at classifying their productions (Patel 2002b, 2003). This preserved ability to signal prosodic contrasts has also been noted in contrastive stees taskls in which speakers with dysarthria utilize combinations of fundamental frequency (F0), intensity, and duration to signal stress (Patel, 2004; Patel & Watkins, 2007; Yorkston et al., 1984).

More recently, Patel and Salata (2006) used an interactive computer game to investigate communicative interactions between five children with severe dysarthria and their caregivers. The children were asked to produce the sustained vowel /a/ at three distinct pitch levels (high, medium, low) and three distinct durations (long, medium, short), as well as nine combinations of these targets, which their caregivers were asked to identify. Caregiver accuracy was variable across the five dyads, but accuracy was above chance level for several targets produced by each speaker. Accuracy for pitch and duration targets in isolation was generally higher than for combined conditions. The current study was designed as an extension study of Patel and Salata's work. Since caregiver accuracy was relatively high for the three duration targets in the Patel and Salata study, we added two further duration parameters. Also, we added two pitch categories, rising and falling, to simulate the lexical tone patterns of Cantonese.

Cantonese has six contrastive lexical tones: high level, mid level, low level, high rising, low rising, and low falling. We were interested in exploring the issue of whether tonal-language listeners are better at perceiving pitch differences than nontonal language listeners. Francis and Ciocca (2003) found that monolingual English-speaking listeners and native Cantonese-speaking listeners performed similarly in tasks that involved discrimination of speech sounds differing in F0; the Cantonese listeners were less sensitive than the English listeners to pitch differences for non-speech sounds. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.