Conference Contribution Details
Mandatory Fields
Lee, A., Kong, A., & Law, S. P.
The Academy of Aphasia 47th Annual Meeting
Acoustic analysis of prosody for normal and aphasic discourse of Cantonese speakers
Boston, MA, United States
Poster Presentation
Optional Fields
Background: Previous studies on prosodic problems in aphasia have focused the investigation on the sentence level. However, the evaluation should be based on connected sample as it is considered the most sensitive task to different dimensions of atypical prosody (Leuschel & Docherty, 1996). Recently, Tseng, Pin, Lee, Wang and Chen (2005) proposed a topdown, multi-layered framework, which considers all relevant levels (prosodic phrase group, breath group, prosodic phrase, prosodic word, and syllable) that constitute discourse prosody. While Tseng et al.ís model is developed based on speech samples of normal Mandarin speakers, the present paper reports a pilot study that has applied this framework to analyze discourse prosody in Cantonese-speaking normal speakers and individuals with aphasia. Method: The subjects were two individuals with aphasia (one female of 47 years old with transcortical motor aphasia and one male aged 49 years old with anomic aphasia) and two age-, gender- and education level-matched normal individuals (one 44-year-old female and one 47-year-old male; education level was secondary or below). Each speaker partook in passage reading, picture description, story telling and monologue. The speech samples were recorded using a digital recorder and a condenser microphone in a quiet room. Broad phonetic transcription, boundary break annotation (based on Tseng et al.ís guidelines) and acoustic analysis were conducted using Praat version 5.1.02 (Boersma & Weenink, 1992-2009). The following parameters were measured: fundamental frequency, peak intensity, duration of each syllable (in millisecond), duration of each break, speaking rate and articulation rate (in syllable per minute). Results: The speakers with aphasia showed lower speaking rate and articulation rate than the normal speakers because more and longer breaks and prolonged syllables were used. For intonation, smaller variations in fundamental frequency and intensity were observed in the speakers with aphasia compared to normal speakers. The results of some of the acoustic measures also differed according to speech tasks. In general, the speakers showed lower speech rate and smaller intensity variation in reading than in picture description, story telling and monologue. Discussion and conclusion: The lower speech rates were probably due to word retrieval problems in the speakers with aphasia. The smaller fundamental frequency and intensity variations in the speakers with aphasia might be related to the use of more breaks, which has disrupted the overall intonation pattern. Tseng et al.ís framework is shown to be applicable for discourse by normal Cantonese speakers. It can be extended to capture the prosodic disturbance in aphasia through modification of the boundary breaks annotation guidelines and additional parameters, such as measuring the use of inappropriate breaks based on language analysis. References: Boersma, P., & Weenink, D. (1992-2009). Praat (Version 5.1.02). Amsterdam. Leuschel, A., & Docherty, G. J. (1996). Prosodic assessment of dysarthria. In D. A. Robin, K. M. Yorkston & D. R. Beukelman (Eds.), Disorders of motor speech: Assessment, treatment, and clinical characterization (pp. 155-178). Baltimore, MD: Paul H. Brookes. Tseng, C., Pin, S., Lee, Y., Wang, H., & Chen, Y. (2005). Fluent speech prosody: Framework and modeling. Speech Communication, 46, 284-309.