The influence of dialect upon the perception of dysarthic speech.

The purpose of this study was to determine the influence of dialect upon the perception of dysarthric speech. Speakers and listeners were self-identifying as either Caucasian American or African American. Three speakers were Caucasian American, three were African American. Four speakers had experienced a CVA and were dysarthric. Listeners were age matched and were equally divided for gender. Readers recorded 14 word sentences from the Assessment of Intelligibility of Dysarthric Speech. Listeners provided ratings of intelligibility, comprehensibility, and acceptability. Own-race biases were found for all measures; however, significant findings were found for intelligibility and comprehensibility in that the Caucasian Americans provided significantly higher scores for Caucasian American speakers. Clinical implications are discussed.

Acoustic characteristics of ataxic speech in Japanese patients with spinocerebellar degeneration (SCD).

BACKGROUND: In English- and German-speaking countries, ataxic speech is often described as showing scanning based on acoustic impressions. Although the term 'scanning' is generally considered to represent abnormal speech features including prosodic excess or insufficiency, any precise acoustic analysis of ataxic speech has not been performed in Japanese-speaking patients. This raises the question of what is the most dominant acoustic characteristic of ataxic speech in Japanese subjects, particularly related to the perceptual impression of 'scanning'. AIMS: The study was designed to investigate the nature of speech characteristics of Japanese ataxic subjects, particularly 'scanning', by means of acoustic analysis. METHODS & PROCEDURES: The study comprised 20 Japanese cases with spinocerebellar degeneration diagnosed to have a perceptual impression of scanning by neurologists (ataxic group) and 20 age-matched normal healthy subjects (control group). Recordings of speech samples of Japanese test sentences were obtained from each subject. The recorded and digitized acoustic samples were analysed using 'Acoustic Core-8' (Arcadia Inc.). OUTCOMES & RESULTS: Sentence duration was significantly longer in the ataxic group as compared with the control group, indicating that the speaking rate was slower in the ataxic subjects. Segment duration remained consistent in both vowels and consonants in the control group as compared with the ataxic group. In particular, the duration of vowel segments, i.e. the nucleus of Japanese mora, was significantly invariable in the control group regardless of differences between subjects as well as in segments compared with the ataxic group. In addition, the duration of phonemically long Japanese vowels was significantly shorter in the ataxic group. CONCLUSIONS & IMPLICATIONS: The results indicate that the perceptual impression of 'scanning' in Japanese ataxic cases derives mainly from the breakdown of isochrony in terms of difficulty in keeping the length of vowel segments of Japanese invariable during speech production. In addition, the tendency toward irregular shortening of the length of phonemically long Japanese vowels is thought to reinforce the impression of 'scanning' in ataxic speech in Japanese cases.

Intonation contrast in Cantonese speakers with hypokinetic dysarthria associated with Parkinson's disease.

PURPOSE: Speech produced by individuals with hypokinetic dysarthria associated with Parkinson's disease (PD) is characterized by a number of features including impaired speech prosody. The purpose of this study was to investigate intonation contrasts produced by this group of speakers. METHOD: Speech materials with a question-statement contrast were collected from 14 Cantonese speakers with PD. Twenty listeners then classified the productions as either questions or statements. Acoustic analyses of F0, duration, and intensity were conducted to determine which acoustic cues distinguished the production of questions from statements, and which cues appeared to be exploited by listeners in identifying intonational contrasts. RESULTS: The results show that listeners identified statements with a high degree of accuracy, but the accuracy of question identification ranged from 0.56% to 96% across the 14 speakers. The speakers with PD used similar acoustic cues as nondysarthric Cantonese speakers to mark the question-statement contrast, although the contrasts were not observed in all speakers. Listeners mainly used F0 cues at the final syllable for intonation identification. CONCLUSION: These data contribute to the researchers' understanding of intonation marking in speakers with PD, with specific application to the production and perception of intonation in a lexical tone language.