Prosodic prominence in parkinsonian speech : a dynamical approach

Thies, Tabea and Mücke, Doris and Auris, Bastian and Lowit, Anja and Barbe, Michael T. (2018) Prosodic prominence in parkinsonian speech : a dynamical approach. In: LabPhon16, 2018-06-19 - 2018-06-23, University of Lisbon.

[img]
Preview
Text (Thies-etal-LabPhon-2018-Prosodic-prominence-in-parkinsonian-speech-Abstract)
Thies_etal_LabPhon_2018_Prosodic_prominence_in_parkinsonian_speech_Abstract.pdf
Final Published Version

Download (338kB)| Preview
    [img]
    Preview
    Text (Thies-etal-LabPhon-2018-Prosodic-prominence-in-parkinsonian-speech)
    Thies_etal_LabPhon_2018_Prosodic_prominence_in_parkinsonian_speech.pdf
    Final Published Version

    Download (983kB)| Preview

      Abstract

      In the present study, we investigate dynamic changes in prosodic highlighting strategies in PD patients and compare them to the productions of neurotypical speakers. Therefore, we compare the production of target words in divergent focus structure, contrastive focus and background. We recorded 40 German speakers: 20 patients with idiopathic Parkinson in medication ON condition, 14 males and 6 females, aged between 54 - 80 and 20 healthy aged and gender matched controls. All speakers were classified in terms of duration of the disorder, severity of the disorder, motoric activity level (UPDRS III, [6]), level of cognition and speech problems in terms of dysarthria. As speech material, we used a question-answer scenario presented on a computer screen to manipulate focal structure by means of contextualizing contexts. Nine target words were placed in either contrastive focus or background position in sentences such as <Die Fliege hat die grüne NAse berührt.> (“The fly has touched the green nose.”) related to pictures on a computer screen [cf. fig.1]. Target words were always disyllabic (CV.CV structure), containing one of the three long vowels /i:/, /a:/ or /o:/, in the stressed syllable, such as <Nase> /na:z@/. In total, we recorded 1440 tokens (9 target words x 40 speakers x 2 focus structures x 2 adjectives). For acoustic measurements, we analyzed the voice range, syllable duration, formant means and contours, the mean intensity and the F0 contours in terms of pitch height and tonal onglide [7]. For the upcoming conference, the results of all intonational and articulatory variables will be presented. We will discuss how much variation is tolerated in a dynamical system of prosodic prominence before the expression of prosodic functions is getting instable.