Naslov
Lingual coarticulation index for epg: voiced vs. voiceless

Autori
Liker, Marko ; Gibbon, Fiona E.

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
EPG2008 / - , 2008

Skup
Fifth international electropalatography symposium (EPG2008)

Mjesto i datum
Edinburgh, Ujedinjeno Kraljevstvo, 01.09.2008. - 02.09.2008

Vrsta sudjelovanja
Poster

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
lingual coarticulation; lingual coarticulation index; electropalatography (EPG); articulation

Sažetak
Introduction: The problem of coarticulatory differences between voiced and voiceless consonants is relatively controversial, especially when electropalatographic (EPG) data are taken into consideration. The aim of this paper is to present a new index for measuring coarticulation differences between voiced and voiceless consonants using electropalatography. Method: Speech material consisted of simultaneous EPG and acoustic data for nonsense symmetrical VCV sequences (C = s, z, t, d ; V = i, a, u). Eight speakers were recorded: 4 Croatian (2 female and 2 male) and 4 English speakers (2 female and 2 male). Croatian data was recorded in a silent room at the Department of Phonetics in Zagreb. English data was extracted from the EUR-ACCOR database (Marchal & Hardcastle, 1993). Segmentation and annotation were performed using the Articulate Assistant software and Praat. The index was developed and analysis was performed by MS Excel. Results: The Difference area index (DAI) provides a detailed 3D visualization and a single relative numerical value as a measure of coarticulaticulatory effects. The dynamics of coarticulatory characteristics is also captured by the DAI. Preliminary results for 2 speakers (1 Croatian and 1 English) show that stops coarticulate more than fricatives and that voiced consonants generally coarticulate more than voiceless. They also show that coarticulatory effects in fricatives are distributed over larger areas of the palate, while in stops they are more localized. Voiceless stops and consonants coarticulate more at the beginning and the end while in their voiced counterparts coarticulation is more consistent throughout the hold phase. Acoustic analysis of voicing duration concurs with EPG results. Additional speakers are being analysed. Conclusions: The results obtained by the DAI agree with some already available data on coarticulation in voiced and voiceless, but they also provide some new data on coartculatory dynamics conditioned by voicing. References. Marchal, A., & Hardcastle, W. J. (1993). ACCOR: Instrumentation and database for the cross-language study of coarticulation. Language and Speech, 36, 137 - 153.

Izvorni jezik
Engleski

Znanstvena područja
Kliničke medicinske znanosti, Filologija

POVEZANOST RADA