Effect of speaking style and speaker on the estimation of the minimum sample length for speaking rate

• Grant number: 19/01661-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: May 01, 2019
• End date: December 31, 2019
• Field of knowledge: Linguistics, Literature and Arts - Linguistics - Linguistic Theory and Analysis
• Principal Investigator: Pablo Arantes
• Grantee: Verônica Gomes Lima
• Host Institution: Centro de Educação e Ciências Humanas (CECH). Universidade Federal de São Carlos (UFSCAR). São Carlos , SP, Brazil

Abstract

The present project builds on a previous research project by the proponents, in which they developed a methodology to help determine the minimum sample length required to derive estimates of speaking rate that are stable enough to be considered representative of the speaker's long term behavior. The sample length defines what we called the speaking rate stabilization time. In the previous project our method was compared to the only methodology that has been reported in the literature. We then compared different types of rates (articulation and speech rates), different linguistic units used to determine the rate (phones, syllables, vowel onset-to-vowel onset (VV) units and phonological words) and rate levels (slow, normal and fast). The speech material studied was read speech by Brazilian Portuguese (BP) speakers. Results show that both methods yield rather similar stabilization times, but much shorter ones compared to the ones reported on the only study already published on the subject. With the present project we aim to investigate the effects of speaking style and speakers on stabilization times of articulation rate. The speech material will be semispontaneous and read speech, recorded from ten BP speakers. Five estimates of stabilization time will be derived for each speaker in each style in order to gauge within- and between-speaker variability and determine whether stabilization time is an speaker-specific trait or can be generalized to baggers speaker samples. Results yielded by the project will have theoretical, methodological and practical implications, specially in the fields of speech corpora design and forensic phonetics.