Voice

About: Voice is a research topic. Over the lifetime, 2393 publications have been published within this topic receiving 56637 citations.

Contextual Variation in L2 Spanish: Voicing Assimilation in Advanced Learner Speech

Abstract: Abstract The present study examines whether, and to what degree, regressive voicing assimilation of Spanish /s/ (as in rasgo/rasgo/ [ˈraz.ɣ̞ o]) occurs in the speech of advanced second language (L2) learners of Spanish. Acoustic analyses of L2 productions of /s/ in the voicing context (preceding a voiced consonant) and in the non-voicing context (preceding a voiceless consonant) elicited from a contextualized picture-description task revealed a contextual voicing effect in the speech of only a limited number of the advanced L2 speakers. The low occurrence of the assimilation process even amongst the advanced learners may be attributed in part to the variable nature of voicing in the input and to the complexity of the process (i.e. subject to different stylistic, linguistic, and social factors). The study also provides a phonetic description of the variants of L2 Spanish /s/ and finds that when voicing does occur, it is phonetically similar to native Spanish voicing in terms of the phonetic contexts in which voicing occurs, patterns of durational differences of /s/ according to voicing, and the variable nature of its occurrence.

The Effect of Instructed Second Language Learning on the Acoustic Properties of First Language Speech

Abstract: This paper reports on a comprehensive phonetic study of American classroom learners of Russian, investigating the influence of the second language (L2) on the first language (L1). Russian and English productions of 20 learners were compared to 18 English monolingual controls focusing on the acoustics of word-initial and word-final voicing. The results demonstrate that learners’ Russian was acoustically different from their English, with shorter voice onset times (VOTs) in [−voice] stops, longer prevoicing in [+voice] stops, more [−voice] stops with short lag VOTs and more [+voice] stops with prevoicing, indicating a degree of successful L2 pronunciation learning. Crucially, learners also demonstrated an L1 phonetic change compared to monolingual English speakers. Specifically, the VOT of learners’ initial English voiceless stops was shortened, indicating assimilation with Russian, while the frequency of prevoicing in learners’ English was decreased, indicating dissimilation with Russian. Word-final, the duration of preceding vowels, stop closures, frication, and voicing during consonantal constriction all demonstrated drift towards Russian norms of word-final voicing neutralization. The study confirms that L2-driven phonetic changes in L1 are possible even in L1-immersed classroom language learners, challenging the role of reduced L1 use and highlighting the plasticity of the L1 phonetic system.

Tone Splits and Voicing Shifts in Thai: Phonetic Plausibility*

Abstract: It has been shown for a number of languages that right after the release of a voiced stop consonant the fundamental frequency (FO) of the voice is likely to be lower than after the release of a voiceless stop and that such FO perturbations can influence phonemic judgments of voicing. This led to the designing of two experiments to test the phonetic plausibility of the argument: (1) CV syllables were synthesized with three values of voice onset time (VOT) acceptable as Thai Ib p phi. Each of these was combined with a continuum of FO contours that had previously been divided perceptually into the high, mid and low tones. These syllables were played to native speakers of Thai for tonal identification. (2) Labial stops with nine values of VOT separable into Ib p phi categories were coupled on synthetic mid-tone and low-tone CV syllables with upward and downward FO onsets varying in extent and duration. The resulting syllables were played for iden­ tification of the initial consonants. The historical argument receives modest support, especially from the second experiment, suggesting that during a period of tone splitting, under the influence of audible FO perturbations, speaken could have brought about the rephonemicization of the old consonant categories. Thus, these results give direct support to the argument that pitch factors led to voicing shifts but only indirect support to the claim that they gave rise to tone splits.

A “Self-Voicing” Test for Individuals with Visual Impairments:

Abstract: A \"Self-Voicing\" Test for Individuals with Visual Impairments Eric G. Hansen, Moon J. Lee, Douglas C. Forer For test takers who are visually impaired (that is, are blind or have low vision), using human readers during tests can have several disadvantages. The problems may include an inconsistent quality of reading, the test taker's anxiety and embarrassment at having the reader reread the material, the reader's mistakes in recording answers, fatigue caused by the slowness and intensity of the reader/test-taker interaction, and a greater need for extra testing time. A computer-based testing system that is operable by keyboard input and speech output (synthesized and/or prerecorded) may reduce or eliminate the need for a human reader for some test takers who are visually impaired. This study examined the use of a prototype testing system that employs synthesized speech to deliver questions on reading and listening comprehension tests. The system is termed a \"self-voicing\" test because it provides the speech output capability within a testing application itself, rather through the use of a distinct assistive technology, such as screen reader software. With funding from the Educational Testing Service (ETS), the Graduate Record Examinations (GRE) program, and the Test of English as a Foreign Language (TOEFL) program, researchers investigated the use of speech output technology for tests for individuals with visual impairments.