Impact of Voice Onset Time on Phoneme Perception and Implications for a Language-Agnostic Hearing Assessment

Abstract

Purpose: The aim of this study was to evaluate the effect of native language on phoneme perception in normal-hearing (NH) adults representing the three most commonly spoken languages in the United States (English, Spanish, and Mandarin Chinese). Method: A prospective cross-sectional study was performed at a tertiary care institution. Ninety-five NH young adults completed Experiment 1: 36 native English, 32 native Mandarin, and 27 native Spanish speakers, with an average age of 21.5 years, recruited from a university population. A subset of 30 NH participants drawn from the Experiment 1 cohort (10 per language group) completed Experiment 2. In Experiment 1, a two-alternative forced-choice (2AFC) same–different paradigm was developed using paired stimuli drawn from an 11-step continuum spanning English /d/−/t/ voice onset time (VOT). Pairs were categorized as subtle, intermediate, or pronounced based on degree of VOT contrast. Experiment 2 utilized a 3AFC odd-one-out paradigm with 16 tasks, with an expanded stimulus library. The stimuli were consonant–vowel (CV) and VCV phrases that were inclusive of voicing contrast but also place, manner, and vowel contrasts. Results: Overall performance in Experiment 1 did not differ significantly across language groups (χ 2 = 4.81, p = .090). However, when stratified by degree of VOT contrast, a significant group effect emerged for intermediate VOT pairs (χ 2 = 7.59, p = .023), driven by differences between Mandarin and Spanish speakers (χ 2 = 7.59, p = .0059, Bonferroni corrected). No significant differences were observed between English speakers and either of the other groups. In Experiment 2, all three language groups achieved 100% accuracy across phoneme pairs spanning multiple contrast types, with no language-related differences detected. Conclusions: While native language affects discrimination of intermediate VOT contrasts in NH adults, performance converges when using a broader set of stimuli. These results provide preliminary support for discrimination-based approaches to speech perception assessment and highlight the importance of stimulus selection and task design in minimizing language-related bias. Further studies are required to evaluate sensitivity and clinical applicability in more challenging listening environments and with hard of hearing populations. Supplemental Material: https://doi.org/10.23641/asha.33470986