Journal Title

Journal ISSN

Volume Title



This dissertation investigates Voice Onset Time (VOT), which serves as an essential property for differentiating plosive consonants in L1 and L2 Mandarin Chinese. It surveys VOT variations and demonstrates that they are affected by several phonetic and phonological properties, e.g., lexical tone, place of articulation (POA), speech rate, phrase-position, pitch register, and gender. While researchers disagree on the relationship between lexical tone and VOT, this study investigates this relationship explicitly. Moreover, while it has been suggested that VOT varies because of different lexical tones, the question has not been comprehensively explored as to which of the tonal properties are responsible for this effect on VOT. This dissertation also tests whether non-native Mandarin speakers exhibit similar effects. Four experiments were conducted—two for native Taiwan Mandarin speakers and a parallel set of two for L2 Mandarin speakers. Two experiments elicited stop-initial words produced with one open-unrounded vowel, three articulation places, four lexical tones, three different speech rates, and three utterance positions. The other two experiments elicited the same set of stimuli and conditions in one tone with three pitch-levels and at a natural speech rate. A series of linear mixed-effects regression models were employed to model the effects of the properties mentioned earlier on VOT duration. We wanted to know whether these factors affect VOT in native and in non-native speech.Testing 164 participants (68 Taiwanese, 34 Spanish, 40 Japanese, and 22 English speakers of Mandarin), the results reveal that when other factors were kept constant, tone indeed influenced VOT, and the higher the onset tone pitch, the shorter the VOT. POA and the speech rate were also found to be highly significant factors. The results reveal that all non-native groups showed the same effects regardless of their L1. This finding suggests that the tone effect on VOT in Mandarin is a universal tendency due to the physiology of the vocal tract rather than due to language-specific phonology. However, the Spanish and Japanese groups showed extended VOT values, which were not from their native VOTs, but their Mandarin VOTs were still significantly shorter than the native Mandarin speakers. Thus, the significant VOT differences between groups indicate some degree of L1 influence, which suggests that L2 VOT delay is probably mediated by language-specific phonological grammar. This dissertation provides empirical evidence that an acoustic property, such as VOT, is not an isolated phenomenon but is involved with other complex phonological categories such as lexical tone. It also discusses how the effects operate within phonetic and phonological theories. Additionally, it compares Mandarin learners’ L1 and L2 VOT directly. This cross-linguistic survey offers insight for L2 performance variations regarding phonetics, which may provide Mandarin instructors with multi-dimensional comparisons and confirmation of the interlanguage process as relevant to Second Language Acquisition. The observed phenomena may offer L2 classrooms insights into Mandarin accent variations for L2 English, Japanese, and Spanish learners.