Charles B. Chang

A novelty effect in phonetic drift of the native language

Abstract Previous findings on adult second-language (L2) learners showed systematic phonetic changes in their production of the native language (L1) starting in the first weeks of L2 learning [Chang, C. B. (2012). Rapid and multifaceted effects of second-language learning on first-language speech production. Journal of Phonetics, 40 , 249–268]. This “phonetic drift” of L1 production in novice L2 learners was consistent with reports of phonetic drift in advanced L2 learners; however, the fact that novice learners showed relatively pronounced drift was unexpected. To explore the hypothesis that this pattern is due to a novelty effect boosting the encoding and retrieval of elementary L2 experience, the current study compared the inexperienced learners analyzed previously (learners with no prior knowledge of the L2) to experienced learners enrolled in the same language program. In accordance with the hypothesis, experienced learners manifested less phonetic drift in their production of L1 stops and vowels than inexperienced learners, suggesting that progressive familiarization with an L2 leads to reduced phonetic drift at later stages of L2 experience. These findings contradict the assumption that L2 influence on the L1 is weakest at early stages of L2 learning and argue in favor of viewing the L1 and L2 both as dynamic systems undergoing continuous change.

Production of phonetic and phonological contrast by heritage speakers of Mandarin

This study tested the hypothesis that heritage speakers of a minority language, due to their childhood experience with two languages, would outperform late learners in producing contrast: language-internal phonological contrast, as well as cross-linguistic phonetic contrast between similar, yet acoustically distinct, categories of different languages. To this end, production of Mandarin and English by heritage speakers of Mandarin was compared to that of native Mandarin speakers and native American English-speaking late learners of Mandarin in three experiments. In experiment 1, back vowels in Mandarin and English were produced distinctly by all groups, but the greatest separation between similar vowels was achieved by heritage speakers. In experiment 2, Mandarin aspirated and English voiceless plosives were produced distinctly by native Mandarin speakers and heritage speakers, who both put more distance between them than late learners. In experiment 3, the Mandarin retroflex and English palato-alveolar fricatives were distinguished by more heritage speakers and late learners than native Mandarin speakers. Thus, overall the hypothesis was supported: across experiments, heritage speakers were found to be the most successful at simultaneously maintaining language-internal and cross-linguistic contrasts, a result that may stem from a close approximation of phonetic norms that occurs during early exposure to both languages.

Evidence for language transfer leading to a perceptual advantage for non-native listeners

Phonological transfer from the native language is a common problem for non-native speakers that has repeatedly been shown to result in perceptual deficits vis-à-vis native speakers. It was hypothesized, however, that transfer could help, rather than hurt, if it resulted in a beneficial bias. Due to differences in pronunciation norms between Korean and English, Koreans in the U.S. were predicted to be better than Americans at perceiving unreleased stops-not only in their native language (Korean) but also in their non-native language (English). In three experiments, Koreans were found to be significantly more accurate than Americans at identifying unreleased stops in Korean, at identifying unreleased stops in English, and at discriminating between the presence and absence of an unreleased stop in English. Taken together, these results suggest that cross-linguistic transfer is capable of boosting speech perception by non-natives beyond native levels.

Bilingual Perceptual Benefits of Experience with a Heritage Language

Research on the linguistic knowledge of heritage speakers has been concerned primarily with the advantages conferred by heritage language experience in production, perception, and (re)learning of the heritage language. Meanwhile, second-language speech research has begun to investigate potential benefits of first-language transfer in second-language performance. Bridging these two bodies of work, the current study examined the perceptual benefits of heritage language experience for heritage speakers of Korean in both the heritage language (Korean) and the dominant language (American English). It was hypothesized that, due to their early bilingual experience and the different nature of unreleased stops in Korean and American English, heritage speakers of Korean would show not only native-like perception of Korean unreleased stops, but also better-than-native perception of American English unreleased stops. Results of three perception experiments were consistent with this hypothesis, suggesting that benefits of early heritage language experience can extend well beyond the heritage language.

Context effects on second-language learning of tonal contrasts.

Studies of lexical tone  learning generally focus on monosyllabic contexts, while reports of phonetic learning benefits associated with input variability are based largely on experienced learners. This study trained inexperienced learners on Mandarin tonal contrasts to test two hypotheses regarding the influence of context and variability on tone  learning. The first hypothesis was that increased phonetic variability of tones in disyllabic contexts makes initial tone  learning more challenging in disyllabic than monosyllabic words. The second hypothesis was that the learnability of a given tone varies across contexts due to differences in tonal variability. Results of a word learning experiment supported both hypotheses: tones were acquired less successfully in disyllables than in monosyllables, and the relative difficulty of disyllables was closely related to contextual tonal variability. These results indicate limited relevance of monosyllable-based data on Mandarin learning for the disyllabic majority of the Mandarin lexicon. Furthermore, in the short term, variability can diminish learning; its effects are not necessarily beneficial but dependent on acquisition stage and other learner characteristics. These findings thus highlight the importance of considering contextual variability and the interaction between variability and type of learner in the design, interpretation, and application of research on phonetic learning.

Transfer effects in perception of a familiar and unfamiliar language

Second-language (L2) speech perception is typically worse than first-language (L1) perception, a disparity often attributed to negative transfer (interference) from the L1 of L2 listeners. The current study investigated the hypothesis that L1 transfer is not always negative, but variable depending on the nature of L1 perceptual biases. In Experiment 1, four groups of L2 English speakers whose L1s (Japanese, Korean, Mandarin, and Russian) differ in the relative informativeness of vowel-to-consonant transition cues were tested on their perception of English segments that rely crucially on these cues: final unreleased voiceless stops. In comparison to L1 English listeners, L1 Japanese, Russian, and Mandarin listeners performed significantly worse, whereas L1 Korean listeners performed significantly better. In Experiment 2, when the same groups were tested on similar Korean stimuli, L1 Russian listeners outperformed all other groups except the Korean group. These results provide evidence that L1 transfer effe...

Acoustic measures of sound change in language obsolescence

Following recent acoustic investigations of obsolescing languages such as Babel (2007), I present an acoustic study of phonetic change in Southeastern Pomo (Northern Hokan, Pomoan) based upon recordings of two generations of speakers. This paper focuses on the realization of two phonological contrasts: A velar/post‐velar contrast and a dental/alveolar contrast. First, acoustic measurements show that both contrasts have undergone phonetic change convergent with English, the dominant contact language, which contains neither contrast. The contrast between velars and post‐velars is considerably diminished in Generation 2 with respect to Generation 1, as seen in the significantly smaller difference between the second formant resonances of the velar and post‐velar fricatives. The dental/alveolar contrast undergoes a similar change: The difference between the burst peak frequencies of the dental and alveolar ejectives and the difference between the third formant resonances of the following vowels both shrink. Ho...

LEXTALE_CH: A quick, character-based proficiency test for Mandarin Chinese

Given the growing interest within linguistics in how bilinguals, and second language (L2) learners more generally, process their L2 (e.g., Talamas et al., 1999; Prior et al., 2007), there has long been a need among language researchers for reliable and valid measures of L2 proficiency. Much of the research on L2 processing within experimental psychology (e.g., Arêas Da Luz Fontes & Schwartz, 2010; Leonard et al., 2010) has relied on questionnaire tools to provide the primary, or sometimes sole, source of data on L2 proficiency. Such questionnaires typically measure proficiency by asking participants for a subjective self-assessment of their own language ability, operationalized in terms of an overall rating and/or ratings by skill (i.e., listening, speaking, reading, writing). However, there are a number of problems with subjective assessments that limit their utility in experimental research. For one, even within one well-defined population (e.g., adult learners of the same L1 background acquiring the same L2 under the same learning conditions), learners at the same proficiency level may perceive their language abilities differently, leading to variation in self-assessments that does not reflect actual proficiency differences. Different studies also use different rating scales, such that proficiency measures are often not directly comparable across studies. Furthermore, the evidence for the external validity of self-assessments as measures of proficiency is relatively weak (Lemmon & Goggin, 1989; Delgado et al., 1999). Thus, despite the fact that questionnaire-based methods of proficiency measurement are relatively simple and easy to administer, there has continued to be a need for reliable and valid methods of measuring proficiency—in particular, methods which can evaluate L2 proficiency as quickly and easily as a questionnaire. In the case of L2 English, Lemhöfer and Broersma (2012) proposed an objective alternative to subjective self-assessments, a vocabularybased test called LexTALE (Lexical Test for Advanced Learners of English). LexTALE was designed as a standard and efficient tool for evaluating the English proficiency of L2 learners within a short amount of time (5–10 minutes). The test requires participants to identify the lexical status (i.e., real

Perceptual attention as the locus of transfer to nonnative speech perception

Abstract One’s native language (L1) is known to influence the development of a nonnative language (L2) at multiple levels, but the nature of L1 transfer to L2 perception remains unclear. This study explored the hypothesis that transfer effects in perception come from L1-specific processing strategies, which direct attention to phonetic cues according to their estimated relative functional load (RFL). Using target languages that were either familiar (English) or unfamiliar (Korean), perception of unreleased final stops was tested in L1 English listeners and four groups of L2 English learners whose L1s differ in stop phonotactics and the estimated RFL of a crucial cue to unreleased stops (i.e., vowel-to-consonant formant transitions). Results were, overall, consistent with the hypothesis, with L1 Japanese listeners showing the poorest perception, followed by L1 Mandarin, Russian, English, and Korean listeners. Two exceptions occurred with Russian listeners, who underperformed Mandarin listeners in identification of English stops and outperformed English listeners in identification of Korean stops. Taken together, these findings support a cue-centric view of transfer based on perceptual attention over a direct phonotactic view based on structural conformity. However, transfer interacts with prior L2 knowledge, which may result in significantly different perceptual consequences for a familiar and an unfamiliar L2.

Effects of age, sex, context, and lexicality on hyperarticulation of Korean fricatives

Seoul Korean is known for a rare three-way laryngeal contrast among lenis, fortis, and aspirated voiceless stops, which has recently undergone a change in phonetic implementation: whereas older speakers rely more on voice onset time (VOT) to distinguish lenis and aspirated stops, younger speakers rely more on onset fundamental frequency (f 0) in the following vowel. This production difference is reflected in disparate strategies for enhancing the contrast in clear speech, supporting the view that younger and older speakers represent the three laryngeal categories differently in terms of VOT and f 0 targets (Kang & Guion, 2008). In the current study, we used the clear speech paradigm to test for change in the representation of the two-way contrast between fortis (/s*/) and non-fortis (/s/) fricatives. Native Seoul Korean speakers (n = 32), representing two generations and both sexes, were recorded producing the coronal stops and fricatives in different vowel contexts, item types (real vs. nonce words), and...