When a bilingual switches languages, do they switch their "voice"? Using a new conversational corpus of speech from early Cantonese-English bilinguals (N = 34), this paper examines the talker-specific acoustic signature of bilingual voices. Following prior work in voice quality variation, 24 filter and source-based acoustic measurements are estimated. The analysis summarizes mean differences for these dimensions, in addition to identifying the underlying structure of each talker's voice across languages with principal components analyses. Canonical redundancy analyses demonstrate that while talkers vary in the degree to which they have the same "voice" across languages, all talkers show strong similarity with themselves.

A recent model of sound change posits that the direction of change is determined, at least in part, by the distribution of variation within speech communities (Harrington, Kleber, Reubold, Schiel, & Stevens, 2018; Harrington & Schiel, 2017). We explore this model in the context of bilingual speech, asking whether the less variable language constrains phonetic variation in the more variable language, using a corpus of spontaneous speech from early Cantonese-English bilinguals (Johnson, Babel, Fong, & Yiu, 2020). As predicted, given the phonetic distributions of stop obstruents in Cantonese compared to English, intervocalic English /b d g/ were produced with less voicing for Cantonese-English bilinguals and word-final English /t k/ were more likely to be unreleased compared to spontaneous speech from two monolingual English control corpora (Pitt, Johnson, Hume, Kiesling, & Raymond, 2005; Swan, 2016). Cantonese phonology is more gradient in terms of voicing initial obstruents (Clumeck, Barton, Macken, & Huntington, 1981; W. Y. P. Wong, 2006) than permitting releases of final obstruents, which is categorically prohibited Bauer & Benedict (2011); Khouw & Ciocca (2006). Neither Cantonese-English bilingual initial voicing nor word-final stop release patterns were significantly impacted by language mode. These results provide evidence that the phonetic variation in crosslinguistically linked categories in bilingual speech is shaped by the distribution of phonetic variation within each language, thus suggesting a mechanistic account for why some segments are more susceptible to cross-language influence than others in studies of mutual influence.

In Cantonese and several other Chinese languages, /n/ is merging with /l/. The Cantonese merger appears categorical, with /n/ becoming /l/ word-initially. This project aims to describe the status of /n/ and /l/ in bilingual Cantonese and English speech to better understand individual differences at the interface of crosslinguistic influence and sound change. We examine bilingual speech using the SpiCE corpus, composed of speech from 34 early Cantonese-English bilinguals. Acoustic measures were collected on pre-vocalic nasal and lateral onsets in both languages. If bilinguals maintain separate representations for corresponding segments across languages, smaller differences between /n/ and /l/ are predicted in Cantonese compared to English. Measures of mid-frequency spectral tilt suggest that the /n/ and /l/ contrast is robustly maintained in English, but not Cantonese. The spacing of F2-F1 suggests small differences between Cantonese /n/ and /l/, and robust differences in English. While cross-language categories appear independent, substantial individual differences exist in the data. These data contribute to the understanding of the /n/ and /l/ merger in Cantonese and other Chinese languages, in addition to providing empirical and theoretical insights into crosslinguistic influence in early bilinguals.

This study examines whether social evaluation and intelligibility affect judgments of perceived credulity. Canadian English listeners (i) completed a speech-in-noise task, (ii) judged if sentences were true, and (iii) rated voices on social dimensions. A Bayesian analysis showed that neither social evaluation nor intelligibility substantially influenced perceived truthfulness. While recent work suggests talker accent may impact perceived credulity, the question is far from settled in the literature. We conclude that perceived credulity is negligibly affected by assessments of natively-accented voices.

While crosslinguistic influence is widespread in bilingual speech production, it is less clear which aspects of representation are shared across languages, if any. Most prior work examines phonetically distinct yet phonologically similar sounds, for which phonetic convergence suggests a cross-language link within individuals [1]. Convergence is harder to assess when sounds are already similar, as with English and Cantonese initial long-lag stops. Here, the articulatory uniformity framework [2, 3, 4] is leveraged to assess whether bilinguals share an underlying laryngeal feature across languages, and describe the nature of cross-language links. Using the SpiCE corpus of spontaneous Cantonese-English bilingual speech [5], this paper asks whether Cantonese-English bilinguals exhibit uniform voice-onset time for long-lag stops within and across languages. Results indicate moderate patterns of uniformity within-language—replicating prior work [2, 6]—and weaker patterns across languages. The analysis, however, raises many questions, as correlations were generally lower compared to prior work, and talkers did not adhere to expected ordinal VOT relationships by place of articulation. Talkers also retained clear differences for /t/ and /k/, despite expectations of similarity. Yet at the same time, more of the overall variation seems to derive from individual-specific differences. While many questions remain, the uniformity framework shows promise.

Listeners need to accommodate pronunciations that vary widely. Lexically-guided perceptual adaptation has been well documented in the literature, but relatively little is known about its limits. Moreover, there are at least two plausible mechanisms supporting adaptation for sound categories: targeted shifts towards the novel pronunciation or a general relaxation of criteria. This paper examines a limit of perceptual adaptation — asymmetries in adaptation to the voicing and devoicing of coronal fricatives in English — and suggests that typology or synchronic experience affects how listeners adapt. A corpus study of coronal fricative substitution patterns confirmed that North American English listeners are more likely to be exposed to devoiced /z/ than voiced /s/. Across two perceptual adaptation experiments, listeners in test conditions head naturally produced devoiced /z/ or voiced /s/ in critical items within sentences, while control listeners were exposed to identical sentences with canonical fricative pronunciations. Perceptual adaptation was tested via word endorsement rates in a lexical decision test, with devoiced /z/ or voiced /s/, as well as a novel alveopalatalized pronunciation, to determine whether adaptation was targeted in the direction of the exposed variant or reflected more general relaxation. Results indicate there was targeted and word-specific adaptation for /z/ devoicing. Conversely, there was some evidence of /s/ voicing eliciting a more general category relaxation. These results underscore the role of prior perceptual experiences, and suggest a need for an evaluation stage in lexically-guided perceptual learning, where listeners assess whether a heard item merits an update to the representation.