Journal of Second Language Pronunciation - Volume 1, Issue 2, 2015

When compared to native-speaker language ‘targets’, second language (L2) speakers are known to exhibit within-speaker variation in their L2 language performance according to the identity of their interlocutor (Beebe & Zuengler, 1983), the topic that is being discussed (Dolgova Jacobsen, 2008), or both (Rampton, 2011). However, previous studies have rarely applied the same methodology to different first language (L1) groups and have rarely used data from a range of speakers, so they have been unable to explore differences between speaker groups. This paper examines situational style-shifting in L2 speakers of New Zealand English in two groups of speakers, one with Korean as their L1 and the other with German as their L1. Participants were recorded in three different settings: discussing family, discussing work, and in an authentic service encounter. Male and female speakers showed similar patterns, and all were found to exhibit stylistic variation in their production of English vowels such that tokens were most nativelike in service encounters, followed by discussion of work and then family. This pattern was more robust for the Korean L1 speakers than for the German L1 speakers, which is explained by different social and linguistic histories of the two groups in New Zealand. This paper adds to our current understanding of sociolinguistic variation in L2 speakers by applying audience design (Bell, 1984) and identity construction (Eckert, 2000) accounts to L2 speakers’ production of vowels.

This study investigates how speakers who speak Brazilian Portuguese as their first language and English as their second language perceive the English phonemes /h/ and /ɹ/, and how they and monolingual Brazilian Portuguese speakers map these phonemes onto Portuguese sound categories. Participants took part in three experiments: an AXB discrimination test, an identification test, and a cross-language assimilation test, which was also taken by monolinguals. Lower and higher proficiency groups were able to hear the distinction acoustically, but only the higher proficiency group used the distinction to identify English words. Monolingual Brazilian Portuguese speakers and the higher proficiency group assimilated English /h/ primarily to Portuguese /h/. However, the phonological environment had an effect for monolinguals, but not for the higher proficiency group. The lower proficiency group, which one might expect to fall in between these two groups, showed a failure to assimilate English sounds to the Portuguese categories.

This study assesses the claim that English late learners of Spanish do not perceive stress like native Spanish speakers, and that a short targeted stress perception training intervention during a study abroad Spanish language course has clear positive effects on stress perception. Fifteen English speakers were exposed to 90 hours of Spanish lessons during a three–week study abroad experience in Mar del Plata, Argentina. The trained group (N = 8) received 10 minutes of perceptual training on vowel and stress contrasts with nonce words three days a week, while the L1 English control group (N = 7) received communicative training focused on consonants, and the native Spanish control group (N = 7) received no training. Participants’ perception was assessed at pretest and posttest, both consisting of identification tasks with nonce words. Results indicated that all English speakers experienced difficulties in perceiving Spanish stress when compared to native Spanish speakers in the pretest. At posttest, however, the English trained group performed comparably to the native Spanish group and differed significantly from the control group, indicating an effect of training on the perception of L2 stress. The results show that English speakers evidenced perceptual difficulties when learning Spanish stress, which could be overcome with a small dose of targeted training with nonce words. Even though L2 immersion in a study abroad context was beneficial for the acquisition of Spanish stress, only students receiving stress training performed like native speakers.

The intelligibility, comprehensibility and foreign accentedness of native and Polish-accented English sentences were evaluated by six Polish, six Spanish and six English speakers. The nonnative data were also analyzed for segmental and word stress errors. Results indicated that the three measures were partially independent of one another, supporting earlier findings that accented speech can be intelligible and comprehensible. An interlanguage speech intelligibility detriment was observed for Spanish listeners, but no clear evidence of an interlanguage speech intelligibility benefit was found, as nonnative listeners never outperformed native listeners. The number of segmental errors, rather than lexical stress errors, was found to correlate with comprehensibility and accentedness ratings of nonnative speech, but not with intelligibility scores. In general, the results point to a greater effect of stimulus properties than of listeners’ L1s in the perception of nonnative speech.

This paper reports results on the acquisition of the English /p/–/b/ contrast by native speakers of Arabic. This contrast does not exist in the participants’ native language (NL). The central finding of this study is that some of the research participants exhibited a covert contrast between these segments in their interlanguage productions. That is, two of the five Arabic-speaking participants who were transcribed as having no contrast between [p] and [b] did, in fact, produce a statistically reliable distinction in voice onset time lags between the two target segments. The existence of such an intermediate stage of covert contrast in the learning of L2 phonology is eminently plausible, in view of the progressive nature of phonological acquisition. Our results help bring the learning of second-language contrasts into conformity with findings of the same phenomenon in the areas of L1 acquisition and phonologically disordered speech.