Language dominance and order of acquisition affect auditory translation priming in heritage speakers

Abstract

Late second language (L2) learners show translation priming from the first language (L1) to the second language (L1-L2), while L2-L1 effects are inconsistent. Late L2 learners also acquire the L2 after the L1 and are typically less dominant in the L2. As such, the relative contribution of language dominance and order of acquisition is confounded in these results. Here, Cantonese heritage and native speakers are tested in an auditory translation priming paradigm. As heritage speakers first learn Cantonese (L1) but later become dominant in English (L2), this profile allows for the potential dissociation of dominance and order of acquisition in translation priming. If order of acquisition is the primary factor, stronger priming is expected in the L1-L2 (Cantonese-English) direction; however, if dominance plays a stronger role, priming is expected in the L2-L1 (English-Cantonese) direction. Native speakers showed stronger L1-L2 priming, consistent with previous findings, while heritage speakers showed priming in both directions, and marginally larger L2-L1 priming. Treating language dominance as a continuous variable revealed that L1-L2 priming correlated with increased Cantonese dominance, while L2-L1 priming marginally correlated with increased English dominance. Collectively, these results suggest that both language dominance and order of acquisition help explain translation priming findings and bilingual lexical processing, generally. Overall, they invite a rethinking of the role of both variables in bilingual lexical access for speakers with different language dominance profiles.

Keywords: Language dominance; bilingualism; heritage speakers; lexical access; psycholinguistics.

Figure 1.

(a) Distribution of language dominance scores as measured with the Bilingual Language Profile (BLP; Gertken et al., 2014) for heritage and native Cantonese speakers included in our sample. (b) Priming magnitude density distributions (reaction times for related trials subtracted from unrelated trials on a by-participant basis) in the auditory translation priming experiment. A positive priming magnitude indicates that the related items were responded to more quickly compared with the unrelated items. The aggregate mean and standard error of the mean are also presented. Points represent individual participant priming magnitudes for each translation direction (L1–L2: Cantonese–English, blue; L2–L1: English–Cantonese, red). (c) Correlation of priming magnitudes (ms) with BLP dominance scores in each translation direction (L1–L2: Cantonese–English, blue; L2–L1: English–Cantonese, red).