Linguistic Approaches to Bilingualism - Volume 15, Issue 3, 2025

The article reports on a longitudinal study of syntactic cross-linguistic influence (CLI) among L1 Polish learners of L2 English and L3 Norwegian. The study mainly aimed to determine the influence of gradience in L1 on third language acquisition. To this end, four syntactic properties were tested, two of which exhibit similarity between Polish and Norwegian (subject-oriented possessive pronouns and adverb placement), and the other two – between English and Norwegian (definite and indefinite articles). A group of 24 learners of Norwegian participated in an acceptability judgment task, which was administered at three data collection times in all three languages. It aimed not only to determine the presence (and sources) of CLI, but also to observe how gradience in L1 affects the assessment of equivalent properties in L3. In order to assess the role played by gradience, the trilinguals’ performance was compared to that of a control group of English-Norwegian bilinguals. The data were analyzed with mixed-effects ordinal logistic regression modelling, which showed statistically significant differences in the ratings of articles between the two groups. We attribute this finding to gradient acceptability characterizing subject-oriented pronouns and adverb placement in Polish, which is a potential source of non-facilitative CLI.

We explore potential effects of prosody on pronoun interpretation in Italian, building on previous research which has shown that second language learners/users (L2ers) assign non-target interpretations to overt pronouns. We investigate effects of contrastive stress and pause, proposing that these will result in changes to default antecedent preferences for overt and null pronouns, for L2ers and for native speakers. An experiment was conducted, involving English-speaking L2ers of Italian and Italian native speakers. Participants were presented with auditory stimuli like Lorenzo ha scritto a Roberto quando Ø/lui si è trasferito a Torino ‘Lorenzo wrote to Roberto when (he) moved to Turin’ and indicated their preferred antecedent for the pronoun. Overt versus null pronouns, presence versus absence of stress on overt pronouns, and presence versus absence of pause between clauses were manipulated. The results yielded significant differences for antecedent choices between null and overt pronouns, consistent with earlier literature. In addition, stress was significant for both groups. Implications of a prosodic approach to ambiguity resolution are discussed.

Research on heritage language acquisition at the school age has shown protracted development and early stabilisation in morphosyntax and the lexicon. Our study examined the properties of resultative verb compound (RVC), a structure at the crossroads of the lexicon and morphosyntax, in second-generation child heritage speakers in the UK who had continuous input in Mandarin Chinese since birth. We analysed three subclasses of RVCs produced by the heritage children (n = 27, age 4–14) and their parents (n = 18) in an oral narration task and compared them with those by children in Beijing (n = 48, age 4–9) from existing databases. Our results show that the heritage children produced RVCs quite frequently and felicitously yet highly repetitively and conservatively, with a remarkably large proportion of their RVCs consisting of a strongly lexicalised subclass with direct lexical equivalents in English. Correlational analyses show that the heritage children’s RVCs improve with age, rather than provision of RVC in the parental input, indicating the role of cumulative input in RVC acquisition. Overall, the development of RVC in heritage Mandarin is delayed rather than stabilised or attrited, supporting the lexical account for grammatical vulnerabilities in proficient heritage speakers.

In this study, we investigate the contact effects of stability, convergence, and divergence regarding the use of the same linguistic construction in the same contact situation. To do that, we collected experimental production and judgment data by native German speakers living in the Netherlands regarding their usage of the complementizer um ‘to’ in German and compared those data to those of a control group of German speakers not in contact with Dutch. The results show that most speakers show evidence for some contact-induced language change in their German. At the same time, speakers seem to experience different contact effects, demonstrating that it is not the structural properties of the construction that result in one effect over the other, but rather factors that pertain to the individual speakers. In particular, we argue that speakers can either focus on the similarities or on the differences between their languages, to some extent driven by their attitudes towards their languages and language change, and then over-generalize these similarities or differences to new contexts. Overall, this result clearly underlines the importance of focusing on individual speakers as the initiators of language change, which is in line with a usage-based approach.

This study presents new evidence for the structural unergative–unaccusative distinction, in second language (L2) grammars, focusing on elementary-level Japanese-speaking learners of English (JLEs). The underlying distinction of unergatives–unaccusatives is often obscured on the surface strings due to independent syntactic properties such as feature-driven subject movement (in English) or headedness (in Japanese). Nevertheless, based on previous findings, elementary-level JLEs are expected to have reset headedness but have not acquired subject movement. Then, the resulting representation would not involve the properties obscuring the underlying unergative–unaccusative distinction and potentially exhibit it on the surface strings in L2 English. Following these observations, we carefully designed test sentences with un/grammatical word orders that elementary-level JLEs would generate and conducted an acceptability judgment task with native speakers of English and elementary-/intermediate-level JLEs. The results showed that, in contrast to native controls and intermediate learners, who exhibited target-like patterns, elementary-level JLEs incorrectly accepted ungrammatical word orders only with unaccusatives (e.g., *When did arrive the train?) but not those with unergatives (e.g., *Where did dance the man?). This discrepancy can be attributed to the sensitivity to the structural distinction of unergative–unaccusative verbs, and our data provide evidence for the creative construction of an interlanguage in L2 acquisition.