Linguistic Approaches to Bilingualism - Volume 10, Issue 6, 2020

This study examines the influence of previously acquired languages – namely, Colloquial Arabic (CA) and English – on the acquisition of Standard Arabic (SA) by L3 and L2 learners. It reports on the role of typological and structural proximity in language transfer and whether transfer patterns change over time. The study involved 105 participants: 41 CA-L1, English-L2 learners of SA, 47 English-L1 learners of SA, and 17 Arabic-L1 speakers. The participants completed three written tasks focusing on: definite article use, verb subcategorization rules, and sentential negation. The results indicate that CA plays a positive role in L3 learners’ acquisition of SA, mainly in forms where SA and CA converge, whereas English seems to play more of a negative role for both L3 and L2 learners. Thus, structural proximity seems to play a positive role in transfer to the L3. Negative transfer, irrespective of proximity/distance, diminishes as learners advance in their study of SA.

We provide a snapshot of childhood morphology development in our investigation of two profiles of bilinguals (age 9–10) in an English-Spanish dual immersion academic setting: Spanish heritage language (SHL, n = 21) and second language (SL2, n = 41) children. Three tasks were given to the 62 bilinguals and 15 age-matched controls (Spanish first language, SL1): oral comprehension of 20 singular-plural present verbs, written sentence production of 10 similar verbs, and a meaning-focused writing task. SHL children were comparable to controls in production of number agreement, and showed no asymmetry between comprehension and production. SL2 learners showed lower accuracy than both SHL and SL1 children. A similar pattern was observed when person agreement and tense, aspect, mood and vowel errors were considered. The most common error among SHL and SL2 children was overregularization of stem vowels, a typical developmental error. The Feature Reassembly model of grammar can accommodate the range of possibilities represented by the data we present.

Visual world studies with bilinguals have demonstrated spontaneous cross-linguistic activations. In two experiments, we examined whether concurrent visual working memory (VWM) load constrains bilingual parallel activation during spoken word comprehension. Hindi-English bilinguals heard a spoken word in Hindi (L1) or English (L2) and saw a display containing the spoken word-referent, a phonological cohort of the spoken word’s translation and two unrelated objects. Participants completed a concurrent WM task of remembering an array of five coloured squares and judging its similarity with a test array. Participants were asked to click on the spoken word-referent in Experiment 1 but not in Experiment 2. Reduced parallel activation and enhanced target activation was observed under the load for L2 spoken words in Experiment 1 (where the task-demands were high). The findings suggest that a VWM load can constrain the spontaneous activation of an irrelevant lexicon, under certain conditions.

The present study describes the results of two experiments that analyze the effects of language dominance on Catalan/Spanish bilingual speech production. We examined the acoustics of the Catalan [a]~[ə] alternation (a phonological process induced by lexical stress) and of the Catalan mid vowel contrasts /e/-/ɛ/ and /o/-/ɔ/ (two phonemic contrasts) in the speech of Catalan- vs. Spanish-dominant early Catalan/Spanish bilinguals from Majorca (Spain). The results indicate that, contrary to the analysis of the Catalan mid vowels, stressed and unstressed Spanish and Catalan /a/ yielded no significant difference as a function of language dominance. These findings suggest that unstressed vowel reduction, a phonological process, may be relatively easier to acquire than phonemic contrasts with a low functional load (/e/-/ɛ/, /o/-/ɔ/), perhaps because its predictability and high frequency may attract attention and/or relieve cognitive resources, which could be conducive to phonological learning. These findings demonstrate that language dominance effects in the production of phonemic contrasts, widely researched in the literature on early bilingualism, do not guarantee the same effects on the implementation of phonological processes.

Recent research proposes that language bias and proficiency modulate cross-language activation in comprehension and production, but it is unclear how they operate and whether they interact. This study investigates whether stress differences between Spanish-English cognates (material, final-syllable stress in Spanish) affect how native-English second-language-Spanish bilinguals recognize Spanish words (materia “subject/matter,” second-syllable stress in Spanish). In a Spanish-English eye-tracking experiment (and parallel production task), participants heard/produced trisyllabic Spanish targets with second-syllable stress (materia) and saw four orthographic words, including the target and a Spanish-English cognate competitor. Cross-language activation was examined by manipulating the stress of the cognate in English. In comprehension, English cognates with the same stress as the Spanish target (materia vs material) were predicted to cause more cross-language interference than English cognates with a different stress (litera “bunk bed,” vs literal), but the reverse pattern was expected in production. Participants were assigned to a Spanish-bias condition (20% of English (filler) items), or an English-bias condition (65% of English (filler) items). Results indicate that English cognates with the same stress as the Spanish target interfered with the recognition of the Spanish target only in the English-bias condition (but facilitated its production), while increasing Spanish proficiency helped reduce this cross-linguistic interference.