Multiple wh-interrogatives in child heritage Romanian: On-line comprehension and production

Abstract

This study compared the online comprehension and the production of multiple interrogatives in 18 Romanian-English bilingual children aged 6;0–9;2 (MAGE = 8;0) living in the UK who have Romanian as heritage language (L1) and English as majority language (L2) and 32 Romanian monolingual children aged 6;11 to 9;8 (MAGE = 8;3). We examined whether differences emerge between heritage and monolingual children in the online comprehension and in the production of multiple interrogatives in Romanian, which requires fronting of all wh-phrases, contrary to English. The main aim was to uncover to which extent similarities or differences in morphosyntactic properties between the L1 and the L2 systems affect the acquisition and processing of the heritage language/L1. Online comprehension was assessed in a self-paced listening task, while production was assessed using an elicitation task. The results reveal that Romanian heritage children show similar online comprehension patterns to monolingual children for multiple interrogatives in Romanian. A different pattern emerges for production as heritage children produce less complex multiple questions in Romanian and avoid movement of two wh-phrases in all elicited structures. Given that their predominant responses for multiple interrogatives only make use of the structural option present in English, namely one fronted wh-phrase and one in-situ, we take this to show that there is transfer from the majority language to the heritage language. Thus, language production in the children’s L1 seems to be affected by properties of the dominant L2, under cross-linguistic influence. Taken together, the results for both comprehension and production suggest that heritage children are able to establish the underlying representation of multiple wh-movement structures, similarly to monolinguals, but have difficulties activating the more complex structure in production.

Language processing in bilingual children

“…Most of the studies comparing children's production and online comprehension have focussed on sequential bilingual children's later acquired societal language (e.g., Blom & Vasić, 2011;Chondrogianni & Marinis, 2012;Chondrogianni, Vasić et al, 2015). Bentea and Marinis (2022), however, compared online comprehension and production of fronting of Romanian wh-phrases in child heritage learners of Romanian with English as societal language. Like some of the studies testing children's societal language, Bentea and Marinis observed an asymmetry between production and online comprehension in the heritage language: bilingual children were less accurate in their production of wh-fronting in Romanian than monolingual children, but nevertheless showed similar behaviour online.…”

The problem of pseudoclefts in French

“…) and D-linked 1 object wh-questions (e.g. Cerejeira 2009;Friedmann & Novogrodsky 2011;Baião 2014;Bentea, Durrleman & Rizzi 2016;Bentea 2017;a.o.). This previous research defined intervention effects in terms of featural similarity between the moved element and the intervening element (see Rizzi 2018): it has been suggested that feature identity corresponds to a configuration which is rejected by both children and adults and feature disjunction (as in the case of non-D-linked wh-questions and free relatives) is accepted and easily processed by both children and adults; in contrast, feature inclusion, as in D-linked wh-questions and headed relative clauses, is avoided by children, even though it is a possibility in the adult grammar (see the summary and general proposal in Friedmann, Belletti & Rizzi, 2009).…”