Resumen de Do L2 learners show attachment preferences?: the processing of participial relative clauses in L2 Russian by English–Russian bilinguals

Zoe Pei sui Luk, Liubov Darzhinova

• Aims and objectives:This study investigated whether first language (L1) English and second language (L2) Russian bilinguals are sensitive to structural information during L2 processing by testing their attachment preference when reading participial relative clauses with three preceding noun phrases in Russian.

  Methodology:The participants read sentences containing a participial relative clause preceded by three noun phrases connected by genitive marking and responded to comprehension questions by indicating which noun phrase the participial relative clause modified or that the sentence was ungrammatical. Their reading times and post-reading responses were recorded.

  Data and analysis:The reading times were log-transformed and fitted to a generalized linear mixed-effects model. Ungrammaticality judgment rates and noun phrase selection accuracy rates were fitted to generalized linear mixed-effects models using a binary logistic distribution.

  Findings:Reading times and ungrammaticality judgment rates showed that the participants preferred to attach the participial relative clause to the first and the third noun phrases. The second noun phrase was least preferred. These results contrast findings in previous studies, which have shown no attachment preference for L2 learners. We argue that the participants were sensitive to structural information, in that they were constrained by the recency preference and the predicate proximity effects, similar to native speakers.

  Originality:To the best of our knowledge, this is the first study that tested L2 learners with relative clauses preceded by three noun phrases and showed that L2 learners are also constrained by the recency preference and predicate proximity effects.

  Significance/Implications:We argue that our findings deviate from the null attachment preference found in previous studies, which has been used as evidence that L2 processing is fundamentally different from L1 processing. Our findings suggest that L2 learners were influenced by predicate proximity, similar to native Russian speakers, and were also mildly influenced by recency preference, which could be from this L1.