Reverse transfer in trilingual acquisition of relative clauses

Abstract

This study reports results of an experiment testing Hong Kong trilingual children’s comprehension of relative clause in their Cantonese, English and Mandarin. Comparisons of the trilingual children with their monolingual counterparts are made with a focus on the directionality of cross-linguistic influence. The experiments test the predictions of the structural distance and linear distance accounts for the acquisition of relative clauses (Chan, Matthews and Yip 2011; Kidd, Chan and Chiu 2015). The experimental group included twenty-two trilingual five- and six- year olds from an international school with an English immersion environment who were tested on their comprehension of subject and object relative clauses in Cantonese, Mandarin and English. In addition, 24 monolingual Cantonese children in Hong Kong, 24 monolingual English children in UK, and 25 monolingual Mandarin children in China were tested as control groups. All children were asked to perform a pointing task requiring them to identify the referent of the head noun in subject and object relative clauses. The results show a highly significant advantage for subject over object relatives in the trilinguals’ second language English, and a non-significant subject advantage in both their first language Cantonese and second language Mandarin. There was no evidence of cross-linguistic influence from first language Cantonese to second language English or Mandarin. In Cantonese, however, the trilingual group made significantly more errors than their monolingual peers with object relatives, even though Cantonese is the first language for both groups. Analysis of errors in head identification suggests that the trilinguals parsed Cantonese object relatives using an English-based head initial analysis. This is treated as an instance of reverse transfer from the children’s second language, English to their first language, Cantonese.