CODE-MIXING OF TRILINGUAL CHILDREN: AN INVESTIGATION ON THEIR LANGUAGES CHOICES

The general aim of the present study was to investigate the vocabulary interaction of the three languages of bilingual primary school children who are currently learning their L-3 English in an educational context. To this end, 49 primary education children aged between 9 and 12, all coming from an immigrant background, born and raised in Greece and all being trilingual (Greek and Albanian, L1 or L2 and L-3 English) were asked to narrate a picture story in their L-3 in order to elicit their speech productions. We wished to investigate the distribution and the trends of the three languages involved within the code-mixes found in the corpus, as well as the influence of participants' heritage and societal languages on their L-3 speech productions. The results showed that the majority of the code-mixes produced were English-Greek and that Greek, the societal language, was the main source of cross-linguistic influence, and that the participants' heritage language, Albanian, was used to a rather small extent. The participants' code-mixes were also analysed and it was found that English (L-3) was used as the Matrix Language while Greek was the embedded one. The participants were also found to follow the English grammatical constraints. They used the equivalent Greek words in the place of the English ones when they could not retrieve them during their speech productions.