Abstract
This study explores the interface between learner beliefs and multilingualism in the under-researched English as a Foreign Language (EFL) context of Turkey. The study investigates the underlying constructs of a modified Beliefs about Language Learning Inventory (BALLI) completed by 168 EFL learners in Turkey using an exploratory factor analysis (EFA). One-way analyses of variance (ANOVAs) were run using two different group divisions for the independent variables: one analysis was run using the groups division of bilinguals versus multilinguals and the second analysis was run using the Perceived Positive Language Interaction (PPLI) group division. These analyses reveal significant differences with both categorizations. That is, knowledge of languages other than English, as well as seeing positive interactions between languages studied, can have a significant effect on learners' beliefs. The uniqueness of the study lies in its discussion of multilingualism within the realm of Dynamic Systems Theory (DST) in the Turkish EFL context. The results of this study underscore the impact of multilingualism on language learning and promote the study of interaction between multilingualism and learner beliefs in diverse contexts.
Notes
1. Two items from the original BALLI were removed: ‘The language I am trying to learn is structured in the same way as English’ and ‘If I speak this language very well, I will have many opportunities to use it’. Four questions were added to the questionnaire of this study: ‘I would like to learn English so that I can better understand people who speak English as a native language’, ‘The most important part of learning English is learning how to translate from English to the L1’, ‘I want to learn to speak English well’, and ‘Language learning involves a lot of memorization’. The wording was also revised to update certain concepts (i.e. ‘language labs’) and to make the questionnaire context specific (i.e. replacing ‘the foreign language’ with ‘English’). One question ‘If someone spent an hour a day learning a language, how long would it take him/her to become fluent’ was not included in the EFA because of the different format (multiple choice versus Likert scale). This is why the total number of questions in our study is 36, with 35 included in the analysis.
2. Some participants provided conflicting information on the background questionnaire and were eliminated accordingly.
3. The Cronbach's alpha in the present study was higher than those found in other studies (e.g. 0.631 in (Öz, Citation2007).