Does learners’ shared L1 use in higher order thinking tasks hinder L2 development?

A task helps second language learners experience and connect themselves to real-world communication. However, one of the challenges is to bring authenticity into the classroom of monolingual speakers (Japanese) enough to put real communicative demands on them. One way is to give discussion tasks in which learners speculate about issues that are generated from their own concerns. The benefits of L1 use in learner-learner interaction in terms of scaffolding (e.g., Bao & Duo, 2015) and task planning (Kunitz, 2018) have been shown, nonetheless, little is known about the impact of L1 in the activities where learners can engage in at higher cognitive level on L2 development. The data was collected in pre-/post design from an intermediate level of university students (B1, B2 at CEFR); one group (n = 53) who were allowed to use L1 in their discussions about current news and the other group (n = 23) used English only over two terms in six months. The discussion tasks (e.g., opinion-exchange, problem-solving, decision-making) were conducted following some information-exchange and reading tasks. To backup the research context, retrospective questionnaires about participants' L1 use were collected every after the class. The participants' language production in the number of sentences, plainness of English, lexical diversity, errors, and interpersonal metadiscourse markers (Bax, Waller, & Nakatsuhara, 2014) were analyzed using ANOVA. The findings showed that, the use of L1 did not hinder L2 production in lexical diversity and the volume of sentences, however, the students used English only produced more difficult/formal English against the other group and increased relational markers in the post-test. For pedagogical implications, provided that the use of shared L1 affects L2 development of discourse abilities, it is worth considering to minimize the use of L1. More details of the findings and implications will be also addressed.