Debates on a range of current affairs topics are commonly used by L2 teachers within a task-based pedagogical methodology to develop learners’ oral fluency. However, debates cannot be expected to boost fluency unless they are designed in a specific way to promote the L2 learner’s motivation to successfully engage in this task for genuine communication purposes. Whilst it is clear that this task offers valuable oral practice for learners, the potential debates offer for fluency development is largely undiscovered, in particular, in terms of how these can boost cognitive processes for creative automatisation within the context of meaningful communicative interaction. This study is underpinned by Segalowitz’ (2010) creative automaticity model and fluency methodology ACCESS, conceived to enable learners to automatise their speech in a creative manner with the aim of enhancing oral fluency through practice and within a genuine communicative context. It is also based on Dörnyei’s (2005) motivational model leading to the perception of motivation as a tool that promotes task engagement in the L2 classroom. The main aim of this study is to investigate the impact of a conclusion-based debate on fluency in L2 Spanish. Producing a conclusion at the end of a debate involves a process of listening to the contributions of all participants and deciding what the main views presented are, which is different to simply contributing to the debate. This study is aimed at investigating the impact of this reasoned process on L2 fluency. The task designed for this study involved participating in a debate on current global news. 56 second year university students of Spanish took part in this debate in 10 different classes. Each class was divided into group A (conclusion outcome) and group B (discussion only). Participants in all classes and groups were audio and video recorded for the duration of the debate. This study also aimed at shedding light on the causes for dysfluencies during the speech process. For this purpose, cards containing two potential explanations for pausing during speech, linked to the conceptualization and formulation stages of Levelt’s (1989, 1999a) speech production model, were used by the participants to indicate the causes for their pausing. Prior to the debate, all participants had 10-minute planning time with the support of a stimulus sheet containing a selection of news headlines selected and adapted from a national Spanish newspaper in order to help them focus their contributions during the debate. At the end of each debate, all participants completed a questionnaire on their perceptions about fluency development. A set of assumptions was presented with regards to the participants’ perceptions on fluency and contrasted with their actual fluency outcomes in the debate. These assumptions regarding fluency and motivational aspects were all met or partially met, including whether reaching an outcome at the end of a debate leads to higher fluency outcomes. Participants’ oral performances were analysed using a range of fluency measures based on Skehan’s framework which represent the three aspects of fluency in SLA research, that is, speed, breakdown and repair. The findings revealed that oral fluency increased in two of its measures, namely articulation rate and frequency of repair, specifically, reformulations and self-corrections, for the A group participants who produced the conclusion as they uttered a slightly higher number of words within the 20 second sample (excluding pausing). It was also revealed that these participants showed a lower rate of mid-clause pausing. It can, therefore, be concluded that reaching a conclusion at the end of a debate had a partial effect on the speech production process and, ultimately, on fluency. These findings give way to a range of theoretical implications regarding the contribution this study makes to creative automatisation within the framework of Segalowitz’ automaticity model and ACCESS, as well as motivation research following Dörnyei’s model. The main theoretical implication is that the processes of planning, priming and monitoring during the task would have led to an increased cognitive fluency in the conclusion utterers speeding up lexical retrieval and freeing up attentional resources for the formulation of new language chunks. They also have methodological implications, in particular, with regard to the effectiveness of the use of dysfluency explanatory cards, the benefits of adopting a mixed-methods approach which elicits unique insights otherwise unidentifiable if statistical analyses were used in isolation, and the contribution this study has made in terms of its findings with regards to grammatical encoding having more weight than lexical retrieval as the main cause for dysfluency in speech in L2 learners resulting from the word by word speech analysis carried out. Finally, the pedagogical implications drawn from the current study focus on the benefits of integrating a conclusion at the end of debates and ways in which to reduce dysfluencies to further enhance fluency in speech from both the teacher’s and the L2 learner’s perspective.
