The majority of the second language (L2) fluency studies have looked at oral fluency in a monologic mode except for (Witton-Davies, 2014; Tavakoli, 2016; and Peltonen, 2017) who have addressed the topic fluency in a dialogic mode. Dialogue is the natural way of everyday communication (Garrod and Pickering, 2004). However, little is known about how fluency is measured in a dialogic performance for example, in terms of filled/unfilled pauses, and pause locations. Additionally, it has been argued that the individual differences in working memory capacity could predict the oral task performance, where working memory is able to temporary store the information that is important to complete cognitive activities, such as speaking. The relationship between working memory capacity and oral production is of great interest. Thus, this study aims to fill the gap in the previous literature regarding whether the individual differences in working memory capacity can predict L2 oral fluency performance in monologue and dialogue as measured by speed breakdown, and repair. The current study sets out to test the relationship between sixty-four L2 English learners’ oral fluency and WMC in monologic and dialogic tasks. The participants were undergraduate students in English language and literature department at the university of Jeddah, Saudi Arabia. The participants were asked to complete two working memory tests: Backward digit span and operation span test. Additionally, the participants completed dialogic and monologic tasks. Monologue task was a decision-making task. The participants were asked to answer the question with justifications for their choice. As for the dialogic task, the participants in pairs (34 pairs) were asked to discuss a popular topic about their hometown. The oral speaking tasks were analyzed by PRAAT. Pause locations, pause durations, and filled/silent pauses were identified manually using PRAAT. Monologue and dialogue were transcribed into Analysis of speech units (ASU). Seventeen fluency measures were used to assess the participants’ oral fluency in both task modes. For example, measures of mid and end clause filled/ silent pauses, speech rate, articulation rate, and repairs. The results showed no relationship between monologue and dialogue performances, and working memory capacity, meaning that working memory was not a Predictive power for L2 fluency in monologue and dialogue. Furthermore, dialogic performance was more fluent than monologic performance in terms of speed and breakdown but not repair fluency measure. This result was in line with previous studies in L2 fluency.
