Despite the status of English as the Lingua Franca of the world (Seidlhofer, 2011) and the growing number of L2 users (Pennycook, 2017), researchers have mostly investigated L2-accented English based on native speakers' perceptions. In...
moreDespite the status of English as the Lingua Franca of the world (Seidlhofer, 2011) and the growing number of L2 users (Pennycook, 2017), researchers have mostly investigated L2-accented English based on native speakers' perceptions. In particular, a number of previous studies have looked at native English speakers' perceptions of fluency (e.g., Bosker, Pinget, Quené, Sanders, & de Jong, 2013). Only a limited number of studies have examined L2 speakers' perceptions of the same phenomenon (for a rare exception, see Rossiter, 2009). Therefore, we know very little about how L2 users conceptualize fluency, and what their subjective conceptualizations of fluency are. Thus, the current study took the first step to investigate the factors affecting L2 users' intuitive perceptions of L2 fluency using an explanatory sequential mixed-methods design. Background Disentangling Fluency The importance of fluency has been recognized in successful L2 communication and language assessment contexts. Thus, previous studies have investigated which utterance features can predict fluency judgements by native speakers of target languages. Prior research has commonly reported that NSs' perception is associated largely with speed and breakdown fluency and secondarily with repair fluency (e.g., Bosker et al., 2013). However, prior research has exclusively employed native speakers as fluency judges while a number of studies have reported that L2 English communication takes place more frequently between L2 users than between L1 and L2 users. For the sake of the ecological validity of research findings, fluency research should be extended by examining L2 users' fluency judgements (Rossiter, 2009). Qualitative Approach to L2 learners Despite being limited in number, qualitative methods have uncovered affective and perceptual aspects of L2 learners. For example, Derwing, Munro and Thomson (2008) conducted individual interviews with 32 L2 English speakers in Canada (L1 Chinese and Russian/Ukrainian). Their results show that affective-cognitive, social, and motivational aspects play a role in L2 speakers' self-reported willingness to communicate in English, suggesting the importance of various aspects of L2 speech production. Regarding L2 learners' perception of fluency (the focus of this study), Rossiter's (2009) qualitative findings point towards temporal features of speech such as pausing and self-repetitions rather than their social perceptions. The raters in Rossiter (2009) also negatively reacted to non-temporal features of speech during their fluency judgements including pronunciation, vocabulary, and perceived confidence. Notably, both novice and expert raters gave a lot of negative comments on lexical errors perhaps due to their strong awareness of lexical choices and appropriateness than that of L2 speakers. Similarly, Tavakoli and Hunter (2018) explored teachers' understanding of fluency using a mixed-methods approach. The results demonstrated that teachers define fluency as an overall speaking ability; teachers conflated the distinction between fluency and speaking ability. Hence, their qualitative data showed that classroom teachers may address fluency in an incompatible way with findings from fluency research which commonly specifies fluency as temporal performance of speech. Our precursor research (Saito, Ilkan, Magne, Tran, & Suzuki, 2018) quantitatively explored a set of utterance fluency measures that differentiated between low, mid, and high levels of fluency judged by native speakers of British English in the context of picture narrative speech produced by 90 Japanese learners of English and 10 native speakers of Canadian English. The results showed that articulation rate differentiated high and nativelike