Session 3B

It is generally accepted that bilinguals' L2 speech is less fluent than monolinguals; however, the relationship between bilinguals' fluency in their two languages is less clear (e.g., De Jong et al., 2005; Duran-Karaoz & Tavakoli, 2020). This study contributes to this topic by investigating the oral fluency of Tibetan-Mandarin bilinguals. Specifically, we ask whether their oral fluency (1) differs from Mandarin monolinguals (2) differs or correlates between their Tibetan and Mandarin (3) differs as a function of task (read vs. semi-spontaneous speech) and (4) associates with their language background and Mandarin test scores.

We collected data of read speech and semi-spontaneous speech from 52 participants, including 26 Tibetan-Mandarin bilingual (Tibetan and Mandarin languages, mean age of acquisition of Mandarin = 4.38 years) and 26 Mandarin monolingual (Mandarin language only) high school students. The data were collected in two waves, 32 in the first wave and 20 in the second. We used a Praat script (De Jong & Wempe, 2009) to extract fluency measures: speech rate, articulation rate, mean pause duration, and pause per second. Participants filled out Language History Questionnaire ([LHQ3], Li, Zhang, Yu, & Zhao, 2019) and reported their language background, language proficiency, language emersion, and language dominance in each language. In addition, we collected participants' concurrent mid-term Mandarin test scores and followed their college entrance Mandarin examination scores two years later.

Preliminary analyses of first wave (32 participants) yielded four main findings: (1) The difference in spoken Mandarin fluency between the monolinguals and bilinguals is mainly related to the number of pauses (F(1, 30) = 8.05, p = 0.008) and speech rate (F(1, 30) = 4.03, p = 0.054, partial η2 = 0.12), but not articulation rate (F(1, 30) = 0.15, p = 0.670); (2) Tibetan participants' Mandarin fluency measures are highly correlated with their Tibetan fluency in all measures for both tasks (Table 1); (3) Task type influences bilinguals' Chinese speech rate (F(1,15) = 13.76, p = 0.002) and articulation rate (F(1, 15) = 9.41, p = 0.008), but not their Tibetan speech rate and articulation rate; (4) Chinese speech rate predicts bilingual participants' mid-term Mandarin scores (r = 0.54, p = 0.04), though no fluency measure significantly associates with participants' college entrance examination scores or any LHQ scores (p's > 0.1).

This is the first study to examine Tibetan-Chinese bilinguals' oral fluency. In accordance with studies on the relationship between L1 and L2 fluency, we found a link between the oral fluency of L1 and L2. This link is stronger than English-Chinese bilinguals, English Slavic bilinguals (Derwing et al., 2009), and English-Turkish bilinguals (Duran-Karaoz & Tavakoli, 2020), suggesting that such a relationship may be related to the typological closeness between the two languages and the context of bilingual acquisition. We plan to include analyses of additional fluency measuresd-filled pauses and self-repairs-in our full-sample analysis.

Research investigating L2 fluency has indicated that pause location, as opposed to overall pause rate or pause duration, is particularly informative for differentiating L2 speech across different proficiency levels or between L1 and L2 speech (De Jong, 2016; Skehan et al., 2016): While L1 and L2 speakers have similar pausing at clause/message boundaries, L2 learners typically pause more often (and for longer) within clause/message boundaries (Tavakoli, 2011). These findings have been connected to Levelt's model of L2 speech production via arguments that mid-clause pausing might reflect Formulation difficulties. Nevertheless, much work has relied on findings from similar speaking tasks and cross-sectional designs. Given potential task effects (Foster & Skehan, 1996) and individual differences in L1/L2 fluency (Peltonen, 2018), it is desirable to examine pause location from L1 and L2 data and L2 data over time from the same speakers on both monologic and dialogic tasks. 
The current study addressed these issues by investigating the rate, duration, and proportion of mid-clause pauses in two tasks (picture-based narrative and semi-structured interview) using data from a publicly-available, two-year longitudinal corpus which tracked English L1 learners of L2 French (n=29) or Spanish (n=27) required to spend a year residing abroad. This corpus is ideal to investigate these questions because it includes L1 and L2 speech from the same speakers and L2 speech from the same speakers at multiple time points, including before and during study abroad where proficiency increased.
Data included 11.5 hours of speech transcribed in CLAN following CHAT conventions, separated into Analysis of Speech Units, and coded for clauses. Three measurements of mid-clause pausing were calculated representing (a) frequency, (b) duration, and (c) proportion of mid-to-end-clause. Results indicated that speaking task had an impact on fluency in both L1 and L2 speech. In their L1, participants demonstrated higher fluency on the narrative task. In the L2, participants showed improvement on each measure in the narrative but remained less fluent in their L2 vs. their L1. In the interview, the only measure that consistently differentiated L1 from L2 speech was frequency. The finding that L1-L2 mid-clause pausing patterns differ across tasks provides important methodological implications and raises new questions about the relationship between mid-clause pausing and Formulation stages of L2 speech production models.


References
De Jong, N. H. (2016). Predicting pauses in L1 and L2 speech: the effects of utterance boundaries and word frequency. International Review of Applied Linguistics in Language Teaching, 54, 113-132.
Foster, P., & Skehan, P. (1996). The influence of planning and task type on second language performance. Studies in Second Language Acquisition, 18, 299-323.
Peltonen, P. (2018). Exploring connections between first and second language fluency: A mixed methods approach. The Modern Language Journal. doi:10.1111/modl.12516
Skehan, P., Foster, P., & Shum, S. (2016). Ladders and Snakes in Second Language Fluency. International Review of Applied Linguistics in Language Teaching, 54, 97-111. 
Tavakoli, P. (2011). Pausing patterns: differences between L2 learners and native speakers. ELT Journal, 65, 71-79.

Fluency is a central aspect of L2 oral proficiency, which has been extensively studied in various fields, including SLA and learner corpus research. While some pedagogically oriented studies on the effects of fluency instruction on L2 fluency development in formal instruction have been conducted (e.g., Author(s) 1; Tavakoli et al. 2016), the effects of teaching on fluency development remain somewhat inconclusive, potentially due to the different teaching approaches. Despite recommendations for effective fluency teaching, especially involving the use of learner corpora and authentic examples of learner speech (e.g., Author(s) 2; Derwing 2017), there is a lack of empirical studies putting these pedagogical recommendations into practice and investigating their efficiency in facilitating L2 speech fluency development.
To fill this gap in research, the present study examines the influence of a corpus-based teaching approach on the development of L2 fluency. In the study, 20 university students of English participated in a course aimed at improving their spoken L2 English. The course included twelve 90-minute language laboratory sessions once a week. The students were divided into an experimental group (n=12) and a control group (n=8), and both groups were taught by the first author. Pre- and post-tests, including a read-aloud task and a picture description task, were administered for both groups. The control group received traditional English pronunciation and fluency teaching based on listen-and-repeat and reading aloud exercises, while the experimental group received additional instruction based on the novel learner corpus approach targeting three fluency features: the role of L1 individual speaking style in L2 speech fluency, fluency-enhancing strategies, and formulaic sequences (e.g., Götz 2013; Author(s) 3). The three-step teaching approach included teacher-led awareness-raising of a target feature, exposure to the feature via authentic learner corpus examples, and practice of the feature. 
The data were analyzed with widely used temporal fluency measures capturing three aspects of L2 fluency: speed, breakdown, and repair (e.g., Skehan 2009). A mixed-methods approach involving quantitative statistical analyses and qualitative analyses was used to examine group level and individual L2 fluency development. The quantitative analysis demonstrated some differences in fluency development across the control and experimental groups, but not for all dimensions of fluency. The complementary qualitative analysis provided insights into individual developmental trajectories. Methodologically, the study advocates for a more wide-spread use of authentic learner corpus data in the teaching of L2 fluency. Based on the results, practical suggestions for the development of L2 fluency teaching practices will be discussed.
References
Author(s) 1. Details removed.
Author(s) 2. Details removed.
Author(s) 3. Details removed.
Derwing, T. M. (2017). L2 fluency development. In S. Loewen & M. Sato (Eds.), The Routledge handbook of instructed second language acquisition (pp. 246–259). Routledge.
Götz, S. (2013). Fluency in native and nonnative English speech. John Benjamins.
Skehan, P. (2009). Modelling second language performance: Integrating complexity, accuracy, fluency, and lexis. Applied Linguistics, 30, 510–532. 
Tavakoli, P., Campbell, C., & McCormack, J. (2016). Development of speech fluency over a short period of time: Effects of pedagogic intervention. TESOL Quarterly, 50, 447–471.

Focusing on L2 Russian learners, this paper explores the impact of a short-term study abroad (SA) sojourn on the development of sociopragmatic variation patterns realised in the speech acts of request, apology and offer. By means of a pre- and post-SA written discourse completion task (WDCT), it examines the acquisition trajectory of strategies and linguistic devices deployed to realise the speech acts in paired scenarios differentiated in terms of social distance, imposition and respect for the interlocutor. Specifically, it aims to answer the following research questions:
- How does a short-term SA impact on sociopragmatic variation patterns?
- To what extent are changes informed by social variables?
This paper adds to the small body of research (e.g. Dubinina & Malamud, 2017; Krulatz, 2012; Owen, 2001; Shardakova, 2005) into the development of interlanguage pragmatics in L2 Russian speakers.
The participants consist of a cohort of 18 third-level learners of Russian who self-reported belonging to three proficiency levels – beginner, intermediate and advanced. The sojourn lasted eight weeks. The learners attended daily lessons and lived together in student accommodation. A WDCT was compiled using four paired scenarios differentiated in terms of social variables laid out above. The WDCT was administered at the beginning of the sojourn, before classes started, and at the end, when the course had finished. The data were analysed following an adapted version of the cross-cultural speech act realisation tool developed by Blum-Kulka & Olshtain (1983) taking into account the strategies of head act, IFID, adjuncts, modifiers and nominal address terms (NATs). The final strategy is particular to Russian, and, to the researcher's knowledge, has not previously been studied in interlanguage pragmatics. The system of NATs in Russian is highly complex, and choices carry sociopragmatic weight (Offord & Gogolitsyna, 2005).
Results indicate that all levels of learners enact sociopragmatic variation to varying degrees before SA. After SA, differential patterns were found, unconnected to proficiency level, mainly within the strategies of modifiers and NATs. Learners used modifiers and NATs to a significantly higher degree after SA and deployed both differentially according to the social constraints of the scenario. We posit that this happened due to increased attendance to the variable of respect and, to a lesser extent, imposition.
References
Blum-Kulka, S., & Olshtain, E. (1983). Requests and Apologies : A Cross-Cultural Study of Speech Act Realization Patterns (CCSARP). Applied Linguistics, 5(3), 196–213.
Dubinina, I. Y., & Malamud, S. A. (2017). Emergent communicative norms in a contact language: Indirect requests in heritage Russian. Linguistics, 55(1), 67–116. 
Krulatz, A. M. (2012). Interlanguage pragmatics in Russian: The speech act of request in email. University of Utah.
Offord, D., & Gogolitsyna, N. (2005). Using Russian: A guide to contemporary usage. Cambridge: Cambridge University Press. 
Owen, J. S. (2001). Interlanguage pragmatics in Russian: A study of the effects of study abroad and proficiency levels on request strategies. BrynMawr College, BrynMawr, PA.
Shardakova, M. (2005). Intercultural pragmatics in the speech of American L2 learners of Russian: Apologies offered by Americans in Russian . Intercultural Pragmatics, 2, 423–451.