No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Fluency in monologic and dialogic task performance: Challenges in defining and measuring L2 fluency
Abstract
The study reported in this paper challenges current models of measuring second language fluency by comparing monologic versus dialogic task performance, and providing a novel insight into the measurement of the interactive aspects of dialogic performance. The data that constitute 35 monologic and dialogic task performances from second language learners were coded using a battery of established measures known to tap different aspects of fluency, and subjected to statistical analysis to test for overlaps or differences. Interactive aspects of fluency in dialogue, e. g. interruptions, overlap and unclaimed between turn pauses were also investigated to compare with common measures of monologic speech. While the results confirm previous research findings suggesting that performance is in general statistically more fluent in a dialogue in terms of speed, length of pause and repair measures, they indicate that performances in the two modes are not different in terms of number and location of pauses. The analysis of the dialogues indicates that the decisions researchers make about measuring the interactive aspects of fluency would have an impact on the outcome of measurements of fluency. These findings highlight the need for developing a more systematic and reliable approach to measuring second language (L2) fluency.
... Gaps, as per Sacks et al., are shorter silences between speaker turns or at transition-relevance places, which are points where an utterance can potentially be completed. Alternative terms for gaps include turn pauses (Peltonen, 2017;Van Os et al., 2020) or between-turn pauses (Peltonen, 2022;Tavakoli, 2016). Lapses are similar to gaps but of longer duration. ...
... The collected conversations were carefully examined for even distribution of turns and balanced participation between interlocutors. Following Tavakoli (2016), conversations that exhibited significant imbalance, with one speaker dominating the dialogue for over 70% of the time or the other speaker remaining quiet for such a long period, were excluded from the analysis. In addition, the data were scrutinized to ensure that each participant had a minimum of two turns within the 3-minute duration of the dialogue. ...
... 2 In the case of dialogic speech samples, turn pauses (i.e., the silence between speaker changes, Trouvain & Werner, 2022) were annotated and differentiated from the silent pauses within individual turns. As Tavakoli (2016) noted, it can be challenging to attribute such between-turn pauses to individual speakers. There are two feasible approaches for handling these pauses in L2 fluency research: one is to exclude them from the analysis, and the other is to divide the pauses equally between the two speakers. ...
Second language (L2) utterance fluency is crucial for speaking-proficiency assessment. The measurement of L2 utterance fluency relies heavily on silent pause identification. However, empirical studies establishing specific silent pause thresholds for L2 monologic speaking are scarce, and even fewer exist for L2 dialogic speaking. This study thus aims to explore the optimal silent pause threshold for measuring L2 utterance fluency in both speaking contexts. Following De Jong and Bosker, the study assumes that silent pause measures calculated at optimal thresholds can best predict L2 proficiency and perceived fluency. Multiple linear regression models were conducted using four silent pause measures (mid-Analysis of Speech Unit [ASU] silent pause rate, end-ASU silent pause rate, mid-ASU silent pause duration, end-ASU silent pause duration) at various thresholds (100–1000 ms) to predict participants’ L2 proficiency and perceived fluency scores. For the monologic task, silent pause measures calculated at a threshold of 200 ms best predicted both L2 proficiency and perceived fluency scores. For the dialogic task, silent pause measures calculated at a 200-ms threshold had the best predictive power for L2 proficiency score, whereas the optimal threshold for predicting the perceived fluency score was 350 ms.
... Despite the importance of exploring L2 fluency in interactive tasks, there is a lack of conclusive evidence regarding the distinctions between monologic and dialogic L2 performances in terms of speed, breakdown, and repair. Previous studies (e.g., Michel, 2011 [1]; Witton-Davies, 2014 [2]; Tavakoli, 2016 [3]; Peltonen, 2017 [4]) have produced inconsistent findings regarding the differences in breakdown and repair measures between monologue and dialogue, necessitating further investigation. ...
... Researchers in second language acquisition, such as Tauroza and Allison (1990), Gilbert et al. Sato (2014) [20], Witton-Davies (2014) [2], Tavakoli (2016) [3], and Peltonen (2017) [4], delved into utterance fluency in dialogue tasks, suggesting that L2 speakers exhibit greater fluency in dialogues than monologues. Tauroza and Allison (1990) and Gilabert et al. (2011) endorsed the efficacy of dialogic tasks in measuring L2 fluency, revealing faster syllable rates in dialogue compared to monologue. ...
... Researchers in second language acquisition, such as Tauroza and Allison (1990), Gilbert et al. Sato (2014) [20], Witton-Davies (2014) [2], Tavakoli (2016) [3], and Peltonen (2017) [4], delved into utterance fluency in dialogue tasks, suggesting that L2 speakers exhibit greater fluency in dialogues than monologues. Tauroza and Allison (1990) and Gilabert et al. (2011) endorsed the efficacy of dialogic tasks in measuring L2 fluency, revealing faster syllable rates in dialogue compared to monologue. ...
Studies on second language (L2) speech fluency recommend studying fluency in a dialogic context. In response to the researchers' calls, this study introduced monologic and dialogic tasks to investigate the various aspects of speed, breakdown, and repair fluency in the oral performancemost L2 fluency studies have looked at oral fluency in a monologic task. Dialogue is the more authentic and natural way of communication, which is apparent in everyday language use. Currently, there is a scarcity of research examining dialogue fluency in non-native bilingual speakers who share the same L2. The existing body of research on language learning and processing has underscored significant connections between individual differences (IDs) in working memory capacity (WMC) and models of L2 speech production in both first language (L1). Nevertheless, it remains unclear whether variations in WMC are linked to dysfluency in L2 monologue and dialogue. Therefore, this study also aimed to fill this gap in the literature by investigating the correlation between utterance fluency in both monologue and dialogue and WMC. A total of 64 undergraduate Saudi students were given various tasks as part of the study. An argumentative task was presented as a monologue focused on a prevalent topic in the participants’ country. In contrast, during a dialogic discussion task, 32 pairs engaged in exchanging opinions on a popular subject in their country. Additionally, participants underwent two challenging working memory (WM) tests: the Operation Span Test and the Backward Digit Span Test. The findings aligned with prior research, indicating that L2 participants demonstrated greater fluency in dialogue compared to monologue, as evident in speed and breakdown measures of utterance fluency. Interestingly, WMC did not emerge as a robust predictor for variations in L2 oral performances between monologue and dialogue.
... From a pedagogical perspective, there are clear advantages for both L2 learners and language teachers to develop empirically-based understanding of the language features that have a positive impact on judgements of perceived fluency Tavakoli & Hunter, 2018). However, from a theoretical standpoint, much of our understanding of what influences listeners' perception of L2 fluency has been limited to the study of monologic speech and observable temporal features of fluency (Segalowitz, 2016), in terms of breakdown, repair, and speed (Tavakoli, 2016). While speed and pausing phenomena (frequency, duration, and placement) have a clear impact on listener-based judgments of fluency (Suzuki et al., 2021), features of speech characteristic of interactional competence and the co-construction of fluent communication have generally been overlooked in experimental research. ...
... Research investigating perceptions of spoken fluency has traditionally been limited to the examination of monologic rather than dialogic speech due to practicality and control, but this limitation represents a significant gap in the development of a complete understanding of fluency as a construct (McCarthy, 2009;Suzuki et al., 2021;Tavakoli, 2016). A growing number of studies have sought to identify the features that differentiate fluent production between monologic and dialogic speech. ...
... A growing number of studies have sought to identify the features that differentiate fluent production between monologic and dialogic speech. For example, Tavakoli (2016) and Witton-Davies (2013) found that dialogic performance resulted in faster speech rates, less pausing, and fewer repairs than in monologues. They speculated that increased listening and planning time contribute to ease in conceptualization and delivery. ...
... Speaking a language fluently is often the ultimate goal for language learners (Kormos & Dénes, 2004, p. 145). Consequently, a substantial body of research has been conducted to examine the definitions of oral fluency (Freed, 2000;Lennon, 1990;Tavakoli, 2016), its aspects (Koponen & Riggenbach, 2000;Préfontaine et al., 2016;Segalowitz, 2010), and the practices that promote it (Boers, 2014;Rossiter et al., 2010;Tavakoli et al., 2016;Tran & Saito, 2021). From a research perspective, fluency has been defined as the 'flow, continuity, automaticity, or smoothness of speech' (Koponen & Riggenbach, 2000, p. 6). ...
... Lennon, 1990;Segalowitz, 2010;Tavakoli & Skehan, 2005). One possible explanation for this lack of clarity is the complex and multifaceted nature of fluency (Freed, 2000;Segalowitz, 2010Segalowitz, , 2016Tavakoli, 2016), which makes it challenging to define precisely. During the interviews, when we asked the teachers to elucidate the relationship between fluency, speaking ability, and general L2 proficiency, they unanimously agreed that fluency is distinct from the other two concepts. ...
... Chambers, 1997;Figure 3. Effect plot of the interaction between Region and Certification. Freed, 2000;Segalowitz, 2010;Tavakoli, 2016). They provide additional evidence from the teachers' perspective, further supporting the notion that understanding fluency poses challenges for teachers (Tavakoli & Hunter, 2018). ...
This study, a partial replication of Tavakoli and Hunter, examined the perceptions and self-reported practices of 72 second language (L2) teachers of English or Japanese in China regarding oral fluency. The research employed a mixed-methods approach, collecting both quantitative and qualitative data through questionnaires and interviews. The findings aligned with the original study conducted in the UK, suggesting that L2 teachers in China often viewed fluency as encompassing speaking ability in a broad sense. Therefore, these teachers reported a large majority of free production tasks that enhanced speaking ability, rather than specifically focusing on fluency development. Collectively, these findings highlight a discrepancy between the scholarly definition of fluency and teachers’ understanding of it, and underscore the importance of examining the impact of teacher perceptions on their practices for promoting fluency. Additionally, this study provided novel insights by exploring the influence of teacher variables on teachers’ confidence in their knowledge about fluency. The findings demonstrated that certified teachers, teachers of multiple L2s, more experienced teachers, and teachers in more socioeconomically advantaged regions exhibited greater confidence in their knowledge about fluency. Furthermore, the interaction effects among various teacher variables were found, underlining the complexity of teachers’ roles in perceiving fluency. Considering the findings from both studies, adopting a narrower definition of fluency and incorporating more fluency-focused activities in the classroom could enhance the effectiveness of fluency teaching. Moreover, this study emphasizes the importance of considering teachers’ variability in their understanding of fluency.
... Speaking a language fluently is often the ultimate goal for language learners (Kormos & Dénes, 2004, p. 145). Consequently, a substantial body of research has been conducted to examine the definitions of oral fluency (Freed, 2000;Lennon, 1990;Tavakoli, 2016), its aspects (Koponen & Riggenbach, 2000;Préfontaine et al., 2016;Segalowitz, 2010), and the practices that promote it (Boers, 2014;Rossiter et al., 2010;Tavakoli et al., 2016;Tran & Saito, 2021). From a research perspective, fluency has been defined as the 'flow, continuity, automaticity, or smoothness of speech' (Koponen & Riggenbach, 2000, p. 6). ...
... Lennon, 1990;Segalowitz, 2010;Tavakoli & Skehan, 2005). One possible explanation for this lack of clarity is the complex and multifaceted nature of fluency (Freed, 2000;Segalowitz, 2010Segalowitz, , 2016Tavakoli, 2016), which makes it challenging to define precisely. During the interviews, when we asked the teachers to elucidate the relationship between fluency, speaking ability, and general L2 proficiency, they unanimously agreed that fluency is distinct from the other two concepts. ...
... Chambers, 1997;Figure 3. Effect plot of the interaction between Region and Certification. Freed, 2000;Segalowitz, 2010;Tavakoli, 2016). They provide additional evidence from the teachers' perspective, further supporting the notion that understanding fluency poses challenges for teachers (Tavakoli & Hunter, 2018). ...
This study, a partial replication of Tavakoli and Hunter, examined the perceptions and self-reported practices of 72 second language (L2) teachers of English or Japanese in China regarding oral fluency. The research employed a mixed-methods approach, collecting both quantitative and qualitative data through questionnaires and interviews. The findings aligned with the original study conducted in the UK, suggesting that L2 teachers in China often viewed fluency as encompassing speaking ability in a broad sense. Therefore, these teachers reported a large majority of free production tasks that enhanced speaking ability, rather than specifically focusing on fluency development. Collectively, these findings highlight a discrepancy between the scholarly definition of fluency and teachers’ understanding of it, and underscore the importance of examining the impact of teacher perceptions on their practices for promoting fluency. Additionally, this study provided novel insights by exploring the influence of teacher variables on teachers’ confidence in their knowledge about fluency. The findings demonstrated that certified teachers, teachers of multiple L2s, more experienced teachers, and teachers in more socioeconomically advantaged regions exhibited greater confidence in their knowledge about fluency. Furthermore, the interaction effects among various teacher variables were found, underlining the complexity of teachers’ roles in perceiving fluency. Considering the findings from both studies, adopting a narrower definition of fluency and incorporating more fluency-focused activities in the classroom could enhance the effectiveness of fluency teaching. Moreover, this study emphasizes the importance of considering teachers’ variability in their understanding of fluency.
... However, previous studies have focused on automated scoring of L2 oral proficiency in spoken monologues. Although acoustic characteristics of monologic speech are different from those of dialogic one [6], scholars have suffered from the lack of knowledge of valid operationalization of dialogic fluency features and the technologies capturing those features. To extend L2 research and pedagogies of L2 fluency, therefore, it is expected to develop an automated scoring system compatible with spoken dialogue with the better understanding of how different features contribute to human ratings of fluency in dialogues. ...
... A feature of disfluency phenomena is also captured by the frequency of self-correction and repetitions in an utterance. In addition to those three aspects of utterance fluency, it has been found, albeit only recently, that the smoothness of interactions can make a unique contribution to human judgments of L2 oral fluency in dialogue [6]. Different studies have tried to extract features such a dialogic aspect of fluency, including the number of turns, the duration of silent intervals between a speaker and an interlocutor, and the number of repeating words delivered by an interlocutor [8], [9]. ...
... Recently, researchers have also admitted that the nature of L2 fluency can differ between monologic and dialogic speaking tasks [6]. For example, utterance fluency measures can vary depending on the turn and topics in dialogue, even though they are calculated from same speakers' utterances [11], [17]. ...
We propose an automated scoring method of fluency that is compatible with second language dialogic responses. Human judgements of L2 oral fluency in dialogue tasks has
different nature from scoring of monologue, and it is necessary to capture a dialogic aspect of fluency. Because utterances in dialogue tend to be less fluent than in monologue, procedures such as pruning disfluency words and classifying pauses by
their syntactic locations are essential for automated scoring systems to extract utterance features strongly correlated to human ratings. However, existing automated utterance feature extractors have suffered from the difficulties to detect disfluency words
and pauses locations due to the technical challenges. Moreover, conventional automated scoring methods of L2 spoken dialogue often predict oral proficiency for each turn, and dialogic features have not been considered properly. To address these gaps between
the nature of fluency in dialogue and existing automated scoring methodologies, we refine an automated utterance feature extractor and design a fluency scoring model based on dialogic features. Experiments showed that the substantial agreement of disfluency
word and pause location detection between our feature extractor and human (Cohen’s κ >0.61). We also found that the proposed scoring method outperformed in predicting subjective fluency scores (QW κ = 0.833) than a conventional turn-level scoring
model (0.654) and even a manual rating (0.799). We additionally compared the current assessment approach considering disfluency features and pause location, and it improved the accuracy on predicting subjective fluency scores. These results may suggest what and how utterance and dialogic features should be utilized in automated scoring of spoken dialogue.
... They also spoke more fluently in the dialogic task, which might be a result of turn-taking behaviors. Tavakoli (2016) compared various fluency measures in monologic and dialogic conditions. It was reported that the dialogic task performance was more fluent in all aspects of fluency except for the number of clause-internal and clause-external pauses. ...
... Speech rate was calculated by dividing the total number of syllables in unpruned speech (i.e., including self-repetitions, self-repairs, replacement) by the total speaking time (Yuan & Ellis, 2003). Repair fluency was measured by the average number of filled pauses (e.g., um, er) per minute (Tavakoli, 2016). ...
... First, the existence of the interlocutor may have pushed participants to speak faster. Tavakoli (2016) found that the participants were more willing to speak faster when they took their interlocutors into consideration. It was found in this study that when the participants spoke too slow, they felt worried about being embarrassed in front of their partner or wasting their partner's time. ...
The present study sets out to explore the effects of pre-task planning and unpressured on-line planning on L2 learners’ oral performance and their choices of planning strategies in a dialogic task condition. Forty-eight intermediate Chinese EFL learners were invited to perform the task and were then assigned to four groups, each with a different planning condition. Complexity, accuracy, and fluency of their oral production were measured. Results indicated that in the dialogic task condition, unpressured on-line planning raised syntactic complexity. Strikingly, pre-task planning did not improve L2 performance in all dimensions. Additionally, a trade-off effect was found between complexity and accuracy. Retrospective interviews were conducted to explore strategies employed by the participants and their perceptions of task preparedness. Results showed that the participants preferred to use metacognitive strategies and social/affective strategies in the dialogic task. Both advantages and limitations were identified by the participants regarding different planning conditions.
... Finally, research on L2 speaker fluency (e.g., Suzuki et al., 2021) and engagement (e.g., Lambert et al., 2017;Qiu & Lo, 2017; see Hiver et al., 2024, for a recent review) has revealed verbal behaviors associated with more versus less optimal interactive performances. In dialogue, fluency often manifests through low frequency of filled pauses and repairs (Tavakoli, 2016), and dialogic speech is perceived as fluent when interlocutors provide backchannels, use discourse markers as hedges (e.g., sort of, like), and produce overlapping speech to show collaboration (Brown et al., 2023;Sato, 2014;van Os et al., 2020). In terms of speaker engagement, which broadly captures a person's action with respect to another person or task (Mercer, 2019), engaged speakers provide more conversational content, elaborate their ideas, and show more responsiveness toward their interlocutor through backchanneling (Dao & McDonough, 2018;Lambert et al., 2017;Nakamura et al., 2021;Qiu & Lo, 2017). ...
... Therefore, we may have inadvertently not included relevant linguistic markers of interpersonal insecurity relevant to metaperceptions. For instance, unfilled pauses as a marker of breakdown fluency might have been more consequential to speakers' metaperceptions than our measure of filled pauses, especially in dialogic performances (Tavakoli, 2016). Without video recordings, we also could not examine speakers' nonverbal behaviours and assess their comparative role in metaperceptions. ...
People worry about how they are seen by others, but their insights (called metaperceptions) are often too negative. For instance, many speakers believe that their interlocutors like them less than they actually do, and these overly negative metaperceptions inform speakers' actions such as asking for advice or pursuing friendships. Our goal was to understand if low, underconfident metaperceptions are associated with specific interactional behaviors for second language (L2) speakers, as a way of identifying a "linguistic signature" of insecure metaperceivers. We analyzed 10-min dyadic conversations by 37 L2-speaking university students discussing academic texts. Following the conversation, students provided their metaperceptions (how much they thought their partner liked them) and their actual assessments (how much they liked each other). We coded the conversations for eight measures of utterance fluency (repetitions, repairs, filled pauses, discourse markers) and speaker engagement (lexical content, mean length of turn, backchannels, overlapping speech). Whereas several measures predicted students' meta-perceptions, none contributed to their actual assessments. Speakers who felt appreciated by their partner provided more lexical content across shorter conversational turns, whereas those who felt insecure assumed a dominant role speaking in long turns. These findings provide initial insights into how speakers' metaperceptions manifest in their interactional behavior.
... The impact of task type on the fluency-proficiency relationship remains underexplored. Studies have frequently reported fluency differences between controlled and open-ended tasks (e.g., Cucchiarini et al., 2002;Eren et al., 2022;Lei, 2021) and between monologic and dialogic tasks (e.g., Tavakoli, 2016;Wright, 2020). These studies often refer to major theories to shed light on task complexity, such as Skehan's trade-off hypothesis (1998) and Robinson's cognition hypothesis (2001). ...
... Moreover, on dialogic tasks, the relationship between speed and proficiency is relatively weaker than on monologic tasks, whereas the correlation for repair features is slightly stronger. Previous studies revealed that speech on dialogic tasks is generally more fluent (e.g., Tavakoli, 2016). However, the differences in the fluency-proficiency relationship might suggest that on dialogic tasks the interlocutors focus more on interactive elements (e.g., managing turns for a smooth conversation), whereas on monologic tasks the speaker is more conscious of demonstrating the ability to deliver fluent runs independent of the help or influence from an interlocutor. ...
Abundant research has indicated fluency features as meaningful predictors of second language proficiency. However, the extent to which different fluency dimensions and features can predict proficiency remains underexplored. This meta‐analysis employed a multilevel modeling approach to synthesize fluency–proficiency relationships from 71 empirical studies from 1959–2023. Additionally, we examined several moderator variables, including task type, learning context, age, and proficiency measure. The correlations found were strongly positive for speed ( r = .55), moderately positive for productivity ( r = .38), moderately negative for breakdown ( r = −.33), and weakly negative for repair ( r = −.11). Moderator analyses revealed that task, learning context, and proficiency measure influence fluency–proficiency relationships in the repair, productivity, and speed dimensions, respectively. Post hoc analyses also suggested that the operationalization of breakdown features might make a difference in fluency–proficiency relationships. This study has both theoretical and methodological implications for second language fluency research.
... [6], focusing on speech rate [SR]). Furthermore, task-based research comparing L2 speech in monologic and dialogic conditions has tended to show higher fluency in dialogue (e.g., [7][8][9]), but these studies have explored other fluencemes besides SPs only to a relatively limited extent. Apart from taking the task type into consideration, researchers have also started to acknowledge the influence of first language (L1) speaking behavior on speakers' L2 performance. ...
... Regarding the distribution of fluencemes, they were found to be relatively similar across L1 and L2 interactions as well, again being potentially linked with the task-based nature of the interactions and advanced level of L2 proficiency. SPs were found to be the most frequent type in both L1 and L2 interactions, which is in line with previous L2 fluency research (e.g., [3,4,8]). In L2 interactions, FPs were the second most common type, while false starts were the second most common type in the L1 interactions. ...
... Studies of second language (L2) speech fluency have largely focused on monologs, while dialogues are rarely studied (McCarthy, 2010;Tavakoli, 2016;Foster, 2020). In dialogues, two or more interlocutors take turns contributing to the flow of interaction. ...
... The non-monadic perspective analyzes dialogue as a whole system with interaction and interdependence between interlocutors (McCarthy, 2010;Segalowitz, 2016;Tavakoli, 2016;Peltonen, 2017b;Tavakoli and Wright, 2020). As such, speaker stance, interactional competence, and interlocutors' cognitive factors are viewed as potentially contributing to fluency in L2 dialogues. ...
... Likewise, fluency is also multi-dimensional as the other two constructs in the tripartite CAF framework. Following Skehan and other researchers (Skehan, 2003(Skehan, , 2009Tavakoli, 2016;Tavakoli et al., 2016), fluency can be examined through its subdimensions such as speed fluency (the rate and density of linguistic units produced), breakdown fluency (number, length, and location of pauses) and repair fluency (false starts, misformulations, self-corrections, and repetitions) . According to Lambert and Kormos (2014), fluency metrics that are conceptualized based on speech rate (i.e., a ratio of syllables produced to time taken to produce them) are the most frequently used measures. ...
... Furthermore, the findings of this study lent some support to Skehan's Trade-off Hypothesis. The triarchic CAF framework generally defines language proficiency as the complex interplay of the three constructs, i.e., complexity, accuracy, and fluency (Tavakoli, 2016), which may be distinctively manifested under different conditions of L2/FL use. The three constructs may be differentially developed by different types of learners and under different learning conditions ). ...
Although the automatic speech recognition (ASR) technology is increasingly used for commercial purposes, its impact on language learning has not been extensively studied. Underpinned by the sociocultural theory, the present work examined the effects of leveraging ASR technology to support English vocabulary learning in a tertiary flipped setting. A control group and an experimental group of college students participated in a 14-week study. Both groups had their English classes in a flipped fashion, but the experimental group was assigned with ASR-assisted oral tasks for pre-class self-learning. The pre- and post-intervention in-class task performance of both groups was audio-recorded and transcribed for data analysis. The triadic complexity-accuracy-fluency (CAF) framework was adopted to evaluate the participants' vocabulary learning. The between- and within-subjects effects were examined mainly through procedures of MANCOVA and mixed-design repeated measures ANCOVA. Results showed that on all the metrics of lexical complexity and speed fluency, the experimental group outperformed the control group, and had significant growth over time. On the other hand, the control group only improved significantly overtime on the G-index. On lexical accuracy, there was no significant difference between the two groups, and the within-subjects effect was not significant for either group. The findings lent some support to Skehan's Trade-off Hypothesis and discussions were conducted regarding the triarchic CAF framework.
... Speaking rate (SR) is the most frequently used general measure of L2 fluency and can be unpruned or pruned speech rate. Unpruned speech rate refers to the mean number of unpruned syllables (i.e., including all dysfluencies such as false starts, repetitions, corrections, and non-lexical filled pauses) per second or minute (including silent pause time) (Chambers, 1997;Kormos & Dénes, 2004;Tavakoli, 2016). Pruned speech rate refers to the number of pruned syllables (i.e., the total number of syllables excluding all dysfluencies such as false starts, repetitions, corrections, and non-lexical filled pauses) per second or minute (including silent pause time). ...
... Recent research also suggests that measures of fluency need to be chosen carefully because one measure may overlap with others (Kormos, 2006;Skehan, 2014;Tavakoli, 2016;Tavakoli & Skehan, 2005). With respect to the relation between utterance fluency measures and cognitive processes of speech production, as discussed in Section 5.2, research suggests that speech rate relates to all stages of L2 speech production, the number of between-clause pauses reflects breakdowns in the conceptualization stage, the number of within-clause pauses is linked to breakdowns in the formulation stage, and the number of repairs (i.e., reformulations and replacements) represents monitoring. ...
This study investigates the effects of task-based instruction on EFL learners’ speech production. The study employed an experimental pre-test/post-test design with the participation of 102 Vietnamese EFL learners who were randomly assigned to two task-based instructional groups (input-based task instruction versus combined input-output task-based instruction) and a comparison group. The results indicate that instruction combining input- and output-based tasks was more effective in improving EFL learners’ speech production capacity than the use of input-based instruction alone.
... Measuring fluency has always been an important aspect of determining student proficiency (Fulcher, 2003) and heavily researched as it is an essential component of communicative language ability (Tavakoli, 2016) as well as an important descriptor of L2 development (de Jong et al., 2012). In a broad sense, this concept can be seen as an equivalent to overall speaking proficiency (Chambers, 1997). ...
... Students performed a weekly narrative production task which consisted of a oneminute spontaneous monologue explaining what happened in each student's week. Dialogue recordings were not considered due to issues arising over the complex pragmatics involved in measuring the interactive aspect of dialogues such as unclaimed pauses between turns, overlap, and interdependence of the interlocutor's performances (Tavakoli, 2016). Due to simplicity and reliability, it would, therefore, be prudent to analyse individual narratives to measure each student's spontaneous speaking ability. ...
... From the perspective of L2 fluency studies, different task designs, such as closed (e.g., picture narrative) and open tasks (e.g., argumentation), result in different cognitive demands (Suzuki & Kormos, 2023), possibly differentiating how oral proficiency and cognitive processes are reflected in temporal speech characteristics including hesitations and pausing behaviors across tasks. Task modality (monologue vs. dialogue) also impacts the triad of UF because in dialogue a speaker can utilize an interlocutor's utterances as speech production resources (i.e., alignment), potentially leading to faster speech and fewer pauses and disfluency than monologue (Tavakoli, 2016). ...
Machine learning (ML) techniques allow for automatically annotating various temporal speech features, particularly by the cascade connection of ML-based modules. Although such systems are expected to enhance scalability of second language (L2) speech research, their annotation accuracy is potentially moderated by speaking tasks and proficiency levels due to the mismatch between training and real-world data. Accordingly, we developed and validated an ML-based temporal feature annotation system on L2 English datasets split by speaking tasks (monologic vs. dialogic tasks) and proficiency levels, operationalized as overall fluency levels (low, mid vs. high). We compared the annotations by experts and the system in terms of the agreement between manual and automatic annotations, correlations between manual and automatic measures, and the predictive power for listener-based fluency judgments. Results showed a substantial degree of agreement in the annotations for monologic tasks and a general tendency of strong correlations between manual and automatic measures regardless of tasks and overall fluency levels. Furthermore, automatic measures yielded substantial predictive power of fluency scores in monologic tasks. These findings suggest the substantial applicability of ML-based annotation systems to monologic tasks possibly without biases by holistic levels of fluency.
... User engagement is defined as the system's ability to maintain the user's interest and encourage ongoing interaction [30]. While, fluency means that the generated responses are grammatically correct and overall easy to understand which includes the smoothness of the conversation, clear communication, and maintaining a natural conversational flow [31]. The relevance is described as the appropriate and on-topic the system's responses are with the user's responses ensuring they contribute meaningfully to the dialogue [32]. ...
The integration of conversational agents into our daily lives has become increasingly common, yet many of these agents cannot engage in deep interactions with humans. Despite this, there is a noticeable shortage of datasets that capture multimodal information from human-robot interaction dialogues. To address this gap, we have developed a Personal Emotional Robotic Conversational sYstem (PERCY) and recorded a novel multimodal dataset that encompasses rich embodied interaction data. The process involved asking participants to complete a questionnaire and gathering their profiles on ten topics, such as hobbies and favourite music. Subsequently, we initiated conversations between the robot and the participants, leveraging GPT-4 to generate contextually appropriate responses based on the participant's profile and emotional state, as determined by facial expression recognition and sentiment analysis. Automatic and user evaluations were conducted to assess the overall quality of the collected data. The results of both evaluations indicated a high level of naturalness, engagement, fluency, consistency, and relevance in the conversation, as well as the robot's ability to provide empathetic responses. It is worth noting that the dataset is derived from genuine interactions with the robot, involving participants who provided personal information and conveyed actual emotions.
... Speech rate, calculated as the number of syllables produced per minute of total speaking time including pauses, provides a comprehensive overview of fluency by accounting for both production speed and pausing behavior (Kormos & Dé nes, 2004). Articulation rate, measured as the number of syllables produced per minute of phonation time excluding pauses, offers a more focused measure of speech production speed (Tavakoli, 2016). These measures were selected based on their demonstrated strong predictive power for perceived fluency in previous studies. ...
This study examines the effects of online task planning conditions on L2 learners' utterance fluency and self-perceived fluency among Chinese EFL learners. Ninety participants were randomly assigned to Pressured Online Planning (POP), Unpressured Online Planning (UOP), and Hybrid Online Planning (HOP) groups. Participants completed a narrative task based on a Mr. Bean video clip under their respective planning conditions. Utterance fluency was measured using temporal and linguistic indicators, while self-perceived fluency was assessed through CEFR self-assessment and an Analytic Fluency Perception Scale. Results indicate that the hybrid condition yielded significantly higher speech and articulation rates, with fewer disfluencies compared to other conditions (p < .001). Notably, self-perceived fluency did not consistently align with objective measures or rater evaluations, particularly in the pressured condition. The study reveals complex relationships between task planning conditions, objective fluency measures, and learners' self-perceptions, contributing to our understanding of L2 speech production processes and informing task-based language teaching practices.
... Because conversation is coconstructed (Carragher et al., 2023;Clark, 1996;Goodwin, 1979;Hengst, 2020), it seeks to minimize joint effort and increase efficiency by reducing the complexity of referring expressions (Clark & Wilkes-Gibb, 1986). This provides many opportunities for a speaker to use elliptical speech that takes advantage of shared knowledge and information in prior utterances (e.g., Speaker 1: Where did you eat? Speaker 2: In the 7th floor caf), and this might render an overestimate of fluency (Tavakoli, 2016). Monologic narrative, on the other hand, better reflects a speaker's ability to construct and connect utterances independently, just as many other language assessment tasks aim to remove cues from the conversation partner and the context. ...
Purpose
Assessing fluency in aphasia is diagnostically important for determining aphasia type and severity and therapeutically important for determining appropriate treatment targets. However, wide variability in the measures and criteria used to assess fluency, as revealed by a recent survey of clinicians (Gordon & Clough, 2022), results in poor reliability. Furthermore, poor specificity in many fluency measures makes it difficult to identify the underlying impairments. Here, we introduce the Flu-ID Aphasia, an evidence-based tool that provides a more informative method of assessing fluency by capturing the range of behaviors that can affect the flow of speech in aphasia.
Method
The development of the Flu-ID was based on prior evidence about factors underlying fluency (Clough & Gordon, 2020; Gordon & Clough, 2020) and clinical perceptions about the measurement of fluency (Gordon & Clough, 2022). Clinical utility is maximized by automated counting of fluency behaviors in an Excel template. Reliability is maximized by outlining thorough guidelines for transcription and coding. Eighteen narrative samples representing a range of fluency were coded independently by the authors to examine the Flu-ID's utility, reliability, and validity.
Results
Overall reliability was very good, with point-to-point agreement of 86% between coders. Ten of the 12 dimensions showed good to excellent reliability. Validity analyses indicated that Flu-ID scores were similar to clinician ratings on some dimensions, but differed on others. Possible reasons and implications of the discrepancies are discussed, along with opportunities for improvement.
Conclusions
The Flu-ID assesses fluency in aphasia using a consistent and comprehensive set of measures and semi-automated procedures to generate individual fluency profiles. The profiles generated in the current study illustrate how similar ratings of fluency can arise from different underlying impairments. Supplemental materials include an analysis template, extensive guidelines for transcription and coding, a completed sample, and a quick reference guide.
Supplemental Material
https://doi.org/10.23641/asha.27078199
... Subsequently, it allows a more sophisticated and accurate language to be obtained, ultimately consolidating L2 knowledge for enhanced fluency and performance. These objective measures have gained prominence in contemporary L2 acquisition research (e.g., de Jong & Mora, 2019;Tavakoli, 2016;Tavakoli, Campbell, & McCormack, 2016;Tavakoli, Nakatsuhara, & Hunter, 2017). ...
This study examined the effects of synchronous and asynchronous online communication using video conferences and video letters (VLs) on the development of foreign language speaking skills for Common European Framework of Reference for Languages (CEFR) A1 and B1 level learners of English. Two groups, one using the VLs and the other video conferencing (ZOOM), were constituted to compare the impact of the modes. Both groups were provided with the same topic for interaction with native English-speaking university students from the United States. Pre- and post-tests were conducted to examine the effects of each practice mode. The transcribed speaking tests were analyzed using the Complexity, Accuracy, and Fluency (CAF) framework (Housen et al., 2012). The post-test results revealed that in the ZOOM group, sentence complexity score and total number of words decreased significantly. The VL group showed no differences in word complexity and maintained sentence complexity. This study highlighted the influence of synchronous and asynchronous computer-mediated communication (CMC) approaches on speaking skills.
... Second, there is a reasonable degree of control of the incoming message. In other words, the elicited data are relatively comparable and predictable (Tavakoli 2016). An additional argument for the choice of the monologue task in the present research is associated with experiences of LA, which depend on, among others, the perception of the interlocutor (Szyszka 2017). ...
... The problem lies partly in measuring the fluency per se. Previous studies that examined the link between L2 fluency and L2 predictive gender processing used fluency tests that varied from perception test, receptive vocabulary to more standardized tests (see Segalowitz, 2016 andTavakoli, 2016 regarding the difficulty in measuring L2 fluency). ...
Previous studies on the use of morphosyntactic gender cues for linguistic prediction show that non-native speakers' use of grammatical gender information is influenced by various factors. In the present study, we examined the influence of differential cross-linguistic influence (DCLI), knowledge of L2 lexical gender, gender congruency, and L2 fluency. To this end, we investigated L1 Oromo L2 Amharic speakers as well as L1 Amharic speakers, using the Visual World Paradigm (VWP) and supplementary offline experiments. We investigated two groups of L2 Amharic speakers, i.e., L1 Eastern Oromo L2 Amharic and L1 Western Oromo L2 Amharic speakers. The Eastern Oromo dialect patterns with Amharic in terms of gender agreement unlike the Western Oromo dialect which does not have grammatical gender. Analyses of the participants' proportion of eye fixations show that early exposure to the gendered Eastern Oromo dialect facilitates predictive L2 gender processing. L2 fluency, the speakers' knowledge of L2 lexical gender and specific properties of the gender cues modulate predictive L2 gender processing. However, there is no significant influence of lexical gender congruency. The study has ecological significance as it presents empirical data from understudied languages.
... Based on this measure, the number of syllables, silent pauses, and speech time were calculated by the author and were inspected manually by one trained speech analyst for the calculations of articulation rate, phonation-time ratio, mean length of runs, and the number and duration of silent pauses. Previous research cautioned against the use of "internally related" or overlapping measures of fluency in making conclusions about utterance fluency (Bosker et al., 2013;Kormos, 2006;Suzuki et al., 2021;Tavakoli, 2016; but see De Jong, 2018). Therefore, a combination of speed and breakdown fluency measures operationalized following the established indices were used in the present study (De Jong & Wempe, 2009;Tavakoli & Wright, 2020, pp. ...
The purpose of the present study was to examine (a) the effects of audio-visual information and (b) the role of temporal measurements associated with the perceptual ratings of fluency, accentedness, comprehensibility, pronunciation, and oral proficiency in second language (L2) speech samples of International Teaching Assistants (ITAs). American undergraduate students randomly assigned to audiovisual and audio-only (without seeing the speaker) conditions rated the two-minute speech samples of five ITAs on a seven-point scale. For acoustic analysis, several temporal measurements were used that were hypothesized to have an association with the ratings. Findings revealed that listeners without visual information judged L2 speech as less fluent and more accented although no significant difference in comprehensibility was found. L2 speakers were also judged to have better pronunciation and higher oral proficiency when the recordings were presented in video mode. The findings further indicated that among the acoustic measures examined, articulation rate was found to be associated with perceptual ratings. Implications for ITA training and language testing and assessment are discussed.
... Firstly, the way in which anxiety was manipulated may include a confound: the external teacher minimized the amount of interaction in the high anxiety condition, whereas the peer in the low anxiety condition collaborated in interaction. It could be that participants are more fluent when the setting is more like a dialogue (see Tavakoli, 2016), resulting in the confounding variable (dialogue). However, the significant and medium to strong correlation between the difference value between the two conditions for silent pauses per second with the difference value for anxiety cannot be attributed to confounding variables and is additional evidence that anxiety is related to fluency. ...
Although the debilitative effect of foreign language anxiety (FLA) on second-language (L2) performance and L2 speaking has repeatedly been shown, it is unclear in what way FLA affects L2 fluency and the speech production processes that lead to (dis)fluency. The current study investigates the effect of FLA on the utterance fluency of Dutch learners of L2 English and their (recalled) speech production processes (cognitive fluency). Using an experiment with a within-subjects design, 22 L2 learners performed two speaking tasks in a high and low anxiety condition. Their L2 utterance fluency was explored quantitatively by calculating temporal measures, whereas cognitive fluency was explored qualitatively using stimulated recalls, for a subset of seven students. Additionally, participants reported on their experiences during the speaking tasks. It was found that in general, they felt more anxious in the high anxiety condition and judged their performance as lower compared to the low anxiety condition. Of all fluency indices, results showed an effect for filled pauses only. Additionally, heightened anxiety was significantly related to more use of silent pauses. Finally, the stimulated recalls revealed a difference in the number and quality of processing issues between the anxiety conditions, as participants reported more issues related to the content of the message in the high-anxiety condition compared to the low-anxiety condition. This finding contrasts with predominant hypotheses, which state that anxiety primarily impacts linguistic formulation. Thus, this finding calls for experimental research with different ways to manipulate anxiety within participants to further investigate how anxiety impacts cognitive fluency.
... Regarding the latter, van Os et al. (2020) examined perceptions of fluency in dialogic speech and demonstrated that experimentally manipulated turn-taking behaviors had an influence on how raters judged fluency. It is also the case that studies comparing utterance fluency in monologic versus dialogic tasks often indicate higher fluency in dialogues (Sato, 2014;Tavakoli, 2016). In summary, following the line of argumentation in Segalowitz (2016) and what we know about how fluency might differ in monologic versus dialogic contexts, it is necessary to expand the investigation of midclause pausing beyond monologic tasks. ...
This study explored the effect of speaking task on midclause pausing characteristics in the L1 and L2 speech of the same speakers to gain further insights into the potential relationship between pause location and stages of speech production. Participants included English L1 learners of L2 French ( n = 29) or Spanish ( n = 27) from the publicly available, longitudinal LANGSNAP corpus. Participants completed two oral tasks in their L1 and L2: a picture-based narrative and a semistructured interview. The rate, duration, and proportion of midclause pauses were compared between tasks in the L1 as well as in the L2 before and during residence abroad. In the L1, results indicated more fluent performance in the narrative task except for rate. When speaking in their L2, participants showed improvement on each measure in the narrative task but ultimately remained less fluent in their L2 in comparison to their L1. In the interview task, the only measure of midclause pausing that consistently differentiated L1 from L2 speech was midclause pause rate. The findings call for a nuanced interpretation of connections between midclause pausing and formulation and suggest that midclause pause rate is least influenced by speaking task.
... Utterance fluency measures are crucial for assessing L2 speaking proficiency (Bosker et al., 2013;Tavakoli, 2016). ...
Unraveling how learners’ L2 fluency performance relates to their L1 is crucial for promoting L2 speaking assessment. Despite the consensus that learners’ L1 and L2 are correlated in several measures of utterance fluency, there have been variations in the correlation strength resulting from methodological differences. The current meta-analysis aims to present an overall picture of the L1–L2 fluency relationship and investigate the moderating effect of methodological factors. We identified 16 studies with 137 effect sizes and 714 participants. Through a multi-level meta-analysis, we calculated the aggregate effect sizes of the relationship between L1 and L2 fluency and examined the potential moderating effect of five learner and task variables. The overall correlation was strongest in L1 and L2 breakdown fluency, followed by speed and composite fluency, and weakest in repair fluency. Based on the results of the moderator analyses, we assume that the relationship between L1 and L2 composite fluency may be affected by L2 learning context, task type, and task consistency. The findings have implications for future research design and L2 fluency assessment.
... That is, a rate that is too high or too low may work to undermine listener comprehension. Further exploration of this relationship across a wide variety of L1 backgrounds and proficiency levels has yet to confirm the acquisition of speech rate(s) that optimize intelligibility, especially considering the differences in speech rate found in L2 productions of dialogic vs. monologic speech (Tavakoli, 2016) and in relation to task familiarity (Lambert et al., 2017). ...
... The two directions of research above predominantly employed monologic tasks to elicit speech samples with very few studies using dialogic tasks or comparing both kinds of tasks (e.g., Tavakoli 2016). Monologic tasks involve topic narratives and retelling monologues (Riazantseva 2001). ...
Utterance fluency is widely acknowledged to be three-dimensional (i.e., speed, breakdown and repair), but little direct evidence has been offered in support of its dimensionality. This study examines the factor structures of speed and breakdown fluency and their relationships with EFL proficiency by evaluating 162 EFL learners' retelling performances with respect to commonly used speed measures as well as pause length and frequency measures. Findings from structural equation modeling show that speed fluency was represented by articulation rate and mean length of utterance, and breakdown fluency by mid-clause pause length and frequency as well as end-of-clause pause length. Although speed and breakdown fluency were strongly related, EFL proficiency had a direct effect on speed fluency but not on breakdown fluency. When articulation rate was replaced with speech rate as the sole measure of speed fluency, the factor structure of breakdown fluency was changed and EFL proficiency contributed a little more to speed fluency. These findings help to better understand how EFL learners make pauses and how measures of speed and breakdown fluency differ from but simultaneously relate to each other.
... Based on this measure, the number of syllables, silent pauses, and speech time were calculated by the author and were inspected manually by one trained speech analyst for the calculations of articulation rate, phonation-time ratio, mean length of runs, and the number and duration of silent pauses. Previous research cautioned against the use of "internally related" or overlapping measures of fluency in making conclusions about utterance fluency (Bosker et al., 2013;Kormos, 2006;Suzuki et al., 2021;Tavakoli, 2016; but see De Jong, 2018). Therefore, a combination of speed and breakdown fluency measures operationalized following the established indices were used in the present study (De Jong & Wempe, 2009;Tavakoli & Wright, 2020, pp. ...
The purpose of the present study was to examine (a) the effects of audio-visual information and (b) the role of temporal measurements associated with the perceptual ratings of fluency, accentedness, comprehensibility, pronunciation, and oral proficiency in second language (L2) speech samples of International Teaching Assistants (ITAs). American undergraduate students randomly assigned to audiovisual and audio-only (without seeing the speaker) conditions rated the two-minute speech samples of five ITAs on a seven-point scale. For acoustic analysis, several temporal measurements were used that were hypothesized to have an association with the ratings. Findings revealed that listeners without visual information judged L2 speech as less fluent and more accented although no significant difference in comprehensibility was found. L2 speakers were also judged to have better pronunciation and higher oral proficiency when the recordings were presented in video mode. The findings further indicated that among the acoustic measures examined, articulation rate was found to be associated with perceptual ratings. Implications for ITA training and language testing and assessment are discussed.
... The question elicited a mean length of 2.5 min of speech from the participants. Previous research has typically collected 1-3 min of monologue narratives for temporal measures of oral fluency by L2 learners (Tavakoli 2016) and thus mean length of learner monologues in the current study provides the basis of an acceptable measure of temporal fluency. ...
The current mixed-method study investigated two groups of Korean-speaking short-term sojourners in Australia. One group (students) was composed of learners enrolled in English training programs, whereas the other group (workers) was of learners in the workplace. We administered questionnaires and a semi-structured interview to examine their willingness to communicate (WTC) in English as their second language (L2) and explored the relationship between this variable and the sojourners’ amount of L2 contact and their oral fluency in English. Our quantitative analyses show that the student group showed a higher level of WTC and amount of L2 exposure than the worker group. For both groups, WTC significantly predicted sojourners’ amount of L2 exposure. However, oral fluency was found neither to be associated with WTC nor with the amount of L2 exposure. Qualitative theme-based analysis suggests that the two sojourn groups demonstrated similarities and differences in their attitudes and motivations related to WTC and unwillingness to communicate (unWTC). The students demonstrated a stronger tendency to engage in L2 interaction than the workers, aligning with their significantly higher frequency of reported L2 exposure. The workers’ attitudes were characterized by feelings of ambivalence, with co-existence of both WTC and unWTC.
... McCarthy (2010) introduced the concept "confluence" to describe the co-constructed nature of a dialogue. The collaborative nature may be demonstrated by interruption, overlap, or long pauses, measured in terms of duration and frequency (Tavakoli, 2016;van Os et al., 2020). Peltonen (2017) proposed the concept of "dialogue fluency". ...
... Whereas participants in both Towell et al.'s and Huensch and Tracy-Ventura's studies completed a story retelling task, our participants were engaged in a semi-guided interview. Relevant evidence for a task effect comes from Tavakoli (2016), who compared productions by the same speakers engaged in monologic and dialogic tasks. The author found that the mean length of pauses was statistically shorter in the dialogues, but that there was little difference concerning number or location of the pauses. ...
In a recent publication, Foster (2020) identified several promising directions for second language (L2) oral fluency research, including explicit attention to the parallel development of fluency and productive vocabulary knowledge. The present paper focuses on this issue and proposes an analysis of longitudinal data collected from five Francophone learners of L2 English before and after a nine-month stay in an Anglophone environment. The analysis focuses on speed fluency and breakdown fluency as well as on both free and controlled measures of productive vocabulary knowledge. Results point to clear evolution in speed fluency for all five participants and different individual trajectories for breakdown fluency after a stay abroad. Change with respect to productive vocabulary knowledge, as measured by three measures, was mixed. Findings are interpreted in light of previous research. In particular, we argue that the current results do not provide strong evidence for a clear relationship between productive vocabulary knowledge and oral fluency development.
The Complexity, Accuracy, and Fluency (CAF) model has established itself as a crucial framework for examining the multifaceted nature of second language (L2) performance, providing valuable insights into learners' linguistic development. This paper presents a comprehensive overview of the CAF model, exploring its core constructs—complexity, accuracy, and fluency. It explores their definitions, the challenges associated with each construct, and the diverse approaches employed to measure them. The analysis includes a review of current research, identifying commonly used metrics, and outlining potential avenues for future inquiry. By offering a critical evaluation of the CAF framework and its measurement techniques, this paper sets out to enhance the understanding of L2 assessment and contribute to the advancement of research in this domain.
Word searches (WSs) have been widely studied in L1 and L2 interaction but have rarely been combined with fluency research. The present study bridges the gap between these research fields by examining WSs and their connections to fluency in L1 and L2 dialogue data. From a larger project, 50 participants’ dialogues (25 L1 Finnish, 25 L2 English) were analyzed for WSs. The mixed-methods analyses focused on potential differences and connections in WS use across the L1 and L2 dialogues and the connections between WSs and fluency among selected participants. The results demonstrated differences especially in the use of solitary WSs across L1 and L2 and some connections between L1 and L2 WSs among individual participants, despite group-level connections not being statistically significant. The qualitative analyses illustrated the potential fluency-enhancing functions of collaborative WSs. The study has implications for L2 speech fluency research and L2 teaching.
Studies on on-line planning have been based upon the same definition in the past few decades, and they have operationalized on-line planning in similar ways. Drawing upon theories in cognitive psychology, the present study sets out to propose a revised definition of on-line planning that aims at capturing a greater variety of on-line planning behaviors. Two methods of eliciting on-line planning were implemented: providing students with listener’s feedback between two consecutive rounds of task performance, and providing them with criteria for successful task performance. Their effects on second language (L2) narrative task performance and the use of planning strategies were investigated with 90 Chinese learners of English as a foreign language (EFL) who were first-year English majors. To assess the quality of task performance, complexity, accuracy and fluency (CAF) measures and a measure of story completeness were employed. Post-task interviews were conducted to explore planning strategies. Results from the quantitative and the qualitative analyses suggest that utilizing listeners’ feedback successfully elicited an increased number of on-line planning activities. Participants under this on-line planning condition also produced significantly more complex and accurate output than the non-intervention group.
Study abroad research has become an established area of inquiry with theoretical impact and methodological sophistication. The field has incorporated the different approaches and methodological changes that have characterized SLA scholarship, including technological advances and new designs. The present volume contributes an update on and a systematic critical appraisal of the methods employed in study abroad research to identify strengths and weaknesses and to look ahead and point towards new directions. The volume is organized around different areas -approaches, instruments, linguistic levels, and learners and their context-, each one including a number of chapters authored by outstanding experts in the field.
Utterance fluency is widely acknowledged to be three-dimensional (i.e., speed, breakdown and repair), but little direct evidence has been offered in support of its dimensionality. This study examines the factor structures of speed and breakdown fluency and their relationships with EFL proficiency by evaluating 162 EFL learners’ retelling performances with respect to commonly used speed measures as well as pause length and frequency measures. Findings from structural equation modeling show that speed fluency was represented by articulation rate and mean length of utterance, and breakdown fluency by mid-clause pause length and frequency as well as end-of-clause pause length. Although speed and breakdown fluency were strongly related, EFL proficiency had a direct effect on speed fluency but not on breakdown fluency. When articulation rate was replaced with speech rate as the sole measure of speed fluency, the factor structure of breakdown fluency was changed and EFL proficiency contributed a little more to speed fluency. These findings help to better understand how EFL learners make pauses and how measures of speed and breakdown fluency differ from but simultaneously relate to each other.
Silent pauses may be defined as stretches of silence within the speech stream (Cucchiarini et al., 2002) or an unvoiced delay, a temporary suspension of speech activity (Clark, 2006: 244). Historically, interest in silent pauses (SPs) predates that in other markers of disfluency, starting in the 1940s and becoming a notable area of investigation with Goldman-Eisler’s publications in the 1950s and 1960s. Thereafter, researchers expanded interest in SP location, frequency, and length. It was early recognised that the unmarked position of SPs was before a finite clause and their marked, i.e. disfluent position mid-clause. More frequent pauses indicated problems with speech production, as did their duration. As well as indicating production problems, SPs were acknowledged to serve a multitude of functions, from physiological, e.g. breathing, to psychosocial, e.g. speaker affect.
This volume illustrates the high potential of learner corpus investigations for research into the CAF triad by presenting eleven original learner corpus-based studies which are set within solid theoretical frameworks, examine learner corpora with state-of-the-art analytical techniques and yield highly interesting findings. The volume’s major strength lies in the range of issues it undertakes and in its interdisciplinary thematic novelty. The chapters collectively address all three dimensions of L2 performance related to different linguistic subsystems (i.e. lexical, phraseological and grammatical complexity and accuracy, along with fluency) as well as the interactions among these constructs. The studies are based on data drawn from carefully compiled learner corpora which are analysed with the help of diverse corpus-based methods. The theoretical discussions and the empirical results shall contribute to the advancement of the fields of SLA and writing and speech research and shall inspire further investigations in the area of the CAF triad.
The current study investigated disfluency and self‐repair in L2 presentational and interpersonal speech to determine whether speech modes may affect L2 self‐repair. Knowledge of self‐initiated repair behavior in L2 speech is particularly relevant for understanding psycholinguistic processes of L2 speech production. Different speech modes, such as presentational and interpersonal speech, may provide distinct psychological and discourse conditions for eliciting speech. Although presentational and interpersonal communicative skills are deemed crucial subskills of L2 ability, little research has examined whether patterns of L2 disfluency and self‐repair may vary between the two speech modes. The current study examined the distributions and rates of different types of disfluency signals, reparanda, and repair strategies manifested in presentational and interpersonal speech by advanced L2 Chinese speakers. The results revealed that in both speech modes, advanced Chinese speakers allocated significant repairing effort to correct lexical items and reformulate messages. The distributions and rates of different types of self‐repair exhibited significant between‐mode differences. The findings indicate that speaking practice through different modes may facilitate the development of L2 speaking skills in different areas. Further research is still needed to examine disfluency and self‐repair in various speech modes and with different L2 proficiency levels and task conditions. Language instructors design task‐based activities to help their students develop speaking skills, but what about the different modalities? How do modes of speaking affect the speech that learners produce? This article demonstrates how presentational and interpersonal speaking may affect L2 learners’ disfluency and self‐repair patterns.
Inspired by Murray Munro and Tracey Derwing’s 1995 seminal study of intelligibility, comprehensibility, and accentedness, this book revisits the insights of their original research and presents subsequent studies extending this work to new ways of understanding second language speech. By rejecting the nativeness approach upon which previous pronunciation research and teaching were built, Munro and Derwing’s paper became the catalyst for a new paradigm of pronunciation and speech research and teaching. For the first time, pronunciation researchers had an empirically-motivated set of dimensions for assessing L2 speech. Results of many subsequent studies showed that the original insights of three partially-independent measures are indispensable to language teaching, language assessment, social evaluations of speech, and pedagogical priorities. This monograph offers 9 diverse chapters by leading researchers, all of which focus on intelligibility and or comprehensibility. This volume is essential reading for anyone interested in up-to-date coverage of L2 pronunciation matters. Originally published as special issue of Journal of Second Language Pronunciation 6:3 (2020)
In this corpus-based study we explore three measurements of L2 fluency – articulation rate, filler particles, and pauses –, both within and between two registers of spontaneous dialogues spoken by Polish learners of German. The measurements are assessed both in toto (as calculated over the whole dialogue) and in parte (as calculated for specific sections). The sections are identified on a quantitative tier that divides the dialogue into four parts, and qualitatively on two linguistically-informed tiers, comprising sections based on dialogue move and task. We challenge the assessment of fluency as an average measurement over the entire dialogue, showing that a sectionwise analysis offers a better understanding of similarities and differences both within and between the two registers.
One ubiquitous feature of unscripted speech across languages is the use of filled pauses and fillers. However, research has hitherto yielded opposing or ambiguous results about whether filled pauses and fillers in spontaneous speech aid or hinder L2 listeners' comprehension. In the current study, 30 L1 Japanese participants of three English proficiency levels viewed two video texts spoken by two different speakers of General American English. The video texts contained four occurrences of the filled pause um and two occurrences of the filler like. Participants had multiple opportunities to view and comprehend the texts. Comprehension, decoding, and internal processing of the video texts were elicited through a three-task assessment procedure of L1 recalls, L2 repetitions, and verbal reports. Results revealed that some occurrences of um and like caused comprehension difficulties for most participants while other occurrences caused no noticeable difficulty for any participants. The findings suggest that the location of filled pauses and fillers, L2 listeners’ language proficiency, and multiple other factors can determine whether or not filled pauses and fillers will affect L2 listening comprehension of spontaneous speech.
This study investigates the potential of VoiceThread (VT) as a learning tool in order to enhance learners’ performance and confidence, and to provide the basis for continuing independent practice. The participants were students from culturally and linguistically diverse backgrounds on a 10-week English for Academic Purposes (EAP) programme in academic speaking and listening skills in an Australian university. Students carried out a weekly extended speaking task in the form of an academic presentation, on a theme/topic related to their own interests. To prepare for this they used VT to record and rehearse their presentation, uploading their recordings to the Learning Management System for other students to listen to, and for the teacher to provide individual feedback. The recordings were analysed in relation to a number of features to measure utterance fluency, including speed, repairs and pauses. Questionnaires were used to investigate students’ perceptions of the value of the task, of their performance on the task, and of changes in their performance and confidence over time. Interviews were used to investigate the potential of recording and rehearsal for long term independent practice. The study involved a total of 34 students in 10 week courses, 22 students using the Voicethread software, and 12 students in a comparison class following the same curriculum, but without the Voicethread software. The results suggest that, over time, regular VT tasks help students develop their performance and confidence in spoken English. Such voice-recording activities have the potential to improve oral performance by encouraging independent practice of oral activities, and so tackling the limitations on time, practice and feedback inherent in classroom-based courses.
Supplemental data for this article is available online at https://doi.org/10.1080/09588221.2021.2003407 .
Keywords: oral fluency, task types, motivation This study was conducted to investigate the impacts of motivation and task types on the development of L2 oral fluency in higher education in Vietnam. The quantitative method was employed to achieve this goal. The participants were thirteen teachers and thirty second-year students at the university. Questionnaires with close items for students and instructors and semi-structured interview questions for instructors were employed as data collection instruments. The results suggested that teachers consider that a student would be successful with proper motivation and the task success is just being motivated. In addition, the findings reveal that performance is, in general, statistically more fluent in dialogue but also indicate that performances in the two modes are not different. This article is a part of my dissertation project, which examines teachers and students' perspectives on the factors influencing the oral fluency of L2 students in universities in Vietnam.
Keywords: oral fluency, task types, motivation This study was conducted to investigate the impacts of motivation and task types on the development of L2 oral fluency in higher education in Vietnam. The quantitative method was employed to achieve this goal. The participants were thirteen teachers and thirty second-year students at the university. Questionnaires with close items for students and instructors and semi-structured interview questions for instructors were employed as data collection instruments. The results suggested that teachers consider that a student would be successful with proper motivation and the task success is just being motivated. In addition, the findings reveal that performance is, in general, statistically more fluent in dialogue but also indicate that performances in the two modes are not different. This article is a part of my dissertation project, which examines teachers and students' perspectives on the factors influencing the oral fluency of L2 students in universities in Vietnam.
This article introduces three important challenges and possible solutions when using spoken dialogue to measure the use of specific multi-word expressions. The first challenge is deciding whether to count precise and accurate use of target expressions only or whether to extend the count to include variation. The second challenge requires addressing the indirect nature of dialogue as a testing method. The third challenge is organizing data and preparing ways to clearly identify speakers within the dialogue. These challenges are illustrated with examples and potential solutions from my recent research investigating spoken use of multi-word expressions.
This article provides an empirically grounded account of what happens when more persons than one talk at once in conversation. It undertakes to specify when such occurrences are problematic for the participants, and for the organization of interaction; what the features of such overlapping talk are; and what constraints an account of overlapping talk should meet. It describes the practices employed by participants to deal with such simultaneous talk, and how they form an organization of practices which is related to the turn-taking organization previously described by Sacks et al. 1974. This constitutes a previously unexplicated component of that turn-taking organization, and one that provides solutions to underspecified features of the previous account.
A key mechanism in the organization of turns at talk in conversation is the ability to anticipate or PROJECT the moment of completion of a current speaker's turn. Some authors suggest that this is achieved via lexicosyntactic cues, while others argue that projection is based on intonational contours. We tested these hypotheses in an on-line experiment, manipulating the presence of symbolic (lexicosyntactic) content and intonational contour of utterances recorded in natural conversations. When hearing the original recordings, subjects can anticipate turn endings with the same degree of accuracy attested in real conversation. With intonational contour entirely removed (leaving intact words and syntax, with a completely flat pitch), there is no change in subjects' accuracy of end-of-turn projection. But in the opposite case (with original intonational contour intact, but with no recognizable words), subjects' performance deteriorates significantly. These results establish that the symbolic (i.e. lexicosyntactic) content of an utterance is necessary (and possibly sufficient) for projecting the moment of its completion, and thus for regulating conversational turn-taking. By contrast, and perhaps surprisingly, intonational contour is neither necessary nor sufficient for end-of-turn projection.
This report extends earlier context-free
treatments of turn-taking for conversation by describing the
context-sensitive operation of the principal forms of addressing
employed by current speakers to select next speakers. It first
describes the context-specific limitations of gaze-directional
addressing, and the selective deployment and more-than-addressing
action regularly accomplished by address terms (most centrally,
names). In addition to these explicit methods of addressing,
this report introduces tacit forms of addressing that call on
the innumerable context-specific particulars of circumstance,
content, and composition to select a next speaker.
This article is organised in five main sections. First, the sub-area of task-based instruction is introduced and contextualised. Its origins within communicative language teaching and second language acquisition research are sketched, and the notion of a task in language learning is defined. There is also brief coverage of the different and sometimes contrasting groups who are interested in the use of tasks. The second section surveys research into tasks, covering the different perspectives (interactional, cognitive) which have been influential. Then a third section explores how performance on tasks has been measured, generally in terms of how complex the language used is, how accurate it is, and how fluent. There is also discussion of approaches to measuring interaction. A fourth section explores the pedagogic and interventionist dimension of the use of tasks. The article concludes with a survey of the various critiques of tasks that have been made in recent years.
This study investigated how individual differences in linguistic knowledge and processing skills relate to individual differences in speaking fluency. Speakers of Dutch as a second language (N = 179) performed eight speaking tasks, from which several measures of fluency were derived such as measures for pausing, repairing, and speed (mean syllable duration). In addition, participants performed separate tasks, designed to gauge individuals' second language linguistic knowledge and linguistic processing speed. The results showed that the linguistic skills were most strongly related to average syllable duration, of which 50% of individual variance was explained; in contrast, average pausing duration was only weakly related to linguistic knowledge and processing skills. People differ with respect to how fluently they speak. Some speak faster than others, some use more filled pauses such as uhs and uhms than others, some use more silent pauses than others, and some use longer silent pauses than others (e.g., Clark & Fox Tree, 2002; Goldman-Eisler, 1968; Shriberg, 1994). These individual differences exist for nonnative speakers and they also exist for native speakers. This raises the question to what extent aspects of fluency in nonnative speech can be considered indicators of second language (L2) proficiency rather than indicators of nonlinguistic factors such as personality characteristics. In this study, we will explore which aspects of L2 fluency relate to L2 linguistic knowledge and processing skills, and to what extent. The term fluency, usually restricted for describing L2 speech, can be used in at least two ways. Lennon (1990) distinguishes a broad definition and a narrow definition. In the broad definition, fluency can be seen as overall (speaking) profi-ciency, whereas fluency in the narrow definition pertains to smoothness and ease of oral linguistic delivery. In this paper, we will use the term fluency in its narrow sense.
ABSTRACT This study examined the componential structure of second-language (L2) speaking proficiency. Participants—181 L2 and 54 native speakers of Dutch—performed eight speaking tasks and six tasks tapping nine linguistic skills. Performance in the speaking tasks was rated on functional adequacy by a panel of judges and formed the dependent variable in subsequent analyses (structural equation modeling). The following independent variables were assessed separately: linguistic knowledge in two tests (vocabulary and grammar); linguistic processing skills (four reaction time measures obtained in three tasks: picture naming, delayed picture naming, and sentence building); and pronunciation skills (speech sounds, word stress, and intonation). All linguistic skills, with the exception of two articulation measures in the delayed picture naming task, were significantly and substantially related to functional adequacy of speaking, explaining 76% of the variance. This provides substantial evidence for a componential view of L2 speaking proficiency that consists of language-knowledge and language-processing components. The componential structure of speaking proficiency was almost identical for the 40% of participants at the lower and the 40% of participants at the higher end of the functional adequacy distribution (n = 73 each), which does not support Higgs and Clifford’s (1982) relative contribution model, predicting that, although L2 learners become more proficient over time, the relative weight of component skills may change.
Highlights the importance of classroom discourse to any second language teacher education programme. Reflective practice is central to teacher education and development, yet is something that many teachers struggle with. Can reflective practice be refocused by asking teachers to place classroom interaction and discourse at the centre of their reflections? In this accessible textbook, Steve Walsh explains why it is essential to put an understanding of classroom discourse at the centre of any second language teacher education programme, whether it is a formal programme under the guidance of a teacher educator or a more informal, self-directed programme of teacher development. He argues that in order to improve their professional practice, language teachers need to gain a detailed, up-close understanding of their local context by focusing on the complex relationship between teacher language, classroom interaction and learning. In order to do this he revisits and reconceptualises the notion of reflective practice by giving teachers appropriate tools which allow them to reflect on and improve their professional practice. This thought-provoking book not only stimulates debate on classroom discourse and reflective practice, but also contains practical exercises and advice which will be invaluable to both new and experienced language teachers as well as to researchers in applied linguistics.Task commentaries, a glossary of technical terms and an annotated list of further reading are also included.
discourse research;information;encoding;interpretive categories;semantics
This study investigates the effects of a short-term pedagogic intervention on development of second language (L2) fluency among learners studying English for academic purposes at a UK university. It also examines the interaction between development of fluency and complexity and accuracy. Through a pretest and posttest design, data were collected over a period of 4 weeks from learners performing monologic tasks. While the control group (CG) focused on developing general speaking and listening skills, the experimental group (EG) received awareness-raising activities and fluency strategy training in addition to general speaking and listening practice. The data, coded in terms of a range of measures of fluency, accuracy, and complexity, were subjected to repeated-measures multivariate analysis of variance (MANOVA), t-tests, and correlations. The results indicate that after the intervention, although some fluency gains were achieved by the CG, the EG produced statistically more fluent language, demonstrating a faster speech and articulation rate, longer runs, and higher phonation time ratios. The significant correlations obtained between measures of accuracy and learners’ pauses in the CG suggest that pausing opportunities may have been linked to accuracy. The findings have significant implications for L2 pedagogy, highlighting the effective impact of instruction on development of fluency.
This paper reports on a comparative study of pauses made by L2 learners and native speakers of English while narrating picture
stories. The comparison is based on the number of pauses and total amount of silence in the middle and at the end of clauses
in the performance of 40 native speakers and 40 L2 learners of English.1 The results of the quantitative analyses suggest that, although the L2 learners generally pause more repeatedly and have
longer periods of silence than the native speakers, the distinctive feature of their pausing pattern is that they pause frequently
in the middle of clauses rather than at the end. The qualitative analysis of the data suggests that some of the L2 learners’
mid-clause pauses are associated with processes such as replacement, reformulation, and online planning. Formulaic sequences,
however, contain very few pauses and therefore appear to facilitate the learners’ fluency.
Although fluency constitutes an essential component of second language (L2) proficiency, there are mixed results and gaps in the literature on how L2 speakers’ fluency differs from fluent speech production in a first language (L1). The research reported in this article investigated utterance fluency and cognitive fluency of L1 English speakers and Korean learners of L2 English by eliciting and comparing quantitative evidence from temporal measures and qualitative evidence from stimulated recall responses. In addition, the L2 speaker group's proficiency was measured by an in-house institutional test so as to inspect how L2 fluency measures correlated to varying proficiency. The L1 and L2 speakers were different in speed, length of run, and silent pauses. In particular, a striking group difference in silent pause rate within a clause was found, consistent with the claim that pauses within clauses reflect processing difficulties in speech production. Different qualitative patterns in the stimulated recall responses by the lower and higher proficiency learners are discussed in relation to Ullman's declarative/procedural model and Segalowitz's fluency vulnerability points in L2 speech production.
Recent literature in second language (L2) perceived fluency has focused on English as a second language, with a primary reliance on impressions from native-speaker judges, leaving learners' self-perceptions of speech production unexplored. This study investigates the relationship between learners' and judges' perceptions of French fluency under three different task conditions. It also addresses how self-perception of fluency is linked to actual measures of utterance fluency. Participants were 40 adult learners of French at varying levels of proficiency, studying in a university immersion context. The results indicate moderate correlation between L2 learners' and judges' perceptions of fluency and offer evidence to validate self-assessment use in language-learning contexts. Across tasks, two measures of utterance fluency were most strongly related to self-perceptions of fluency: mean length of runs and average pause time. The results also revealed the influence of task complexity on L2 fluent performance and are discussed according to the Cognition Hypothesis (Robinson, 2001, 2003, 2005).
This study investigates the differential effects of two learning contexts, formal instruction (FI) at home and a study abroad period (SA), on the oral production skills of advanced-level Catalan-Spanish undergraduate learners of English. Speech samples elicited through an interview at three data collection times over a 2-year period were quantitatively assessed for fluency, accuracy, and complexity. The results reveal robust fluency gains during SA, moderate improvement in accuracy, and lack of gains in complexity. No gains were obtained during FI. Initial fluency scores, rather than post-FI scores, accounted for a substantial amount of variance in the fluency scores after SA. These findings provide strong evidence for the positive impact of an SA period on the development of oral fluency.
This article
reports on a study that investigated the effect of different amounts of planning time on the
speech performance of L2 speakers. Subjects were 4 groups of learners of German (31 in
total) performing 2 tasks each. The tasks varied in the degree of structure they contained and
the familiarity of information they tapped. The control group had no planning time available;
the 3 experimental groups had 1, 5, and 10 minutes of planning time, respectively, before they
started speaking. Results show fluency and lexical density of speech increase as a function of
planning time. Accuracy of speech improved with only 1 minute planning but did not increase
with more planning time. Complexity of speech was significantly higher for the 10-minute
planning condition only. No significant differences were found for the effect of planning on
the different tasks. This study employed various general and specific constructs for measuring
fluency, complexity, and accuracy of speech. The interrelationships and qualities of these
measures are also investigated and discussed.
An agent's ability to understand an utterance depends upon its ability to relate that utterance to the preceding discourse. The agent must determine whether the utterance begins a new segment of the discourse, completes the current segment, or contributes to it. The intentional structure of the discourse, comprised of discourse segment purposes and their interrelationships, plays a central role in this process (Grosz and Sidner 1986). In this paper, we provide a computational model for recognizing intentional structure and utilizing it in discourse processing. The model is based on the collaborative planning framework of SharedPlans (Grosz and Kraus 1996).
The research reported in this paper concentrates on the contribution of local melodic factors such as pitch accents and boundary tones to the (non)projection of possible turn transition places in a corpus of Dutch task-oriented dialogues. The conclusions drawn are in line with results reported for German and English: syntax seems to be the main projecting factor in the turn-taking mechanism (the developing grammatical structure is used to project upcoming Transition Relevance Places), while local melodic factors play a supporting as well as a constraining role in the process. Melody supports the grammatical structure by signaling incompleteness at positions where a pause does not coincide with a syntactic completion point and by using (low or high) boundary tones where they do coincide. In addition, specific melodic configurations may be used to bridge a syntactic completion point and continue with a further turn.
The present study investigates the role of speech repetition in oral fluency development. Twenty-four students enrolled in English-as-a-second-language classes performed three training sessions in which they recorded three speeches, of 4, 3, and 2 min, respectively. Some students spoke about the same topic three times, whereas others spoke about three different topics. It was found that fluency improved for both groups during training but was maintained on posttests only by the students who repeated their speeches. These students had used more words repeatedly across speeches, most of which were not specifically related to the topic. It is argued that proceduralization of linguistic knowledge represented a change in underlying cognitive mechanisms, resulting in improvements in observable fluency.
Winner of the 31st Annual Kenneth W. Mildenberger Prize, by the Modern Language Association of America, (presented January 7, 2012). Citation: In Cognitive Bases of Second Language Fluency, Norman Segalowitz takes a concept that permeates many teachers’ and students’ understanding of what it means to learn another language—the quest for fluency—and investigates it from multiple vantage points to provide a nuanced, interdisciplinary perspective. Using a dynamical-systems-theory approach, he explores fluency not as an absolute trait that speakers possess but as a nonlinear phenomenon involving interrelated factors: speakers’ first and second language skills and experiences as well as their emotional and cognitive states when speaking; speakers’ interactions within various social and communicative contexts; and listeners’ perceptions. Carefully organized and clearly written, this volume is filled with thoughtful reinterpretations of previous research and new insights on every page that will be of interest to teachers and researchers alike.
The research reported in this paper explores which variables predict native and non-native speaking teachers' perception of fluency and distinguish fluent from non-fluent L2 learners. In addition to traditional measures of the quality of students' output such as accuracy and lexical diversity, we investigated speech samples collected from 16 Hungarian L2 learners at two distinct levels of proficiency with the help of computer technology. The two groups of students were compared and their temporal and linguistic measures were correlated with the fluency scores they received from three experienced native and three non-native speaker teacher judges. The teachers' written comments concerning the students' performance were also taken into consideration. For all the native and non-native teachers, speech rate, the mean length of utterance, phonation time ratio and the number of stressed words produced per minute were the best predictors of fluency scores. However, the raters differed as regards how much importance they attributed to accuracy, lexical diversity and the mean length of pauses. The number of filled and unfilled pauses and other disfluency phenomena were not found to influence perceptions of fluency.
this paper, in textbooks by Allen (1995) and by Jurafsky and Martin (2000), in a survey by Cohen (1996), and in the websites of the Association for Computational Lin2 guistics' Special Interest Group on Discourse and Dialogue (SIGDial) (http://www.sigdial.org) and the Language Engineering Telematics project, MATE (http://mate.nis.sdu.dk/).
The transcription of discourse The handbook of Discourse Analysis
- Jane Edwards
Linguistic skills and speaking fluency in a second language
- De Jong
- Margarita Nivja
- Arjen Steinel
- Florijn
On fluency Individual differences in language ability and language behavior
- Charles Fillmore
Fluency and its teaching. Clevedon: Multilingual Matters
- Maureen Guillot
Fluency The encyclopaedia of applied linguistics
- Pauline Foster
Task-based instruction. Language Teaching
- Peter Skehan
Gender and discourse
- Deborah Tannen
Analysis of silence in interaction
- Dennis Kurzon
The study of fluency and its development in monologue and dialogue
- Giles Witton-Davies
Perceptions of French fluency in second language speech production Canadian Modern Language PrefontaineYvonne Perceptions of French fluency in second language speech productionCanadian Modern Language Review
- Prefontaine
Effects of task complexity and interaction in L2 performance Second Language Task Complexity: Researching the Cognition Hypothesis of Language Learning and Performance
- Marije Michel
Automaticity and attentional skill in fluent performance
- Norman Segalowitz
The ecology and semiotics of language learning: A sociocultural perspective
- Van Lier
- Leo
Exploring utterance and cognitive fluency of and English speakers : Temporal measures and stimulated recall Exploring utterance and cognitive fluency of and English speakers : Temporal measures and stimulated recallLanguage Learning
- Khang
de Holger Mitterer Projecting the end of a speaker s turn : A cognitive cornerstone of conversation de RuiterJanMittererHolgerEnfield Projecting the end of a speaker s turn : A cognitive cornerstone of conversationLanguage
- Ruiter
de Margarita Rob Schoonen Facets of speaking proficiency Studies in Second Language de Facets of speaking proficiencyStudies in Second Language Acquisition
- Jong
Second Language Task Complexity: Researching the Cognition Hypothesis of Language Learning and Performance
- Marije Michel
Working with spoken discourse. London: Sage.Google Scholar
- Debora Cameron
Individual differences in language ability and language behavior
- Charles Fillmore