TBLT AND ATTENTION

One of the strongest arguments in support of TBLT is the theoretical rationale, and corroborating empirical evidence, that well-designed tasks can draw learners’ attention to form within meaning. Long (Reference Long2015) highlights how TBLT provides an ideal middle ground in the balance between form and meaning: Whereas traditional views of language teaching (and classroom learning) encourage a synthetic approach with a focus on forms (i.e., presentation of grammar points that the learner must synthesize for communication), the opposite viewpoint—analytic with a focus on meaning, an extreme view that would include no explanation of grammar—is also not ideal. Long describes TBLT as analytic with focus on form, where the primary focus is on meaning, and form is focused on when need arises during communication (2015, p. 28). Focusing on form within meaning or, even more specifically, within communicative tasks, is a central component of task-based frameworks (Ellis, Reference Ellis2003; Long, Reference Long, Hyltenstam and Pienemann1985, Reference Long1998, Reference Long2015; Long & Norris, Reference Long, Norris and Byram2000; Robinson, Reference Robinson2001, Reference Robinson2011; Skehan, Reference Skehan1998). Robinson (Reference Robinson2011) states, “Tasks provide opportunities for noticing the gap between a participant’s production and input provided and for metalinguistic reflection in the form of output” (p. 2). Ellis (Reference Ellis2003) highlights the role of the researcher and instructor, asserting that, “Task designers should manipulate tasks in such a way as to enhance the probability that language learners will pay attention to particular aspects of the language code in the context of a meaningful activity, because this is believed to strongly promote SLA” (p. 9). Although TBLT is not as concerned with the level of attention or awareness needed (see Leow, Reference Leow1998, Reference Leow2001; Schmidt, Reference Schmidt and Robinson2001; Tomlin & Villa, Reference Tomlin and Villa1994), the belief is that learning is predicated on learners’ attending, at some level, to the target structure(s) while they are already engaged in meaningful task-based interaction (for a discussion of the importance of attention at two levels in TBLT, see Long, Reference Long2015). This relationship between task design and learners’ attention has been supported by considerable empirical research.

For example, studies have manipulated the focus of tasks to make the target form essential within a meaning-based task. One study, Mackey (Reference Mackey1999), which manipulated the type of interaction in which L2 English speakers partook to make the formation of various question forms necessary, was successful in encouraging learners to focus their attention on the linguistic target while completing information gap tasks. Other studies have examined if manipulating the mode of task-based interaction (face-to-face [FTF] as compared to synchronous computer-mediated communication [SCMC]) differentially affects learners’ attention to the linguistic target (e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014; Baralt, Gurzynski-Weiss, & Kim, Reference Baralt, Gurzynski-Weiss, Kim, Sato and Ballinger2016) or perception of feedback (e.g., Gurzynski-Weiss & Baralt, Reference Gurzynski-Weiss and Baralt2014, Reference Gurzynski-Weiss and Baralt2015; Lai & Zhao, Reference Lai and Zhao2006). In Baralt (Reference Baralt, Baralt, Gilabert and Robinson2014), there was an interaction noted in relation to mode and task complexity for the grammatical target, the Spanish imperfect subjunctive: Learners were incapable of focusing on the imperfect subjunctive in the complex SCMC group. No other study found mode to differentially influence learners’ attention to grammatical (e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014; Baralt et al., Reference Baralt, Gurzynski-Weiss, Kim, Sato and Ballinger2016) or lexical (e.g., Gurzynski-Weiss & Baralt, Reference Gurzynski-Weiss and Baralt2014, Reference Gurzynski-Weiss and Baralt2015) targets or to the feedback addressing these linguistic targets. Perhaps one of the most notable strands of TBLT research attempting to manipulate learners’ attention is the work within the cognition hypothesis (Robinson, Reference Robinson2001, Reference Robinson2005). This hypothesis, together with the complementary triadic framework (Robinson & Gilabert, Reference Robinson and Gilabert2007) and SSARC (Stabilize, Simplify, Automatize, Restructure, Complexify) model (Robinson, Reference Robinson2011), outlines a specific formula for directing learners’ attention, either to fluency (along resource-dispersing dimensions) or to accuracy/complexity (along resource-directing dimensions).

Many studies (recently reviewed in Baralt, Gilabert, & Robinson, Reference Baralt, Gilabert and Robinson2014, and Jackson & Suethanapornkul, Reference Jackson and Suethanapornkul2013) have set out to systematically investigate how specific manipulations to individual tasks, and tasks in sequence, may optimally direct learners’ attention to linguistic targets within communicative, meaning-based tasks. Measuring learners’ incidental attention to form has been done most commonly through FFEs (e.g., Loewen, Reference Loewen2005), language-related episodes (LREs; e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014), and/or learners’ modified output following feedback (e.g., Gurzynski-Weiss & Baralt, Reference Gurzynski-Weiss and Baralt2015). FFEs refer to “brief, spontaneous focus on form within a meaning-based interaction” (Loewen, Reference Loewen2005, p. 363). Language-related episodes are another way of operationalizing instances of when learners focus on language during meaning-based tasks. Specifically, LREs occur when “students talk about language they are producing, question their language use, or other- or self-correct their language production” (Swain & Lapkin, Reference Swain, Lapkin, Bygate, Skehan and Swain2001, p. 10). The presence of FFEs or LREs can be considered as instantiations of learner reflection on form or as learning opportunities (e.g., Kim & McDonough, Reference Kim and McDonough2008). Although not highlighted in the current volume, learners’ modified output ^{Footnote 1} is also commonly used as an instantiation of learner reflection on form and is a promising area for future TBLT research on L2 pronunciation. In each case, the presence of FFEs, LREs, and modified output, respectively, is taken as evidence that learners have reflected on form, which, in turn, is seen as a potential task-based learning opportunity.

Empirical findings in this domain indicate that well-designed tasks do indeed encourage learners’ attention to the target structures within meaning-focused tasks (e.g., Kim, Reference Kim2009) and that this attention can relate to L2 development (e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014; Loewen, Reference Loewen2005). However, the target structures that have been examined include grammatical targets such as the imperfect subjunctive in Spanish (e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014; Baralt et al., Reference Baralt, Gurzynski-Weiss, Kim, Sato and Ballinger2016), lexical targets such as locative prepositions and task-specific vocabulary (e.g., Gurzynski-Weiss & Baralt, Reference Gurzynski-Weiss and Baralt2014, Reference Gurzynski-Weiss and Baralt2015), and, most recently, pragmatic targets including requests and suggestions (e.g., Gilabert & Barón, Reference Gilabert, Barón, Mackey and McDonough2013; Kim & Taguchi, Reference Kim and Taguchi2015; Reference Taguchi and KimTaguchi & Kim, in press). To the best of our knowledge, TBLT research has yet to dedicate focus to phonological targets, such as specific aspects of L2 pronunciation, to see if task design may direct learners’ attention to targets beyond grammar, lexis, and pragmatics.

TBLT AND L2 DEVELOPMENT

In TBLT, L2 development is conceptualized as targetlike use following a specific task design implementation or manipulation, and, as is common outside the field, as gains in pre- to posttests (e.g., Révész & Han, Reference Révész and Han2015). In TBLT, these gains are often discussed in terms of measuring changes in complexity, accuracy, and/or fluency (CAF; for a thorough review of CAF measures see Housen, Kuiken, & Vedder, Reference Housen, Kuiken and Vedder2012; for an empirical example see Kim & Tracy-Ventura, Reference Kim and Tracy-Ventura2013). Task-based manipulations with the goal of promoting L2 development have focused on instruction (e.g., De Ridder et al., Reference De Ridder, Vangehuchten and Seseña Gómez2007), task type(s) (e.g., Newton, Reference Newton2013; Révész & Han, Reference Révész and Han2015), task complexity (e.g., Nuevo, Adams, & Ross-Feldman, Reference Nuevo, Adams, Ross-Feldman and Robinson2011), interlocutor (e.g., Kim & McDonough, Reference Kim and McDonough2008), and mode (most often FTF as compared to SCMC; e.g., Baralt, Reference Baralt, Baralt, Gilabert and Robinson2014). Empirical research has found these task-based manipulations to promote the development of L2 accuracy, at least for lexis and grammar. For example, Nuevo et al. (Reference Nuevo, Adams, Ross-Feldman and Robinson2011) examined how L2 development of English locatives could be promoted by task complexity, and operationalized according to reasoning demands. They found that learners in the complex group had more instances of self-repair and that this related to greater L2 development. Kim and McDonough (Reference Kim and McDonough2008) found that interlocutor proficiency significantly affected the presence and resolution of both grammatically and lexically focused LREs for L2 English learners. Finally, Baralt (Reference Baralt, Baralt, Gilabert and Robinson2014), which examined mode (alongside task complexity, operationalized as intentional reasoning, and attention, operationalized as LREs) in relation to Spanish L2 development, found that learners who completed a complex task in the FTF mode significantly outperformed learners in the SCMC mode. Not only were FTF learners more accurate in their L2 production, learners who interacted in SCMC were too cognitively overwhelmed to even attempt the linguistic target, the imperfect subjunctive.

In other words, TBLT has been considered successful for promoting L2 development. However, this statement is applicable primarily for lexical and grammatical linguistic targets, and most often in English (e.g., locatives, Nuevo et al., Reference Nuevo, Adams, Ross-Feldman and Robinson2011; the past progressive, Révész & Han, Reference Révész and Han2015; past counterfactual, Révész, Sachs, & Hama, Reference Révész, Sachs and Hama2014, etc.). Whether task-based designs work for other language domains, including L2 pronunciation, has yet to be soundly empirically tested. For example, some studies, such as Kim and McDonough (Reference Kim and McDonough2008), operationalized a focus on pronunciation as within lexis (p. 218). De Ridder et al. (Reference De Ridder, Vangehuchten and Seseña Gómez2007), which contained an independent measure of L2 pronunciation and intonation (in addition to lexis and grammar), relied on two raters, who were also the course instructors, and only one of four components of the experiment was task-based, limiting the study’s contribution to our knowledge of whether task-based designs may promote gains in L2 pronunciation accuracy. Even for studies examining oral gains, such as Tonkyn (Reference Tonkyn, Housen, Kuiken and Vedder2012), no measure of pronunciation was included; L2 accuracy measures (along with fluency and complexity) were reserved for lexis and grammar.

Thus, previous TBLT research has demonstrated that tasks can direct learners’ attention to linguistic form within meaning and promote L2 development for grammatical, lexical, and even pragmatic targets, but what about L2 pronunciation? Despite its presence within the theoretical rationale for TBLT (see Ellis, Reference Ellis2009), we have yet to examine if communicative tasks designed with pronunciation linguistic targets can (a) successfully manipulate learner attention and (b) promote accuracy and/or development. Looking briefly at the larger role of L2 pronunciation in experimental SLA, this oversight is in keeping with trends outside of TBLT.

L2 PRONUNCIATION

Pronunciation is a central component of the acquisition of a L2. Nevertheless, Derwing and Munro (Reference Derwing and Munro2005) assert that “the study of pronunciation has been marginalized within the field of applied linguistics” (p. 379), with greater attention, even now, being given to other skills and foci, such as vocabulary and grammar. This marginalization, they argue, has led to a reliance on intuition, anecdotal evidence, and personal impression in the teaching of L2 pronunciation in the classroom. In light of this oversight, Derwing and Munro have called for “empirical, replicable studies to inform pronunciation instruction” (p. 380). They argue that, although there has been growing interest in research on speech production and perception, the technical nature of such research, and the fact that it is often conducted in controlled laboratory conditions, makes the immediate relevance and applicability of such research to the classroom less clear.

Within research on TBLT, pronunciation has likewise been marginalized, remarkably absent from most accounts of this approach (the few exceptions cited previously; e.g., Ellis et al., Reference Ellis, Baskturkmen and Loewen2001; Loewen, Reference Loewen2005). In fact, neither pronunciation nor phonetics/phonology are mentioned in the following recent volumes on TBLT: Task-Based Language Learning and Teaching (Ellis, Reference Ellis2003), Task-Based Language Teaching (Nunan, Reference Nunan2004), Task-Based Language Education: From Theory to Practice (Van den Branden, Reference Van den Branden2006), Task-Based Language Teaching: A Reader (Van den Branden, Bygate, & Norris, Reference Van den Branden, Bygate and Norris2009), Task-Based Language Teaching in Foreign Language Contexts (Shehadeh & Coombe, Reference Shehadeh and Coombe2012), Task-Based Language Teaching from the Teachers’ Perspective (East, Reference East2012), and Second Language Acquisition and Task-Based Language Teaching (Long, Reference Long2015). This absence is surprising given the emphasis in previous TBLT research on production measures within communicative outcomes, such as accuracy (including accuracy of specific structures), acceptability and appropriateness, global proficiency measures, and intelligibility (including in relation to the specific interlocutor[s]). Arguably, pronunciation and/or accent play a role in these areas.

Additionally, pronunciation matters in language competence, particularly in using language in real-world/meaningful communication contexts. As Moyer (Reference Moyer2013) describes, pronunciation and accent are “fundamental to communication, for without a reasonable degree of phonological fluency, spoken interaction will falter” (p. 9). She goes on to state that L2 speakers use pronunciation (using accent) to “project individual style and signal [their] relationship to interlocutors” and that the degree of intelligibility of a L2 speaker influences that speaker’s capacity to participate within the target language community (p. 19). It is for these reasons that L2 pronunciation is central to the entire act of communication in the L2, particularly as it relates to accent. Nevertheless, the acquisition of a new sound system is no easy task and is, as Derwing and Munro (Reference Derwing and Munro2005) argue, an undertaking for which L2 speakers need support. A growing body of research on L2 pronunciation learning is demonstrating that helping learners notice and pay attention to the way their pronunciation may differ from the target through training, awareness raising, and/or instruction may indeed be beneficial for L2 phonetic/phonological development.

PRONUNCIATION AND ATTENTION

Attention is not often addressed as a central construct in the existing research on L2 phonetic and/or phonological learning. Nevertheless, it does appear in theoretical discussions of L2 pronunciation teaching and learning. For instance, Moyer (Reference Moyer2013) posits that “phonological instruction is predicated on the learner’s ability to detect the differences between their own output and the native(like) model provided to them” (p. 154), and Derwing and Munro (Reference Derwing and Munro2005) state that “students learning L2 pronunciation benefit from being explicitly taught phonological form to help them notice the difference between their own productions and those of proficient speakers in the L2 community” (p. 388). Beyond underscoring the theoretical importance of attention (i.e., through words like detect [Moyer] and notice [Derwing & Munro]) in L2 pronunciation development research, these quotes further illustrate that instruction has been a primary means by which learners’ attention is drawn to phonetic and/or phonological form. Thus, it comes as no surprise that a fair amount of research on L2 pronunciation has been dedicated to investigating the impact of instruction on attention and subsequent learning.

In the research-based pronunciation teaching literature, a variety of instructional methods have been employed. Attested methods for drawing attention to pronunciation through teaching include the use of contrastive analysis techniques (Arroyo Hernández, Reference Arroyo Hernández2009; Hammerly, Reference Hammerly1982), explicit instruction and oral practice on the articulation of specific sound segments (e.g., Elliott, Reference Elliott1995, Reference Elliott1997; González-Bueno, Reference González-Bueno, Glass and Pérez-Leroux1997; Macdonald, Yule, & Powers, Reference Macdonald, Yule and Powers1994; Saito, Reference Saito2013), explicit phonetics training for perception and/or production (e.g., Aliaga-García & Mora, Reference Aliaga-García, Mora, Watkins, Rauber and Baptista2009; Derwing, Munro, & Wiebe, Reference Derwing, Munro and Wiebe1998; Kissling, Reference Kissling2013; Lord, Reference Lord2005, Reference Lord2008, Reference Lord2010; Missaglia, Reference Missaglia1999; Zampini, Reference Zampini1998), and peer teaching (e.g., Rodríguez-Sabater, Reference Rodríguez-Sabater2005). Research has also examined the impact of feedback targeting pronunciation-related errors (e.g., Saito, Reference Saito2015; Saito & Lyster, Reference Saito and Lyster2012) as well as the use of journal entries in combination with in-class instruction (e.g., Kennedy & Trofimovich, Reference Kennedy and Trofimovich2008). Trofimovich and Gatbonton (Reference Trofimovich and Gatbonton2006) have also demonstrated the beneficial impact of repetition and focus-on-form on L2 pronunciation accuracy, outlining implications for instruction. The use of instruction as a medium for drawing learner attention to phonetic and/or phonological form and interest in the empirical investigation of whether and/or how instruction can influence L2 pronunciation outcomes is growing (see Lee, Jang, & Plonsky, Reference Lee, Jang and Plonsky2015, for a recent meta-analysis on the effectiveness of L2 pronunciation instruction, and Saito, Reference Saito2012, for an earlier research synthesis).

L2 PRONUNCIATION DEVELOPMENT

Empirical studies measuring L2 pronunciation learning generally rely on the elicitation of spoken speech, ^{Footnote 2} either through open-ended, informal conversations and interviews or through more controlled production tasks that elicit particular sounds in specific phonetic contexts. High-quality audio-recorders and the ready availability of acoustic analysis software (e.g., Praat; Boersma & Weenink, Reference Boersma and Weenink2014) have changed the ways in which L2 speech is analyzed (e.g., Hansen Edwards & Zampini, Reference Hansen Edwards, Zampini, Hansen Edwards and Zampini2008) and subsequently interpreted. Acoustic and spectrographic analyses of speech dominate studies of L2 speech production, as researchers quantify phonetic and phonological development through the examination of specific measures that correlate with the adjustment of different aspects of the articulation and realization of L2 sounds. Additionally, methodological designs that include cross-sectional samples or data involving pre- and post- (e.g., treatment, instruction, study abroad) recordings offer information regarding ongoing modifications to learners’ L2 systems (i.e., learning or development). Overall, Ellis (Reference Ellis2008) describes three general characteristics of L2 phonology that are largely accepted by researchers in the field: (a) L1 phonological features are present in learners’ L2 speech, (b) learners tend to substitute unmarked forms in places in which marked forms are required by the target language systems, and (c) unique forms also appear in L2 learners’ production, just as they do during L1 phonological acquisition. Additionally, we know that L2 pronunciation is an area of L2 competence in which, even after many years of exposure or study, learners often fail to achieve targetlike accuracy (e.g., Loewen, Reference Loewen2014; Long, Reference Long1990), but that, overall, learners who receive pronunciation instruction outperform those who do not and that the effectiveness of instruction has been shown to be augmented when feedback on pronunciation is provided (Lee et al., Reference Lee, Jang and Plonsky2015).

Our knowledge of the development of L2 sound systems is quickly growing; nevertheless, as Moyer (Reference Moyer2013) writes, “Despite increasing interest in phonology among applied linguists, neither the cognitive nor the sociolinguistic paradigm offers much practical insight or advice for FL teachers” (p. 147). The question remains, could TBLT be an ideal locus of connection between the theoretical and the practical in the instructed SLA of pronunciation?

TBLT: AN IDEAL MEDIUM FOR L2 PRONUNCIATION?

The studies in this thematic issue combine TBLT designs and measurements with standard phonetic and pronunciation measures to examine if the benefits of TBLT and of tasks, which have been established for grammar and lexis and have begun to be investigated for pragmatics, extend to L2 pronunciation learning. As briefly reviewed, despite a wealth of empirical information on TBLT and on L2 pronunciation and phonetic/phonological learning separately, to date, there is a lack of systematic investigation into the role of tasks and features of task design in promoting opportunities for attention to and the development of L2 pronunciation, which represents a central aspect of L2 acquisition and communicative competence. The five empirical studies in this special issue ^{Footnote 3} use innovative designs, methods, and measures to combine TBLT and L2 phonetics/phonology to expand our knowledge of instructed SLA and of L2 pronunciation development. Additionally, the studies are carried out in various interactional contexts: They include researcher-learner and learner-learner interactions; take place in FTF and SCMC contexts; target Spanish and English; examine foreign and second language contexts; and investigate both segmental and suprasegmental features. This thematic issue exemplifies the way TBLT offers an ideal framework for approaching the empirical study of L2 pronunciation learning in the classroom. The studies in this issue seek to combine and build on what we know about the role of pronunciation in L2 competence, the role of attention in the acquisition of L2 pronunciation, and how pronunciation is learned (as well as how pronunciation learning is measured) while exploring and employing concepts and approaches central to TBLT. In this way, we examine whether TBLT can encourage L2 pronunciation learning as it has been shown to do for L2 grammar, vocabulary, and pragmatics learning. Together, we ask and take the first steps to answer: Do the benefits of tasks extend to L2 pronunciation?

The first empirical study, by Loewen and Isbell, examines the factors of modality (i.e., FTF vs. SCMC), learner L1, and task type in relation to the production of pronunciation-related LREs during task-based interaction by dyads of L2 English learners from a variety of L1 backgrounds. The findings of their study encourage us to consider the role that task characteristics, modality, and learner factors play in orienting learners’ focus toward pronunciation during interaction. The following study, by Parlak and Ziegler, also examines modality but in the context of feedback provision (specifically, recasts) and the acquisition of lexical stress by native Arabic learners of English. Their study is important for demonstrating the potential for recasts to positively influence pronunciation development (in this study, at the suprasegmental level) within TBLT, which is in line with existing pronunciation research outside of TBLT (e.g., Saito & Lyster, Reference Saito and Lyster2012). McKinnon’s study investigates the impact of task-based instruction on prosody by L2 learners of Spanish. His work highlights the potential of task-based instruction for prosodic (specifically, intonation) learning and adds an important contribution to the discussion of the acquisition of a notable pragmatic contrast in L2 Spanish. The study by Jung, Kim, and Murphy extends auditory priming methods and task repetition, a heavily investigated concept in TBLT research, to the examination of English stress patterns by L2 learners. Their study similarly offers positive support for the role of priming and task repetition in the development of a prosodic feature of the L2. The final empirical study, by Solon, Long, and Gurzynski-Weiss, examines whether cognitively simple or cognitively complex dyadic map tasks designed to make pronunciation task-essential encouraged the production of pronunciation-focused LREs and/or accuracy of phonetic form. The findings of their study suggest that some of the predictions of the cognition hypothesis (Robinson, Reference Robinson2003, Reference Robinson2011; Robinson & Gilabert, Reference Robinson and Gilabert2007), previously shown to hold for accuracy on specific grammatical structures, may also extend to pronunciation. The epilogue, written collaboratively by phonological expert Mora and TBLT specialist Levkina, synthesizes and contextualizes conclusions from the empirical studies within both task-based and phonetic/phonological perspectives, outlines promising research questions and future directions, and calls for researchers to utilize both task-based and phonological measures to most robustly address these research questions.

The novel and varied studies in the present issue demonstrate that TBLT is a worthwhile avenue for promoting attention to and development of L2 pronunciation. It is our hope that the research presented in this special issue paves the way for an empirically based discussion of how tasks may be utilized in L2 pronunciation and on the nature of tasks in all areas of L2 development. Such research carries important implications for TBLT research theory and pedagogy as well as for L2 pronunciation learning and teaching.