Pronunciation and Task-Based Instruction: Effects of a Classroom Intervention

Abstract

Explicit pronunciation instruction enhances production of intelligible, comprehensible, and fluent second language (L2) speech. Additionally, task-based instruction (TBI) enhances the production of segmentals and suprasegmentals, and promotes awareness of L2 forms in pronunciation. Since most previous research on TBI in L2 pronunciation learning has been carried out in laboratory settings, it is necessary to investigate how the implementation of tasks that differ in complexity in actual classroom settings could benefit L2 pronunciation learners. This classroom-based study investigated the effects of combining focus on form and TBI in L2 pronunciation instruction in the production of comprehensible, fluent, and L2-accented speech in English-as-a-foreign-language learners. Although the results of the study present modest gains, they make a case for the inclusion of explicit pronunciation instruction and tasks that differ in complexity to help learners develop pronunciation skills. Responding to recent calls for more classroom-based research that demonstrates how pronunciation can be implemented in systematic and principled ways, the results of this study are discussed in terms of implications and pedagogical recommendations for L2 pronunciation teaching and learning.

Keywords

Pronunciation teaching and learning task-based instruction classroom research comprehensibility oral communication

Introduction

In the current globalised world where communication among speakers of different varieties of English is more common every day, intelligible second language (L2) pronunciation is essential for effective oral communication. Although the majority of L2 learners present some type of foreign accent (Flege et al., 1995),¹ explicit pronunciation instruction that balances a focus on form and meaning proves to be effective in helping learners enhance their pronunciation skills (Lee et al., 2015; Thomson and Derwing, 2015). Additionally, recent research findings indicate that task-based instruction (TBI) can be beneficial for L2 pronunciation learning (Gurzynski-Weiss et al., 2017), particularly when learners are exposed to sequences of tasks that increase in complexity (Robinson, 2001, 2005, 2007). This is because TBI prompts more interaction and negotiation of meaning among learners (Loewen and Isbell, 2017; Solon et al., 2017), which improves the perception and production of pronunciation features (Mora and Levkina, 2017, 2018).

While previous research on TBI is promising for pronunciation teaching, further inquiry is necessary to explore how using different tasks could enhance pronunciation learning in actual classrooms. The current study presents the methodology and the results of a short pronunciation classroom intervention that investigated the effects of combining explicit pronunciation instruction and TBI. To facilitate this intervention, three groups of English-as-a-foreign-language (EFL) learners at a small university in Costa Rica received pronunciation instruction centred on suprasegmentals combined with different tasks that varied in complexity during three weeks of instruction. A set of speech samples collected in a pre-test and a post-test evaluated the gains of these learners in terms of comprehensibility, fluency, and accent.

Background

Pronunciation Teaching and Learning

Explicit pronunciation instruction has proven effective in helping learners develop intelligible (actually understandable) and comprehensible (easy to understand) L2 speech in classroom contexts (Derwing and Munro, 2009; Lee et al., 2015; Thomson and Derwing, 2015). Instruction is effective when learners are made explicitly aware of linguistic features of the L2 (Kennedy and Trofimovich, 2010; Venkatagiri and Levis, 2007), when it targets key segmental (vowels and consonants) and suprasegmental (stress, rhythm, and intonation) features that enhance intelligibility (Field, 2005; Hahn, 2004; Munro and Derwing, 2001, 2006), and when explicit phonetic instruction is combined with different types of corrective feedback (Saito, 2013; Saito and Lyster, 2012). Additionally, pronunciation instruction is enhanced through a combination of form-focussed instruction (FFI) with opportunities for learners to focus on meaning (Celce-Murcia et al., 2010; Saito, 2013; Saito and Saito, 2017; Sardegna et al., 2016). This combination of FFI and meaning is key because it allows learners to automatise the language forms learned under more controlled conditions (Gatbonton and Segalowitz, 2005; Segalowitz and Hulstijn, 2005; Trofimovich and Gatbonton, 2006).

Although classroom-based pronunciation interventions tend to be complicated because of logistical issues and other factors (Derwing and Munro, 2015), they have proven beneficial even during short periods of time (Gordon and Darcy, 2016), and when the intervention focusses on factors known to enhance intelligibility and comprehensibility (Levis and Muller Levis, 2018). For example, in a classroom-based study with ESL learners, Derwing et al. (1998) found that explicit instruction during 12 weeks on general speaking habits and prosodic features (e.g. stress, intonation, rhythm) helped learners develop more comprehensible and fluent spontaneous speech than did instruction on segmentals only. These results are evidence of the importance of providing a broad teaching framework combining both suprasegmentals and segmentals in instruction so that learners have the opportunity to develop different speaking aspects (e.g. stress, rhythm, intonation) as opposed to just production of individual segments.

In a more recent study, Darcy et al. (2019) compared explicit pronunciation instruction in three groups of ESL learners: FFI, FFI with communicative activities, and no specific pronunciation instruction (control). While both experimental groups improved their comprehensibility from pre-test to post-test, the control group remained unchanged, and only the group exposed to FFI with communicative activities significantly improved comprehensibility in the post-test. Darcy et al. (2019) argued that the inclusion of FFI with a communicative component in this group generated more pronunciation gains because the communicative activities allowed genuine communicative pressure for learners to interact as in real life. These results reinforce the idea that FFI is more effective when it is implemented with a communicative component (Celce-Murcia et al., 2010). In light of these results, TBI could make a difference for L2 pronunciation learning where a communicative component forces learners to use all their cognitive resources to focus on meaning and develop automaticity.

TBI and L2 Pronunciation

TBI is a type of communicative methodology that aims to develop communicative competence in L2 learners through an engagement in meaning-oriented communication when performing a task (Ellis and Shintani, 2014). In TBI, a task is a ‘workplan’ whose primary focus is on meaning and in which learners have to rely on their own linguistic resources for completion. This workplan usually contains some type of ‘information gap’ and there is a clear communicative outcome to be achieved (Ellis, 2003). From an acquisitional perspective, TBI has been considered suitable for language learning because it promotes interaction and negotiation of meaning in L2, which leads to noticing salient linguistic forms in the input (Doughty and Williams, 1998; Long, 1991; Schmidt, 2001) or producing more language-related episodes—that is, interactional sequences in which the learners focus on a specific linguistic feature and engage in discussion about it (Swain and Lapkin, 2001). Additionally, TBI promotes the development of communicative competence in learners through meaning-focussed communication while also aiming to develop linguistic and interactional competence (Ellis, 2003). This is because learners, while primarily focussed on meaning to produce and understand messages, also need to attend to linguistic forms for learning to take place (Ellis and Shintani, 2014). In this respect, Robinson’s Cognition Hypothesis (2001, 2005) provides an account of how TBI can help learners in L2 development. Robinson’s hypothesis states that, because complex tasks promote more interaction, attention to form, and uptake of information from the input than simpler tasks, they will foster more accurate and complex (though less fluent) L2 language as tasks increase in cognitive complexity.

In L2 pronunciation, TBI is effective in segmental (Mora-Plaza et al., 2018; Solon et al., 2017) and suprasegmental learning (Jung et al., 2017; McKinnon, 2017). Additionally, as task complexity increases, learners engage in more negotiation of meaning, producing more pronunciation-focussed language-related episodes (Loewen and Isbell, 2017; Mora and Levkina, 2018; Mora-Plaza et al., 2018; Solon et al., 2017). This is important because of the increase in awareness of linguistic forms, a necessary condition for pronunciation development, and because of the potential improvement in perception and production of such forms (Kennedy and Trofimovich, 2010; Mora and Levkina, 2017, 2018; Venkatagiri and Levis, 2007). These previous studies highlight the benefits that TBI could provide in L2 pronunciation instruction. However, the majority of these studies were carried out in laboratory conditions, and it is also necessary to understand if the implementation of TBI in actual L2 pronunciation classes could benefit L2 learners. Thus, because of the benefits of TBI in pronunciation learning (Ellis, 2003; Gurzynski-Weiss et al., 2017) and because FFI can be effective in the development of L2 pronunciation (Lee et al., 2015; Thomson and Derwing, 2015), it is critically important to explore how the implementation of FFI with TBI can help L2 learners achieve better pronunciation skills in actual classrooms.

The Current Study

Because of the need to explore the benefits of FFI combined with TBI in intact classes, this study investigated the results of a short-term pedagogical intervention combining FFI and TBI in an EFL classroom context. Varying the task complexity, but keeping instruction the same, allows for the evaluation of explicit pronunciation instruction in relation to task complexity in leading to improvements in comprehensibility. Three EFL classes at a university in Costa Rica were exposed to FFI based on suprasegmentals, with tasks of differing complexity, during three weeks in which instruction was provided two days per week for 30 minutes each day. Additionally, participants recorded spontaneous speech samples in a pre-test and post-test, which were presented to a group of first language (L1)-English-speaking raters to be evaluated for comprehensibility (i.e. the ease or difficulty a listener experiences in understanding L2 speech), fluency (i.e. the degree to which speech flows easily without unnatural pauses and other dysfluencies), and accentedness (i.e. the degree to which L2 speech is different from the expected L1 patterns) (Derwing and Munro, 2009). The study was guided by the following research questions:

Do EFL students improve their pronunciation (in terms of comprehensibility, fluency, and accentedness) by the end of a three-week pronunciation classroom intervention? And if so,

Can explicit pronunciation instruction combined with tasks at different levels of complexity lead to better L2 pronunciation?

The details of the study methodology are presented below.

Method

This study followed a pre-test/post-test design in which different experimental treatments were given to three intact EFL classes (Mackey and Gass, 2016). These classes were composed of students majoring in Tourism and Computer Engineering, and there were a total of 21, 22, and 24 students in each class. They were selected for convenience, as it would have been logistically impossible to randomly assign participants to different groups (Mackey and Gass, 2016). These three classes were the first of a sequence of four basic integrated-skill EFL courses. The student participants were first-year university students with ages ranging from 18 to 21 years, and most of them presented a very basic level of the language, as many of them had recently finished secondary school.² The three classes were taught by different non-native EFL teachers with graduate degrees in EFL teaching and 16, 14, and 13 years of teaching experience. These teachers were given guidelines (i.e. lesson plans and materials) to implement the treatment in their classes in a way they felt comfortable doing.

Treatment: Tasks and Suprasegmental Instruction

The students received treatment two times per week in 30-minute sessions each day over a three-week period. Each session followed a sequence in which the students were exposed to explicit phonetic and phonological information (FFI), engaged in practice exercises, and completed a communicative task following the guidance of Celce-Murcia et al. (2010). These communicative tasks—the central variable of interest in the study—were different in each group and differed in terms of their complexity (see Hewings, 2004; and Appendices A, B, and C for a sample of each task). In the first class, the communicative tasks were simple communicative activities like a conversation based on a set of questions or prompts (Conversation Class). In contrast, in the second class, the students engaged in simple decision-making tasks like finding directions on a map or looking for an apartment (Simple-Decision-Making Class). Finally, the students in the third class completed tasks very similar to those of the second class but with extra elements of complexity (Complex-Decision-Making Class). The criteria used for designing these tasks were based on Robinson’s (2007) resource-directing variable +/– elements. For instance, tasks implemented in one of the classes were simpler because they contained [+few] elements. In contrast, tasks implemented in the other class were more complex and contained [–few] elements. Figure 1 presents the type of treatment given to the three classes.

Figure 1.

Experimental treatment in three EFL classes.

As background for the communicative tasks, the treatment given in each class consisted of explicit phonetic and phonological instruction of suprasegmentals and connected speech. The students received explicit instruction in perception and production of word stress in multisyllabic words, sentence stress (e.g. more prominent syllables in a phrase), rhythm (e.g. stressed and unstressed syllables), thought groups (e.g. groups of words that go together grammatically and are separated by brief pauses), reductions (e.g. vowel reduction in function words), linking (e.g. resyllabification in CC-V sequences (e.g. find_out, kept_under, dreamed_of) and intervocalic consonants in VC-C sequences (e.g. drea_m_on, kee_p_up)), and intonation (e.g. rising, falling, and mid-rise contours). The practice exercises consisted of working in pairs reading dialogues or analysing a short written passage for thought groups, pauses, and reductions (among other features), then practising it out loud. This focus was chosen because it has been shown to be effective in short-term interventions to improve comprehensibility (Gordon and Darcy, 2016).

Comprehensibility, Fluency, and Accentedness Rating Tasks

The students from the three EFL classes recorded spontaneous speech samples the week before (pre-test) and after (post-test) the treatment. Each student watched a short video cartoon (about 2:30 minutes each) and then recorded a description of the main plot (a different video in each test). These videos were part of a popular cartoon found on the internet (Simon’s Cat, 2009, 2010). The students recorded the speech samples individually in a quiet room at a library on the school campus using a personal computer equipped with a microphone. After initial dysfluencies (e.g. hesitations, false starts), a segment of the first 25 seconds of each recording was selected from each participant (from pre-test and post-test) as stimuli for the rating task. In addition to the EFL learners, two L1-English speakers recorded samples to be used as a baseline. These L1-English speakers were undergraduate students, male and female, at an American university in the Midwest, aged 21 and 19, respectively, from the same midwestern region, and not proficient in any L2s. They recorded descriptions of both videos only once. A few participants’ samples had to be excluded from the stimuli for various reasons. For example, some students missed treatment sessions, and others could not complete the pre-test and post-test appropriately. Thus, the stimuli for the rating task consisted of samples from eight students from the Conversation Class, seven students from the Simple-Decision-Making Class, and eight students from the Complex-Decision-Making Class (in addition to the two baseline L1-English speakers).

A total of 32 L1-English speakers listened and rated the speech samples. However, only the scores of 26 raters were used in the final analyses, as some participants did not complete the task properly (e.g. rating all samples with the same score). All raters were undergraduate students at an American university in the Midwest. They were pursuing different majors, and they were enrolled in a first semester of Spanish-as-a-foreign-language class. They did not speak Spanish or any L2s fluently. These raters completed the rating task individually in a computer laboratory using high-quality headphones. They received instructions in oral and written form. First, they watched the video cartoons that the L2 speakers described in the pre-test/post-test to mitigate the effect of familiarity with the story and avoid unbalanced ratings with the first speech samples (Derwing et al., 2004). They were then told that they would hear samples of speakers describing those two video cartoons, and that they had to rate each speech sample using nine-point Likert scales for comprehensibility (1 = impossible to understand, 9 = extremely easy to understand), fluency (1 = extremely dysfluent, 9 = extremely fluent), and accentedness (1 = very strong foreign accent, 9 = no foreign accent). These scales are similar to those used in previous studies (Levis and Muller Levis, 2018).

After a short warm-up to become familiar with the task (using samples from another study), the raters heard the actual samples presented in randomised order. They were told that for comprehensibility they should rate how easy or difficult it was to understand each sample, and that problems with vocabulary or grammar could influence their ratings. For fluency, they were asked to pay attention only to the natural flow of speech and pauses regardless of errors of pronunciation, vocabulary, or grammar. For accentedness, they were asked to rate the difference of the pronunciation in speech from that of an L1-English speaker. The results of these rating tasks are presented below.

Results

The interrater reliability coefficients (Cronbach’s alpha) computed across all ratings were moderate for comprehensibility (pre-test = .78 and post-test= .78), fluency (pre-test = .70 and post-test = .70), and accentedness (pre-test = .84 and post-test = .84). For statistical analysis of the data, a Linear Mixed-Effects Model was used, with Comprehensibility, Fluency, and Accentedness as dependent variables, Group (Complex-Decision-Making Class, Simple-Decision-Making Class, Conversation Class, and Native Speakers) and Time (1 = pre-test, 2 = post-test) as fixed effects, and Speaker as a random effect. Significance levels were set at alpha = .05 for this and all subsequent analyses. For all models, Bonferroni corrections were used for multiple comparisons.

The overall model yielded a significant effect of Group for Comprehensibility (F(3, 21) = 17.24, p < .0001), a significant effect of Group for Fluency (F(3, 21) = 14.38, p < .0001), and a significant effect of Group for Accentedness (F(3, 21) = 54.24, p < .0001). There was a significant Group by Time interaction for Comprehensibility (F(3, 1271) = 2.83, p = 0.0372); however, there were no significant Group by Time interactions for Fluency (F(3, 1271) = 1.53, p = 0.2042) or Accentedness (F(3, 1271) = 0.81, p = 0.4879). Table 1 presents the overall scores obtained by all groups at each time for the three dependent variables. This interaction suggests that the changes in speech patterns from T1 to T2 are modulated by task.

Table 1.

Mean comprehensibility, fluency, and accentedness ratings for all classes.

Group	Comprehensibility				Fluency				Accentedness
	Pre-test		Post-test		Pre-test		Post-test		Pre-test		Post-test
	M	SD	M	SD	M	SD	M	SD	M	SD	M	SD
CDM Class	4.65	1.87	5.13	1.78	5.58	1.88	5.88	1.72	4.28	1.91	4.53	1.90
SDM Class	4.43	1.95	4.37	1.88	5.39	2.03	5.26	1.82	4.26	2.13	4.25	2.03
Conv. Class	4.63	1.88	4.54	1.64	5.66	1.89	5.51	1.74	4.35	1.96	4.20	1.95
NS Group	8.90	0.36	8.85	0.50	8.96	0.19	8.98	0.14	8.96	0.19	8.96	0.19

CDM: Complex-Decision-Making Class; SDM: Simple-Decision-Making Class; Conv. Class: Conversational Class; M: mean; SD: standard deviation.

The interaction obtained for Comprehensibility warrants examining the effect of task more in depth. In order to focus on the effect of task itself (which only applies to the L2 learners), the subsequent analyses were carried out without the group of Native Speakers. These analyses are presented separately for Comprehensibility, Fluency, and Accentedness. The fixed effect of Group has been renamed Task to highlight its relevance to the research question.

Comprehensibility Ratings

A Linear Mixed-Effects Model was conducted with Comprehensibility as the dependent variable, with Task (Complex-Decision-Making Class, Simple-Decision-Making Class, and Conversation Class) and Time (1 = pre-test, 2 = post-test) as fixed effects, and Speaker as a random effect. The analysis yielded no overall significant effect of Task (F(2,20 = 0.66, p = 0.52) nor Time (F(1,1170) = 1.24, p = 0.26), but, most crucially, there was a significant Task by Time interaction (F(2, 1170) = 3.81, p = 0.0225), in which the Complex-Decision-Making Class became significantly more comprehensible than the Simple-Decision-Making and Conversation classes. Table 2 presents the mean differences in the three learner groups, which confirms that unlike the Simple-Decision-Making and Conversation classes, the Complex-Decision-Making Class improved its comprehensibility significantly from pre-test to post-test. Figure 2 summarises the results from pre-test to post-test in the three groups.

Table 2.

Mean differences from pre-test to post-test for comprehensibility in three EFL classes.

Group	Mean diff. T1 to T2	SE	DF	t-value	p-value	95% CI
CDM Class	–0.4808	0.1642	1170	–2.93	0.0035	–0.8029	–0.1486
SDM Class	0.06044	0.1755	1170	0.34	0.7307	–0.2840	0.4048
Conv. Class	0.09615	0.1642	1170	0.59	0.5583	–0.2260	0.4183

SE: standard error

DF: degrees of freedom

CI: confidence interval.

Figure 2.

Comprehensibility ratings across time in three EFL classes.

Fluency Ratings

A Linear Mixed-Effects Model was conducted with Fluency as the dependent variable. Parallel to the model above, Task (i.e. Complex-Decision-Making Class, Simple-Decision-Making Class, and Conversation Class) and Time (i.e. 1 = pre-test; 2 = post-test) were used as fixed effects, and Speaker as a random effect. The analysis yielded no significant effect for Task (F(2, 20) = 0.57, p = 0.57), no significant effect for Time (F(1, 1170) = 0.00, p = 0.95), and no significant Task by Time interaction (F(2, 1170) = 2.12, p = 0.12). The lack of interaction and main effects indicates that the fluency ratings did not meaningfully change over time, and were not modulated by task.

Accentedness Ratings

To replicate the method used previously, a Linear Mixed-Effects Model was applied to analyse accentedness. The model declared Accentedness as the dependent variable, with Task (Complex-Decision-Making Class, Simple-Decision-Making Class, and Conversation Class) and Time (1 = pre-test, 2 = post-test) as fixed effects, and Speaker as a random effect. For this variable, the overall effects of the analysis yielded no significant effect of Task (F(2, 20) = 0.19, p = 0.8262), no effect of Time (F(1, 1170) = 0.08, p = 0.7779), and no Task by Time (F(2, 1170) = 1.12, p = 0.3277) interaction, indicating that the accentedness ratings did not meaningfully change, and that task did not play a role either.

Discussion

Explicit Pronunciation Instruction

The first research question of this study inquired whether L2 learners could improve their pronunciation by the end of a short classroom intervention, and the results demonstrated that the students in the Complex-Decision-Making Class were perceived as more comprehensible than those in the other two classes in the post-test. They also had higher fluency ratings than those in the other classes, though the difference in fluency was not significant.

These results confirm that it is possible to help L2 learners in the development of pronunciation skills through explicit instruction (Lee et al., 2015; Thomson and Derwing, 2015), particularly when such instruction is tailored to develop awareness in learners of phonetic and phonological characteristics of the L2, or of differences between their L1 and L2 (Kennedy and Trofimovich, 2010; Venkatagiri and Levis, 2007). Additionally, these results confirm the key role of prosodic features in the perception of comprehensible L2 speech, and the importance of instruction of suprasegmentals to enhance comprehensibility (e.g. Derwing et al., 1998; Field, 2005; Hahn, 2004). While segmental features are indeed necessary in the production of intelligible and comprehensible speech, and some combination of instruction on segmentals and suprasegmentals could provide learners with positive results (Derwing et al., 1998), the findings presented here add support to the role of prosody in helping learners develop comprehensible L2 speech in a short time (Gordon and Darcy, 2016; Levis and Muller Levis, 2018).

It is important to notice that this treatment consisted of only three weeks and 30 minutes twice per week, and there are three reasons that could potentially explain the results obtained. First, because the entire intervention was in fact only three hours, and the time span from pre-test to post-test was short, it is normal not to expect drastic changes in the speech of the learners. The second reason is that more time may be necessary to help learners develop automaticity that could potentially help them develop fluency, particularly because learners struggle during early learning stages to control linguistic forms while attempting to communicate, and fluency takes larger amounts of input, interaction, and practice to develop (Gatbonton and Segalowitz, 2005; Lahmann et al., 2017; Saito et al., 2019; Segalowitz and Hulstijn, 2005). A third reason is that fluency is only one of several components of comprehensibility (Isaacs and Trofimovich, 2012). Thus, it is possible that the learners in the Complex-Decision-Making Class used other linguistic resources (e.g. proper suprasegmentals, grammar, vocabulary, discourse strategies) in order to be more comprehensible (see also Levis and Muller Levis, 2018).

With regard to the accentedness ratings, the three EFL classes received a similar FFI treatment, but none of the learners were perceived as less accented in the post-test. This result confirms that a speaker may be perceived as comprehensible while still being perceived as accented (Munro and Derwing, 1995). These results thus provide support for the need to anchor pronunciation instruction on intelligibility or comprehensibility principles instead of nativeness principles (Levis, 2005). While a foreign accent is generally a natural consequence of learning an L2 (Flege et al, 1995), it is important to prioritise instruction on features that could enhance comprehensibility and intelligibility in learners since a change in accent may not necessarily make a difference.

TBI and L2 Pronunciation

L2 learners need a communicative component in pronunciation instruction to automatise the forms learned under more controlled conditions (Celce-Murcia et al., 2010), and the results of this study are evidence of the potential benefits of TBI when implementing such a communicative component. The learners in the Complex-Decision-Making Class, who completed more complex tasks as part of their treatment, ended up being more comprehensible during the post-test than in the pre-test. This confirms the potential value of including TBI in classroom pronunciation instruction, given that previous research has demonstrated how tasks help L2 learners enhance the acoustic production of segmentals (Mora-Plaza et al., 2018; Solon et al., 2017) and suprasegmentals (Jung et al., 2017; McKinnon, 2017). In addition, and in line with previous findings, the results of this study indicate that manipulating task complexity (i.e. exposing learners to more complex tasks) can stimulate the production of more accurate linguistic forms (Jung et al., 2017; McKinnon, 2017)—in this case, prosodic features that are necessary to enhance comprehensibility in the L2 (Gordon and Darcy, 2016; Derwing et al., 1998; Levis and Muller Levis, 2018). For instance, the use of complex tasks in the Complex-Decision-Making Class may have prompted more interaction and negotiation of meaning than in the other two classes, creating necessary conditions for learning to take place (Ellis and Shintani, 2014; Schmidt, 2001; Swain and Lapkin, 2001). In pronunciation learning, interaction and negotiation of meaning as a result of TBI may potentially produce more pronunciation-focussed language-related episodes (Loewen and Isbell, 2017). Thus, it is possible that the use of complex tasks created an environment for the learners in the Complex-Decision-Making Class to interact, negotiate meaning, and produce pronunciation-focussed language-related episodes, which allowed them to automatise the phonetic and phonological forms learned in more controlled conditions during the treatment (see Gatbonton and Segalowitz, 2005; Trofimovich and Gatbonton, 2006).

The incorporation of complex tasks as part of the treatment may have promoted the production of more accurate and complex pronunciation forms. Robinson’s (2001, 2005, 2007) Cognition Hypothesis states that pedagogic tasks sequenced in terms of complexity stimulate the use of more cognitive resources in learners for task completion, which promotes development of accuracy and complexity of linguistic forms in the L2—although at the expense of fluency. Therefore, it is possible that the increased complexity of tasks in the Complex-Decision-Making Class may have created proper conditions for learners to develop accuracy of linguistic forms (e.g. accurate use of prosodic features like stress and rhythm) and complexity in speech production (e.g. different discourse strategies) that ended up making their speech more comprehensible. At the same time, the fact that learners in this class did not significantly improve their fluency also aligns with the tenets of Robinson’s hypothesis that complex tasks lead to development of accuracy and complexity at the expense of fluency in the L2 (2001, 2005, 2007).

Another notable finding from this study is that improvement in pronunciation was noticeable to listeners. Although previous studies demonstrated that learners indeed acoustically improve their production of L2 speech as a result of TBI (e.g. Jung et al., 2017; McKinnon, 2017; Solon et al., 2017), the results of this study demonstrate that such acoustic improvement as a result of TBI can potentially be perceived by listeners. While acoustic improvement as a result of instruction is important, ultimately it is the listeners’ perception of improvement that is key because acoustic changes in L2 speech are not always noticeable (Thomson and Derwing, 2015).

Limitations, Key Findings, Pedagogical Implications, and Conclusion

This study presented specific limitations, like the small number of participants that were tested in the final analyses and the short period of the intervention. It is possible that a longer intervention could have yielded different results in a variable like fluency, and more longitudinal and controlled classroom-based research will be necessary to test this. Another limitation of this study is that measuring the number of pronunciation-focussed language-related episodes was outside the scope of this investigation. While task complexity promotes more interaction and negotiation of meaning that lead to an increase in pronunciation-focussed language-related episodes (Loewen and Isbell, 2017), future classroom-based pronunciation research could investigate whether instruction with complex tasks that promote an increase in pronunciation-focussed language-related episodes in interaction has any effects on the production of more comprehensible, intelligible, and/or fluent L2 speech.

Despite these limitations, the results of this study revealed that complex tasks paired with explicit instruction on suprasegmentals have the potential to lead to comprehensibility improvements after short-term explicit instruction. This is meaningful in an important way: this study is one of the first to examine previous claims regarding task complexity in pronunciation instruction with actual learners in real classrooms, and it suggests that higher complexity might indeed lead to larger benefits of explicit instruction. While it is true that more classroom-based pronunciation studies are needed to provide further support for the implementation of different types of tasks in pronunciation instruction, the findings of this study prove that it is possible to help learners develop comprehensible L2 speech by implementing research-based instruction. As stated by Thomson and Derwing (2015), evidence of improvements in L2 pronunciation in language classrooms as a result of instruction could prompt language teachers to follow principled teaching practices, and the results of this study provide pedagogical implications to implement pronunciation instruction systematically.

Based on the results of this study, there are three pedagogical implications for pronunciation instruction that could be implemented in classrooms. The first implication is to teach explicit phonetic and phonological aspects of the L2. While many learners may have an ‘ear’ for different forms of the L2, others will benefit from instruction that directs their attention to specific features of the L2 (Lee et al., 2015; Thomson and Derwing, 2015). Noticing and becoming aware of such features is a first step in L2 learning (Schmidt, 2001). The second pedagogical implication is that teachers should prioritise classroom teaching of aspects of pronunciation that make a difference in comprehensibility and intelligibility. Word and sentence stress, speech rate, and the functional load of certain segments are known to be of high value in perception of intelligible, comprehensible, and fluent L2 speech (Field, 2005; Hahn, 2004; Munro and Derwing, 2001, 2006; Zielinski, 2008). This means that in teaching contexts where there is limited time to fully implement pronunciation teaching, teachers should address pronunciation features that will make a difference in enhancing L2 speech. The third implication is to include a communicative component in teaching that allows learners to automatise the forms they learn in controlled conditions (Gatbonton and Segalowitz, 2005; Trofimovich and Gatbonton, 2006). While explicit FFI and controlled exercises help learners focus on accuracy of forms, it is necessary to include a communicative component so that learners can practise those linguistic forms in authentic speech. It is in this communicative component that the use of tasks that differed in complexity could help learners achieve more intelligible, fluent, and comprehensible L2 speech (Gurzynski-Weiss et al., 2017; Mora and Levkina, 2017).

In pronunciation instruction, tasks need to be implemented according to their complexity and the genuine or authentic nature of the task to promote authentic communication. For example, task complexity can be implemented by sequencing a simple task first (e.g. like the ones used in the Simple-Decision-Making Class), and then a more complex one containing more elements (e.g. like the ones used by the Complex-Decision-Making Class) or more steps to complete it. This sequence of similar steps that go from simple to complex constitutes a repetitive procedure that gives learners the opportunity to automatise forms (Gatbonton and Segalowitz, 2005). Additionally, the genuine nature of a task could promote authentic interaction and negotiation of meaning to make learners aware of the significance of proper pronunciation forms to enhance communication when performing the task. To do this, tasks need to be designed in a way that prompts the use of the forms learnt through explicit instruction. For example, in a situation in which learners analyse a feature like intonation explicitly, a task that requires asking questions and producing statements using different intonation contours to get meaning across would be more suitable than one that relies mainly on a different pronunciation feature (e.g. minimal pairs). Therefore, in addition to their complexity, tasks need to be tailored in a way which shows learners that certain pronunciation forms need to be used for successful task completion.

Finally, more research into pronunciation and TBI is necessary to investigate not only the best ways to implement the findings of laboratory studies in actual classrooms, but also the use of tasks to promote conditions that enhance pronunciation learning (e.g. enhancing interaction and negotiation of meaning). As both pronunciation and TBI are currently relevant research areas in language teaching, future studies combining these two areas could promote their use in actual classrooms. Such findings will be useful for both researchers and language teachers to implement pronunciation teaching practices that can move pronunciation teaching forward and, in this way, promote the development of intelligible, comprehensible, and fluent L2 speech.

Supplemental Material

sj-docx-1-rel-10.1177_0033688220986919 – Supplemental material for Pronunciation and Task-Based Instruction: Effects of a Classroom Intervention

Supplemental material, sj-docx-1-rel-10.1177_0033688220986919 for Pronunciation and Task-Based Instruction: Effects of a Classroom Intervention by Joshua Gordon in RELC Journal

Footnotes

Acknowledgements

I would like to thank my colleagues Lena Barrantes, Isabelle Darcy, Avizia Y Long, and the anonymous reviewers for their comments and suggestions on previous versions of this manuscript. All errors remain my own. I would also like to thank the guest editor Martha C Pennington for her suggestions and crucial input on this manuscript.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was completed through a Summer Fellowship 2019-2020 awarded to the author by the Graduate College of the University of Northern Iowa.

ORCID iD

Joshua Gordon

Supplemental material

Supplemental material for this article is available online.

Notes

References

Celce-Murcia

Brinton

Goodwin

, et al. (2010) Teaching Pronunciation: A Reference for Teachers of English to Speakers of Other Languages. 2nd ed. Cambridge: Cambridge University Press.

Darcy

Hancock

Lee

, et al. (2019) Improving pronunciation in spontaneous speech? A comparison of two instructional methods. In: L2 pronunciation research workshop 2019: Bridging the gap between research and practice, Barcelona, Spain, 6–7 June 2019.

Derwing

Munro

(2009) Putting accent in its place: Rethinking obstacles to communication. Language Teaching 42(4): 476–490.

Derwing

Munro

(2015) Pronunciation Fundamentals: Evidence-Based Perspectives for L2 Teaching and Research. Amsterdam: Benjamins.

Derwing

Munro

Wiebe

(1998) Evidence in favor of a broad framework for pronunciation instruction. Language Learning 48(3): 393–410.

Derwing

Rossiter

Munro

, et al. (2004) Second language fluency: Judgments on different tasks. Language Learning 54(4): 655–679.

Doughty

Williams

(1998) Pedagogical choices in focus on form. In: Doughty

Williams

(eds) Focus on Form in Classroom Second Language Acquisition. Cambridge: Cambridge University Press, 197–285.

Ellis

(2003) Task-Based Language Learning and Teaching. Oxford: Oxford University Press.

Ellis

Shintani

(2014) Exploring Language Pedagogy through Second Language Acquisition Research. London: Routledge.

10.

Field

(2005) Intelligibility and the listener: The role of lexical stress. TESOL Quarterly 39(3): 399–423.

11.

Flege

Munro

MacKay

IRA

(1995) Factors affecting strength of perceived foreign accent in a second language. Journal of the Acoustical Society of America 97(5): 3125–3134.

12.

Gatbonton

Segalowitz

(2005) Rethinking communicative language teaching: A focus on access to fluency. Canadian Modern Language Review 61(3): 325–353.

13.

Gordon

Darcy

(2016) The development of comprehensible speech in L2 learners: A classroom study on the effects of short-term pronunciation instruction. Journal of Second Language Pronunciation 2(1): 56–92.

14.

Gurzynski-Weiss

Long

Solon

(2017) TBLT and L2 pronunciation: Do the benefits of tasks extend beyond grammar and lexis? Studies in Second Language Acquisition 39(2): 213–224.

15.

Hahn

(2004) Primary stress and intelligibility: Research to motivate the teaching of suprasegmentals. TESOL Quarterly 38(2): 201–223.

16.

Hewings

(2004) Pronunciation Practice Activities: A Resource Book for Teaching English Pronunciation. Cambridge: Cambridge University Press.

17.

Isaacs

Trofimovich

(2012) Deconstructing comprehensibility: Identifying the linguistic influences on listeners’ L2 comprehensibility ratings. Studies in Second Language Acquisition 34(3): 475–505.

18.

Jung

Kim

Murphy

(2017) The role of task repetition in learning word-stress patterns through auditory priming tasks. Studies in Second Language Acquisition 39(2): 319–346.

19.

Kennedy

Trofimovich

(2010) Language awareness and second language pronunciation: A classroom study. Language Awareness 19(3): 171–185.

20.

Lahmann

Steinkrauss

Schmid

(2017) Speed, breakdown, and repair: An investigation of fluency in long-term second-language speakers of English. International Journal of Bilingualism 21(2): 228–242.

21.

Lee

Jang

Plonsky

(2015) The effectiveness of second language pronunciation instruction: A meta-analysis. Applied Linguistics 36(3): 345–366.

22.

Levis

(2005) Changing contexts and shifting paradigms in pronunciation teaching. TESOL Quarterly 39(3): 369–377.

23.

Levis

Muller Levis

(2018) Teaching high-value pronunciation features: Contrastive stress for intermediate learners. The CATESOL Journal 30(1): 139–160.

24.

Loewen

Isbell

(2017) Pronunciation in face-to-face and audio-only synchronous computer-mediated learner interactions. Studies in Second Language Acquisition 39(2): 225–256.

25.

Long

(1991) Focus on form: A design feature in language teaching methodology. In: De Bot

Ginsberg

Kramsch

(eds) Foreign Language Research in Cross-Cultural Perspective Amsterdam: Benjamins, 39–52.

26.

Mackey

Gass

(2016) Second Language Research: Methodology and Design. 2nd ed. New York: Routledge.

27.

McKinnon

(2017) TBLT instructional effects on tonal alignment and pitch range in L2 Spanish imperatives versus declaratives. Studies in Second Language Acquisition 39(2): 287–317.

28.

Mora

Levkina

(2017) Task-based pronunciation teaching and research: Key issues and future directions. Studies in Second Language Acquisition 39(2): 381–399.

29.

Mora

Levkina

(2018) Training vowel perception through map tasks: The role of linguistic and cognitive complexity. In: Proceedings of the 9th pronunciation in second language learning and teaching conference (ed J

Levis

), University of Utah, USA, 1–2 September 2017, pp.151–162. Ames, IA: Iowa State University.

30.

Mora-Plaza

Mora

Gilabert

(2018) Learning L2 pronunciation through communicative tasks. In: Proceedings of the 9th pronunciation in second language learning and teaching conference (ed J

Levis

), University of Utah, USA, 1–2 September 2017, pp.174–184. Ames, IA: Iowa State University.

31.

Munro

Derwing

(1995) Foreign accent, comprehensibility, and intelligibility in the speech of second language learners. Language Learning 45(1): 73–97.

32.

Munro

Derwing

(2001) Modeling perceptions of the accentedness and comprehensibility of L2 speech: The role of speaking rate. Studies in Second Language Acquisition 23(4): 451–468.

33.

Munro

Derwing

(2006) The functional load principle in ESL pronunciation instruction: An exploratory study. System 34(4): 520–531.

34.

Programa Estado de la Nación (2017) Sexto informe del estado de la educación costarricense. Consejo Nacional de Rectores. San José, Costa Rica. Available at: www.estadonacion.or.cr (accessed 1 December 2018).

35.

Robinson

(2001) Task complexity, task difficulty and task production: Exploring interactions in a componential framework. Applied Linguistics 22(1): 27–57.

36.

Robinson

(2005) Cognitive abilities, chunk-strength, and frequency effects in implicit artificial grammar and incidental L2 learning: Replications of Reber, Walkenfeld, and Hernstadt (1991) and Knowlton and Squire (1996) and their relevance for SLA. Studies in Second Language Acquisition 27(2): 235–268.

37.

Robinson

(2007) Criteria for classifying and sequencing pedagogic tasks. In: García-Mayo

MDP

(ed) Investigating Tasks in Formal Language Learning. New York: Multilingual Matters, 7–26.

38.

Saito

(2013) Reexamining effects of form-focused instruction on L2 pronunciation development: The role of explicit phonetic information. Studies in Second Language Acquisition 35(1): 1–29.

39.

Saito

Lyster

(2012) Effects of form-focused instruction and corrective feedback on L2 pronunciation development: The case of English /r/ by Japanese learners of English. Language Learning 62(2): 595–683.

40.

Saito

Suzukida

Sun

(2019) Aptitude, experience, and second language pronunciation proficiency development in classroom settings: A longitudinal study. Studies in Second Language Acquisition 41(1): 201–225.

41.

Saito

(2017) Differential effects of instruction on the development of comprehensibility, word stress, rhythm, and intonation: The case of inexperienced Japanese EFL learners. Language Teaching Research 21(5): 589–608.

42.

Sardegna

Chiang

Ghosh

(2016) Integrating pronunciation with presentation skills. In: Jones

(ed) Pronunciation in the Classroom: The Overlooked Essentials. Alexandria, VA: TESOL Press, 43–56.

43.

Schmidt

(2001) Attention. In: Robinson

(ed) Cognition and Second Language Instruction. Cambridge: Cambridge University Press, 3–32.

44.

Segalowitz

Hulstijn

(2005) Automaticity in bilingualism and second language learning. In: Kroll

De Groot

(eds) Handbook of Bilingualism: Psycholinguistic Approaches. Oxford: Oxford University Press, 371–388.

45.

Simon’s

Cat

(2009) Fly guy. Available at: https://www.youtube.com/watch?v=I1qHVVbYG8Y (accessed 19 August 2020).

46.

Simon’s

Cat

(2010) Snow business. Available at: https://www.youtube.com/watch?v=Tuf61OjvoPQ (accessed 19 August 2020).

47.

Solon

Long

Gurzynski-Weiss

(2017) Task complexity, language-related episodes, and production of L2 Spanish vowels. Studies in Second Language Acquisition 39(2): 347–380.

48.

Swain

Lapkin

(2001) Focus on form through collaborative dialogue: Exploring task effects. In: Bygate

Skehan

Swain

(eds) Researching Pedagogic Tasks: Second Language Learning, Teaching and Testing. New York: Longman, 99–118.

49.

Thomson

Derwing

(2015) The effectiveness of L2 pronunciation instruction: A narrative review. Applied Linguistics 36(3): 326–344.

50.

Trofimovich

Gatbonton

(2006) Repetition and focus on form in processing L2 Spanish words: Implications for pronunciation instruction. Modern Language Journal 90(4): 519–535.

51.

Venkatagiri

Levis

(2007) Phonological awareness and speech comprehensibility: An exploratory study. Language Awareness 16(4): 263–277.

52.

Zielinski

(2008) The listener: No longer the silent partner in reduced intelligibility. System 36(1): 69–84.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.10 MB