Introduction

Intelligibility has long been understood to be a fundamental requirement for effective human communication. Recognition of the centrality of this concept has resulted in a vast literature from such diverse fields as communication technology, audio engineering, speech language pathology, and audiology. In English language teaching, agreement about the importance of intelligibility is increasing among researchers and practitioners; however, as Levis (2005) points out, pronunciation teaching continues to be dominated by two contradictory principles: the nativeness principle and the intelligibility principle. The first of these assumes that learners should strive to become native-like in all aspects of pronunciation, whereas the intelligibility principle holds that learners should aim to develop speaking patterns that allow them to communicate with ease, even if their accent retains nonnative characteristics. The strong evidence suggesting that adult language learners rarely, if ever, achieve fully native-like pronunciation (Abrahamsson and Hyltenstam 2009; Flege, Munro, and MacKay 1995) and the compelling evidence that such a goal is unnecessary for effective communication (Munro and Derwing 1995a) have led us to work within a framework following from the intelligibility principle. Our attention below focuses on research about and applications of the intelligibility construct. In view of the fact that a fully satisfactory definition of the term intelligibility has proved elusive and no consensus exists on how to best measure it, we review the historical developments underpinning the current status of this notion.

Definitions

As early as 1900, the famous phonetician, Henry Sweet, raised the issue of intelligibility as a central goal in the learning and teaching of languages. He was one of the first in a long line of British academics who emphasized intelligibility, including Abercrombie (1949), who argued that second language (L2) speakers do not need to emulate native speakers, and Gimson (1962), whose Introduction to the Pronunciation of English is still in print. This work encourages language learners to strive for intelligibility rather than native speaker-like productions. While these pronunciation specialists offered no technical definition of the concept, they evidently assumed a notion of intelligibility as an umbrella term meaning a shared understanding between speaker and listener. Voegelin and Harris (1951: 327), working from an anthropological perspective, proposed an even broader, cross-dialectal definition: the degree to which “people of one community understand the speech of another”.

Attempts to formally define the intelligibility of L2 speech gradually emerged in the mid to late 20th century. However, the result was a diversity of opinions about how to describe the concept and how to evaluate it empirically. Catford (1950) distinguished the intelligibility of an utterance from its effectiveness, arguing that speech is intelligible when the listener identifies words correctly. An utterance is effective if the listener responds according to the speaker’s intention. He did not, however, support this distinction with any phonological examples. Moreover, his conception of unintelligible appears to be tautological in that he described unintelligible speech as including speech that is intelligible but ineffective.

In their landmark article, Smith and Nelson (1985: 334) proposed a tripartite distinction among terms relating to understanding: “(1) intelligibility: word/utterance recognition; (2) comprehensibility: word/utterance meaning (locutionary force); (3) interpretability: meaning behind word/utterance (illocutionary force).” In this hierarchical framework, the third level was the highest. Although it seems essential to distinguish word recognition from other aspects of understanding, this model does not appear to have motivated much work on L2 communication, perhaps because it is unclear how to apply it empirically. For instance, a problem of defining intelligibility as the lowest level in a hierarchy of understanding is that high-level listener processing is sometimes invoked in identifying a word. That is, the comprehension of speech is not a linear process: when a word cannot be recognized on the basis of bottom-up cognitive processes, the listener may nonetheless be able to “fill it in” by exploiting top-down strategies. Thus a transcription task cannot be assumed to exclusively measure Smith and Nelson’s (1985) notion of intelligibility because the task reflects more than just low-level speech processing.

A different approach was taken by Varonis and Gass (1982), who used the term “understanding” in the sense of “intelligibility” and described “comprehensibility” as a perceptual rating scale ranging from “I understood this sentence easily” to “I didn’t understand it at all”. In 1984, however, Gass and Varonis operationalized “comprehensibility” differently, as the percentage of words correctly transcribed by listeners as they heard L2 speakers’ productions; in other words, they used the word “comprehensibility” to refer to what many others have called “intelligibility”.

A perspective on intelligibility from the speech sciences is that of Schiavetti (1992: 13): “the match between the intention of the speaker and the response of the listener to the speech passed through the transmission system”. With respect to second language speech, Munro and Derwing (1995a: 76) characterized intelligibility as “the extent to which a speaker’s message is understood by a listener”. However, these definitions raise the issue of what we mean by such notions as “intention”, “message”, and “understand”. While it seems obvious, for instance, that “understanding a spoken message” entails recognizing and grasping the meaning of most or all the individual words that the speaker has produced, it is unclear how much attention should be focused on paralanguage and the more subtle aspects of speakers’ intentions, such as the unstated implications of their utterances, i.e., the illocutionary force that Smith and Nelson (1985) have called “interpretability”. Our concept of intelligibility is broader and nonhierarchical, encompassing all three of Smith and Nelson’s concepts, and recognizing that nonlinguistic factors, such as degree of shared knowledge and social context, may also affect understanding. Furthermore, it is not tied to any particular measurement—transcription tasks, comprehension questions, and other approaches are all ways of tapping into intelligibility, though no single task is fully satisfactory. While our approach conflates different aspects of understanding, it offers the advantage of relatively straightforward empirical assessment.

Rather than dwell on articulating a formal definition, we believe it is more fruitful to discuss intelligibility in terms of a number of functional properties that have been established in empirical research. The following apply only to spoken language communication:

• Intelligibility arises out of human interaction, in particular, the experience of one or more listeners as they process spoken material from an interlocutor. It therefore does not reside exclusively in either the speaker or listener.
• Intelligibility is a continuous phenomenon, such that the listener may understand all or none of the spoken material, as well as any intermediate amount. Furthermore, listeners may sometimes misjudge how much they have actually understood. They may realize, for instance, that some of the speaker’s words are not intelligible, but they may also assume that they have understood other words, when, in fact, they have not. This potential for misapprehension complicates the assessment of intelligibility (see below).
• Intelligibility is affected by the speech transmission system (a telephone, the Internet, the air, water) as well as the ambient environment (quiet, noisy).
• Intelligibility is at least partially independent of many other commonly discussed dimensions of speech, such as accentedness, comprehensibility, fluency, accuracy, or naturalness.

Within our framework for L2 speech assessment, two concepts related to intelligibility are comprehensibility and accentedness. Munro and Derwing (1995a) and Derwing and Munro (1997) use these terms to mean ease or effort of understanding and degree of difference from some comparison pattern respectively. Each dimension has its own continuum, ranging from high to low. Although intelligibility and comprehensibility are more important than accentedness to language learners’ communication skills, there is no hierarchical relationship among these dimensions. Comprehensibility and accentedness are partially independent of intelligibility, such that it is possible to be fully intelligible but somewhat difficult to understand. Furthermore, a speaker can be perceived to have a heavy accent, and yet be easy to understand and fully intelligible. The possible high–low combinations of intelligibility with comprehensibility are shown in Table 21.1, while the same combinations with accent appear in Table 21.2.

Local versus global intelligibility

An important distinction that is of use to both researchers and teachers is the difference between local and global intelligibility. The former refers to how well listeners recognize relatively small units of speech, such as segments and words, outside of a larger meaningful context, whereas the latter entails larger units of language that include rich contextual information. Although other researchers have not often used this terminology, previous empirical studies of intelligibility generally focus on one or the other. Field (2005), for instance, examined local intelligibility when he evaluated the effect of stress placement on listeners’ identifications of isolated single words. As he points out, this type of identification task has a different locus than that of a sentence or narrative dictation in which the listener may exploit contextual information to assist understanding. The latter task is global and more closely approximates language interaction in real communicative situations. Ou, Yeh, and Chuang (2012) found large differences in intelligibility scores, depending on whether they used a local approach in which individual words were transcribed (43% incorrect) or a global approach in which sentences containing the same words of interest were included (12% incorrect). Within our framework, global intelligibility is the goal of pronunciation instructors who want to enhance their students’ speech; research on local intelligibility is more useful to our understanding of L2 learning processes and to identifying some of the underlying components of global intelligibility. For instance, a local study might help us determine several speaker errors that lead to problems for the listener; however, only some of those may cause difficulties when contextual information is present. A fundamental problem for intelligibility researchers is to identify instances of the latter type. This concern is the one addressed by the functional load concept (Catford 1987). Phonological distinctions with a high functional load are those that “do a great deal of work” in the language. For instance, many minimal pairs in English are distinguished by the /i/ versus /I/ contrast. Moreover, the words in such pairs are of high frequency and often belong to the same lexical classes. Consequently, one can predict, on a purely theoretical basis, that a failure to produce the phonemic distinction between the sounds is likely to result in significant miscommunication (see Brown 1991; Levis and Cortes 2008), perhaps even when detailed contextual information is available.

Measurement

The need for satisfactory assessment of intelligibility has long been recognized in the speech sciences. In reference to speech production by the deaf, for example, Subtelny (1977: 183) described intelligibility as “the most practical single index to apply in assessing competence in oral communication”. Although a number of studies of L2 speech intelligibility have appeared in leading journals in recent years, relatively little attention has been directed to establishing the validity and reliability of intelligibility measures (Harding 2011). The fact that intelligibility is an aspect of interaction, for instance, means that it can be assessed only by reference to listeners’ experience. It is therefore not possible to measure it through an acoustic analysis of speech or through an expert phonetician’s fine-grained impressionistic analysis. Nor can a listener be expected to directly assess the intelligibility of a speaker. For instance, it is not logically possible to collect ratings of intelligibility on a scale; in the absence of any corroborating evidence, listeners may accurately identify their own failure to understand, but may also mistakenly assume successful communication where a breakdown actually occurred. Intelligibility assessment is only possible if the speaker’s intended utterance is known to the researcher and compared with the interpretation that the listener attributes to that same utterance. Given these constraints, intelligibility can be quantified in several ways, each of which has its own advantages and limitations. Here we will consider the following approaches that have been utilized in empirical work to assess intelligibility: word count, sentence verification, cloze and dictation, content summaries, and comprehension questions. We stress that each of these measures provides a window on the same, or closely related, underlying processes experienced by the listener. We therefore see them all as intelligibility measures, each of which is imperfect in its own way. A word count approach, for example, focuses strictly on exact word matches, but does not fully address illocutionary force, which would require further probing, perhaps with comprehension questions. On the other hand, using only comprehension questions entails a high risk of missing some aspects of a mismatch between the speaker’s specific production and the listener’s understanding because of the impossibility of evaluating every possible aspect of comprehension.

Laboratory and classroom-based studies of intelligibility

It is tempting to make a sharp distinction between laboratory and classroom-based research on pronunciation. In fact, however, this dichotomy is artificial and fails to take into account the many ethical and scientific issues that arise in evaluating the effects of instruction on intelligibility. The fact that a study has been carried out in a classroom does not necessarily make its results any more generalizable to typical classrooms than a study carried out in a laboratory. While some studies are classroom-based in the sense that the participants are registered in existing classes and the research is integrated into their regular instruction, the nature of the control and the experimental procedures are often identical to those used in laboratory settings. For example, in Derwing, Munro, and Wiebe (1998), three existing classes of ESL learners participated under three distinct conditions (suprasegmental instruction, segmental instruction, no specific pronunciation instruction). Although instruction took place in the students’ normal learning environment and was delivered by their homeroom teachers, their productions were collected and assessed using controlled laboratory methods. Saito and Lyster’s (2012) study combined several features of classroom and laboratory environments in that participants were recruited (as in a lab study) to form instructional cohorts over a period of two weeks. The students received differential pronunciation teaching and were recorded under laboratory conditions; the resultant data were assessed using acoustic measurements. The finding that /ɹ/ productions were acoustically more native-like when form-focused instruction was used together with corrective feedback led the authors to conclude that the condition was effective in improving intelligibility. However, no direct measures of intelligibility were included in the study. Thus, studies fall on a continuum that ranges from more classroom-like to more laboratory-like, but a key issue for pedagogical purposes is how intelligibility is measured, whether a particular intervention is effective, and what the study indicates about methods and foci.

Although some L2 production studies are concerned with matters of intelligibility, others tend to emphasize the acquisition and consequences of pronunciation accuracy compared to a native speaker norm. Most pronunciation research studies fall into three broad categories: focus-oriented studies that attempt to identify and assess the impact of characteristics of second language speech detrimental to intelligibility; acquisitional studies that trace the development of pronunciation intelligibility in L2 learners (using either cross-sectional or longitudinal designs); and intervention studies that examine whether a given method or technique is effective in bringing about changes in pronunciation.

Listener effects

It may seem trite to repeat the well-known fact that oral communication is at minimum a two-person enterprise, in which speaker and listener have equal responsibility for ensuring a successful outcome. All intelligibility research therefore necessarily entails both speakers and listeners. Although the focus in L2 research is often on the characteristics of L2 speakers’ productions, listeners play a crucial role in establishing the consequences of those characteristics. A number of researchers have pointed to the likely influence of listener variables on their comprehension of L2 speech (Munro 2008). It would not make sense to expect that every listener would respond in the same way to the same utterance. Rather, comprehension may vary depending upon a listener’s familiarity with a given speaker, a specific accent, non-native accents in general, or with a particular topic (Gass and Varonis 1984; Kennedy and Trofimovich 2008). In some instances, sharing a common L1 is thought to enhance comprehension of other speakers; however, such effects appear to be very small and inconsistent (Munro, Derwing, and Holtby 2012; Munro, Derwing, and Morton 2006). Differences in listener aptitude can also be related to intelligibility, such that some listeners appear to have an affinity for L2 accented speech (Munro, Derwing, and Holtby 2012). Moreover, some listeners may feel more motivated and able to interact with an L2 speaker than others, and thus make greater efforts to understand (Derwing, Munro, and Thomson 2008; Derwing, Rossiter, and Munro 2002). The listener’s age also influences understanding of L2 accented speech: elderly listeners appear to be at a disadvantage, even when their hearing is within normal limits for their age (Burda et al. 2003). Children also seem to understand less of L2 accented speech than adults (Munro, Derwing, and Holtby 2012). With all this potential variability, one has to ask whether teaching pronunciation is a viable enterprise. If listeners differ dramatically from each other in the ways in which they react to accented speech, how can an instructor decide on a focus for teaching? Correcting a particular problem might benefit some listeners but have no effect on others. Fortunately, despite the variability across listeners, evidence from rating studies reveals that their listening behavior is more similar than it is different. Presumably listeners process speech in very similar ways. If this were not the case, researchers would not see much interlistener agreement on which speech samples are intelligible and which are not. However, many studies show that diverse groups of listeners agree on which L2 speakers from a set are easy or difficult to understand (comprehensibility) and the same is true of listeners’ actual comprehension (intelligibility).

Teaching priorities in intelligibility-oriented instruction

In this section, we discuss priorities for pronunciation instruction with the understanding that language teachers require at least a basic knowledge of introductory linguistics and of the principles of L2 pronunciation in order to provide effective pedagogy. However, survey research indicates that many English language teachers have not actually received formal training in these areas (Foote, Holtby, and Derwing 2011).

Our conception of “prioritized pronunciation instruction” emphasizes helping learners to produce output that is comfortably intelligible (Abercrombie 1949), but is not concerned with “reducing” foreign accents. A first step in setting priorities within such a framework is to consider the kinds of changes in L2 speech production that might be effected through instruction. Table 21.3 lists eight logically possible outcomes of teaching interventions with respect to intelligibility, comprehensibility, and accentedness. A check mark in a cell indicates improvement on the relevant dimension, whereas an X indicates no change. Here we do not address the issue of whether some aspects of pronunciation might actually become worse as a function of instruction, though we recognize that such a risk exists.

Our interpretation of each combination is based on its consequences for a listener’s understanding of the speaker’s intended message. First, we consider simultaneous improvement of intelligibility and comprehensibility to be optimal, irrespective of whether the learner is heard as less accented. Second, any outcome in which one of intelligibility or comprehensibility improves is positive. It should be noted that the absence of improvement in one of these dimensions may at times indicate less than a fully desirable level of achievement; however, in other cases it may be an indication of a high level of performance on that dimension prior to instruction. For instance, a speaker who is already highly intelligible (and therefore cannot become more intelligible) could conceivably become more comprehensible thanks to instruction. In such a case, the speaker would have been highly intelligible with effort prior to instruction and highly intelligible with less effort afterwards. In short, the benefit for communication is that the listener must work less hard to understand the speaker. Third, when neither intelligibility nor comprehensibility improves, the outcome is negative, even if accent is reduced. This is because accent reduction is not relevant in prioritized pronunciation teaching, and should not be considered an appropriate goal when classroom time for instruction is limited. Of course, learners who wish to modify their pronunciation simply to approach some model are free to do so as they please.

The possibilities in Table 21.3 are based on the assumption that the three dimensions in question are independent enough of each other that a change in one does not automatically entail a change in either of the others. It is important to note that such an assumption is not merely a theoretical conjecture. Rather, empirical support for the actual occurrence of most of them has already been obtained. In an intervention study, for instance, Derwing, Munro, and Wiebe (1997) observed individual instances of Patterns 1, 4, 5, 6, 7, and 8. Intriguingly, each of these patterns was observed in the same group of learners receiving the same type of instruction. Further work to account for why the individual learners responded differently to the same instruction is clearly needed. Additional studies of learners of English and of other languages support the independence of some of the two-way combinations as well. For example, Derwing, Munro, and Wiebe (1998) reported improved comprehensibility of L2 speakers’ oral narratives with no change in accentedness, as well as improved accentedness in sentence productions without improved comprehensibility. Furthermore, when Holm (2008) provided intonation training to learners of Norwegian, she observed improvement in accent without improvement in intelligibility.

Focus priorities

Given the various possible instructional outcomes described in the last section, it is useful to consider the ways in which specific priorities can be identified to maximize optimal and positive outcomes. The first column of Table 21.4 provides a nonhierarchical list of six focus priorities for the promotion of global intelligibility and comprehensibility. Priority 1 is an emphasis on local phonological structures that enhance global intelligibility and comprehensibility. This means that not all local pronunciation problems deserve equal attention, especially since teachers usually have limited time to devote to pronunciation issues.

A problem we face at this point is the limited evidence about what those structures are. Our view is that teachers must base their curricula on the best available evidence about phonological phenomena that influence intelligibility and comprehensibility. In some cases, this may derive from well-constructed, controlled empirical studies (Priority 2), but in others it will come from Priority 3 – a well-motivated, theoretical framework. To ensure the most effective use of time, it is best to direct the most attention to pronunciation problems that are unlikely to resolve themselves in the long run without explicit intervention (focus Priority 4) and to devote class time to difficulties that are shared by many or all students in the class (focus Priority 5). On the latter point, suprasegmentals, which are often problematic for ESL learners from a range of L1 backgrounds, are good candidates for whole-class attention. Finally, it is essential that instructors evaluate individual students’ speech to identify idiosyncratic patterns that interfere significantly with intelligibility. These may be addressed by individualized activities, either through technology or in small groups with the teacher.

Implementation

The second column of Table 21.4 lists means for the implementation of focus priorities, whether in a stand-alone pronunciation class or an integrated skills class. Teachers often find that exclusively lock-step, teacher-centered formats do not lend themselves well to many aspects of pronunciation teaching. For this reason, careful classroom management is important for successful instruction. While shared problems may be addressed through whole-class activities, for idiosyncratic difficulties, students may benefit more from rotating through several work stations in the classroom, completing work independently or with each other, and spending some of that time directly with the teacher in a small group of students who share similar problems. In addition, available technology makes it much easier for teachers to assign beneficial self-administered pronunciation tasks (e.g., Thomson’s English Accent Coach 2012b; University of Iowa’s Phonetics: The Sounds of American English, n/d). A second concern in implementation is the need for attention to units of language at multiple levels. In particular, it is essential that learners gain practice and receive feedback on discourse-level productions, as well as on shorter utterances and word-level language. Work on discourse is particularly important in the teaching of intonation, an aspect of English phonology that is essential to spoken-language effectiveness. Levis and Pickering (2004) point out the value of evolving speech visualization technologies such as pitch displays for instruction. In particular, this type of software allows learners to record and compare their productions visually with models to facilitate acquisition of such intonational phenomena as paratones and tonal composition.

On a third point, a common critique of contemporary pronunciation teaching is that it is excessively formS-focused as opposed to form-focused. This opposition of approaches is not restricted to pronunciation, but has been widely discussed in the second language acquisition literature in connection with grammar (Ellis 2008). While formS-focused instruction draws students’ attention to specific structural details of decontextualized language, form-focused instruction addresses language issues within meaningful interaction; typically students are encouraged to “discover” regularities within the language. Despite several investigations of grammar acquisition, relatively little research has addressed the use of either formS or forms in the area of L2 pronunciation. Moreover, we cannot assume that findings pertaining to grammar are automatically applicable to pronunciation. It is clear, however, that L2 speakers who do not receive any explicit pronunciation instruction in their ESL training frequently show minimal pronunciation development over time, regardless of degree of exposure. For this reason, we advocate a balanced approach that includes some focus on formS that is critical for intelligibility, particularly to guide learners in developing the articulatory skills needed to produce new segments and prosody. However, to ensure transfer to real-world communication, a focus on formS needs to be matched with sufficient practice during communicative activities in which learners must cope with the higher cognitive demands of producing original output.

Conclusions

In attempting to outline a detailed approach to prioritizing issues for the pronunciation classroom, we continually face the problem of limited data and the consequent need to speculate rather than provide empirically grounded bases for recommendations to teachers. Progress has been made in understanding how intelligibility relates to other constructs in L2 speech. In addition, the growing body of empirical research on pronunciation has shed some light on how the most serious pronunciation problems can be identified and effectively addressed. However, much research remains to be done. In particular, further testing of theoretical notions such as functional load is needed to develop a clearer picture of the relationship between local and global intelligibility. In addition, we currently still suffer from a dearth of intervention studies whose findings can be directly applied in classrooms. Far more work of this type is needed to help us understand the diverse ways in which individual learners respond to pronunciation instruction and to help us determine how to efficiently and effectively use class time to address shared and individual problems. Finally, the potential of technology for pronunciation instruction has yet to be effectively tapped. Regrettably, one of the biggest problems in the use of much currently available software is that it comes without priority-setting features. The common one-size-fits-all approach in which practice is offered in “everything” is unhelpful to teachers and students who need to focus their attention on issues that will genuinely improve their communication skills. An important challenge, then, is to find ways to apply the individualized attention that technology offers so that time is not wasted and interactional benefits are maximized.

REFERENCES

1. Abercrombie, D. 1949. Teaching pronunciation. English Language Teaching 3: 113–122.
2. Abrahamsson, N. and Hyltenstam, K. 2009. Age of onset and nativeness in a second language: listener perception versus linguistic scrutiny. Language Learning 59(2): 249–306.
3. Bent, T. and Bradlow, A. 2003. The interlanguage speech intelligibility benefit. Journal of the Acoustical Society of America 114: 1600–1610.
4. Brown, A. 1991. Functional load and the teaching of pronunciation. In: Teaching English Pronunciation: A Book of Readings, A. Brown (ed.), 211–224, London: Routledge.
5. Burda, A.N., Scherz, J.A., Hageman, C.F., and Edwards, H.T. 2003. Age and understanding of speakers with Spanish or Taiwanese accents. Perceptual and Motor Skills 97: 11–20.
6. Catford, J.C. 1950. Intelligibility. English Language Teaching Journal 1: 7–15.
7. Catford, J.C. 1987. Phonetics and the teaching of pronunciation: a systemic description of English phonology. In: Current Perspectives on Pronunciation, J. Morley (ed.), 878–100, Washington, DC: TESOL.
8. Couper, G. 2006. The short and long-term effects of pronunciation instruction. Prospect 21: 46–66.
9. Derwing, T.M. 2008. Curriculum issues in teaching pronunciation to second language learners. In: Phonology and Second Language Acquisition, J.H. Edwards and M. Zampini (eds.), 347–369. Amsterdam: John Benjamins.
10. Derwing, T.M. and Munro, M.J. 1997. Accent, intelligibility, and comprehensibility: evidence from four L1s. Studies in Second Language Acquisition 19: 1–16.
11. Derwing, T.M., Munro, M.J., and Thomson, R.I. 2008. A longitudinal study of ESL learners’ fluency and comprehensibility development. Applied Linguistics 29(3): 359–380.
12. Derwing, T.M., Munro, M.J., and Wiebe, G.E. 1997. Pronunciation instruction for ‘fossilized’ learners: Can it help?” Applied Language Learning 8(2): 217–235.
13. Derwing, T.M., Munro, M.J., and Wiebe, G.E. 1998. Evidence in favor of a broad framework for pronunciation instruction. Language Learning 48(3): 393–410.
14. Derwing, T.M., Rossiter, M.J., and Munro, M.J. 2002. Teaching native speakers to listen to foreign-accented speech. Journal of Multilingual and Multicultural Development 23: 245–259.
15. Derwing, T.M., Munro, M.J., Foote, J.A., Waugh, E., and Fleming, J. 2014. Opening the window on comprehensible pronunciation after 19 years: a workplace training study. Language Learning 64: 526–548.
16. Ellis, R. 2008. The Study of Second Language Acquisition, 2nd Edition, Oxford: Oxford University Press.
17. Field, J. 2005. Intelligibility and the listener: the role of lexical stress. TESOL Quarterly 39: 399–423.
18. Flege, J.E. 1988. Factors affecting degree of perceived foreign accent in English sentences. Journal of the Acoustical Society of America 84: 70–79.
19. Flege, J.E., Munro, M.J., and MacKay, I. 1995. Factors affecting degree of perceived foreign accent in a second language. Journal of the Acoustical Society of America 97: 3125–3134.
20. Foote, J.A., Holtby, A., and Derwing, T.M. 2011. Survey of pronunciation teaching in adult ESL programs in Canada, 2010. TESL Canada Journal 29: 1–22.
21. Gass, S. and Varonis, E. 1984. The effect of familiarity on the comprehensibility of nonnative speech. Language Learning 34: 65–89.
22. Gimson, A.C. 1962. Introduction to the Pronunciation of English, London: Edward Arnold.
23. Hahn, L. 2004. Primary stress and intelligibility: research to motivate the teaching of suprasegmentals. TESOL Quarterly 38(2): 201–223.
24. Harding, L. 2011. Accent and Listening Assessment: A Validation Study of the Use of peakers with L2 Accents on an Academic English Listening Test. Frankfurt: Peter Lang.
25. Hayes-Harb, R., Smith, B.L., Bent, T., and Bradlow, A.R. 2008. The interlanguage speech intelligibility benefit for native speakers of Mandarin: production and perception of English word-final voicing contrasts. Journal of Phonetics 36: 664–679.
26. Holm, S. 2008. Intonational and durational contributions to the perception of foreign-accented Norwegian: An experimental phonetic investigation. Unpublished PhD dissertation, Department of Language and Communication Studies, Norwegian University of Science and Technology.
27. Kennedy, S. and Trofimovich, P. 2008. Intelligibility, comprehensibility, and accentedness of L2 speech: the role of listener experience. The Canadian Modern Language Review 64: 459–489.
28. Levis, J.M. 2005. Changing concepts and shifting paradigms in pronunciation teaching. TESOL Quarterly 39(3): 369–377.
29. Levis, J.M. and Cortes, V. 2008. Minimal pairs in spoken corpora: implications for pronunciation assessment and teaching. In: Towards Adaptive CALL: Natural Language Processing for Diagnostic Language Assessment, C.A. Chapelle, Y.-R. Chung, and J. Xu (eds.), 197–208, Ames, IA: Iowa State University.
30. Levis, J. and Pickering, P. 2004. Teaching intonation in discourse using speech visualization technology.” System 32: 505–524.
31. Munro, M.J. 2008. Foreign accent and speech intelligibility. In: Phonology and Second Language Acquisition, J.G.H. Edwards and M.L. Zampini (eds.), 193–218, Amsterdam: John Benjamins.
32. Munro, M.J. and Derwing, T.M. 1995a. Foreign accent, comprehensibility, and intelligibility in the speech of second language learners. Language Learning 45(1): 73–97.
33. Munro, M.J. and Derwing, T.M. 1995b. Processing time, accent, and comprehensibility in the perception of foreign-accented speech. Language and Speech 38(3): 289–306.
34. Munro, M.J. and Derwing, T.M. 2006. The functional load principle in ESL pronunciation instruction: an exploratory study. System 34: 520–531.
35. Munro, M.J. and Derwing, T.M. 2008. Segmental acquisition in adult ESL learners: a longitudinal study of vowel production. Language Learning 58: 479–502.
36. Munro, M.J., Derwing, T.M., and Holtby, A.K. 2012. Evaluating individual variability in foreign accent comprehension. In: Proceedings of the 3rd Pronunciation in Second Language Learning and Teaching Conference, September 2011, J. Levis and K. LeVelle (eds.), Ames, IA: Iowa State University.
37. Munro, M.J., Derwing, T.M., and Morton, S. 2006. The mutual intelligibility of L2 speech. Studies in Second Language Acquisition 28: 111–131.
38. Munro, M.J., Derwing, T.M., and Saito, K. 2013. English L2 vowel acquisition over seven years. In: Proceedings of the 4th Pronunciation in Second Language Learning and Teaching Conference, J. Levis and K. LeVelle (eds.), August 2012, Ames, IA: Iowa State University.
39. Nilsen, D.L.F. and Nilsen, A.P. 2010. Pronunciation Contrasts in English, 2nd edition, Long Grove, IL: Waveland Press.
40. Ou, S.-C., Yeh, R., and Chuang, Z.-L. 2012. Units of analysis, intelligibility evaluation and phonological cores of EIL. Poster presented at The 4th Annual Conference on Pronunciation in Second Language Learning and Teaching, Vancouver.
41. Perlmutter, M. 1989. Intelligibility rating of L2 speech pre- and post-intervention. Perceptual and Motor Skills 68: 515–521.
42. Rubin, D.L. 1992. Nonlanguage factors affecting undergraduates’ judgments of nonnative English-speaking teaching assistants. Research in Higher Education 33(4): 511–531.
43. Saito, K. and Lyster, R. 2012. Effects of form-focused instruction and corrective feedback on L2 pronunciation development of /ɹ/ by Japanese learners of English. Language Learning 62: 595–633. doi: 10.1111/j.1467-9922.2011.00639.x.
44. Schiavetti, N. 1992. Scaling procedures for the measurement of speech intelligibility. In: Intelligibility in Speech Disorders, R.D. Kent (ed.), 11–34, Philadelphia, PA: John Benjamins.
45. Smith, L.E. and Nelson, C.E. 1985. International intelligibility of English: directions and resources. World Englishes 3: 333–342.
46. Subtelny, J.D. 1977. Assessment of speech with implications for training. In: Childhood Deafness: Causation, Assessment, and Management, F.H. Bess (ed.), 183–194, New York: Grune & Stratton.
47. Swan, M. and Smith, B. (eds.). 2001. Learner English: A Teacher’s Guide to Interference and Other Problems, 2nd edition, Cambridge: Cambridge University Press.
48. Sweet, H. 1900. The Practical Study of Languages: A Guide for Teachers and Learners, New York: Holt.
49. Tajima, K., Port, R., and Dalby, J. 1997. Effects of temporal correction on intelligibility of foreign accented English. Journal of Phonetics 25: 1–24.
50. Thomson, R.I. 2012a. Improving L2 listeners’ perception of English vowels: a computer-mediated approach. Language Learning. doi: 10.1111/j.1467-9922.2012.00724.x.
51. Thomson, R.I. 2012b. English Accent Coach. Accessed October 24, 2012, http://www.englishaccentcoach.com.
52. Thomson, R.I., Nearey, T.M., and Derwing, T.M. 2009. A modified statistical pattern approach to measuring the crosslinguistic similarity of Mandarin and English vowels. Journal of the Acoustical Society of America 126: 1447–1460.
53. Trofimovich, P. and Baker, W. 2006. Learning second language suprasegmentals: effect of L2 experience on prosody and fluency characteristics of L2 speech. Studies in Second Language Acquisition 28: 1–30.
54. University of Iowa. n/d. Phonetics: The Sounds of Spoken English.http://www.uiowa.edu/~acadtech/phonetics/ Accessed July 9, 2013.
55. Varonis, E.M. and Gass, S. 1982. The comprehensibility of non-native speech. Studies in Second Language Acquisition 4: 114–136.
56. Voegelin, C.F. and Harris, Z.S. 1951. Methods for determining intelligibility among dialects of natural languages. Proceedings of the American Philosophical Society 95: 322–329.
57. Zielinski, B.W. 2008. The listener: no longer the silent partner in reduced intelligibility. System 36(1): 69–84.