Pronunciation Myths: Applying Second Language Research to Classroom Teaching

by Linda Grant

Overview

This volume was conceived as a "best practices" resource for pronunciation and speaking teachers in the way that Vocabulary Myths by Keith S. Folse is one for reading and vocabulary teachers. Like others in the Myths series, this book combines research with good pedagogical practices.

The book opens with a Prologue by Linda Grant (author of the Well Said textbook series), which reviews the last four decades of pronunciation teaching, the differences between accent and intelligibility, the rudiments of the English sound system, and other factors related to the ways that pronunciation is learned and taught.

The myths challenged in this book are:

§ Once you’ve been speaking a second language for years, it’s too late to change your pronunciation. (Derwing and Munro)

§ Pronunciation instruction is not appropriate for beginning-level learners. (Zielinski and Yates)

§ Pronunciation teaching has to establish in the minds of language learners a set of distinct consonant and vowel sounds. (Field)

§ Intonation is hard to teach. (Gilbert)

§ Students would make better progress if they just practiced more. (Grant)

§ Accent reduction and pronunciation instruction are the same thing. (Thomson)

§ Teacher training programs provide adequate preparation in how to teach pronunciation (Murphy).

The book concludes with an Epilogue by Donna M. Brinton, who synthesizes some of the best practices explored in the volume.

Product Details

ISBN-13:	9780472120703
Publisher:	University of Michigan Press
Publication date:	08/08/2014
Series:	The Myths Series
Sold by:	Barnes & Noble
Format:	eBook
Pages:	264
File size:	3 MB

Read an Excerpt

Pronunciation Myths

Applying Second Language Research to Classroom Teaching

---

By Linda Grant

The University of Michigan Press

Copyright © 2014 University of Michigan
All rights reserved.
ISBN: 978-0-472-12070-3

---

CHAPTER 1

MYTH 1

Once you have been speaking a second language for years, it's too late to change your pronunciation.

Tracey Derwing and Murray J. Munro

University of Alberta and Simon Fraser University

In the Real World

David Nguyen, originally from Vietnam, moved to Canada in 1980, a time when many Vietnamese people were fleeing their country. David was an engineer and, although it took a long time and a lot of hard work, his credentials were eventually recognized, and he was hired in a large engineering firm. His professional skills were very strong, but his employers often complained that they had difficulty understanding him, despite the fact that he had taken several ESL courses when he first arrived and had a good grasp of both spoken and written English. The problem, as they put it, was his "heavy accent."

Sixteen years after his arrival in Canada, David enrolled in a Clear Speaking course offered two evenings a week for twelve weeks at a local college. Along with his classmates, he received instruction intended to make him more intelligible. On the first night, the students were invited to participate in a study that would entail collecting samples of their English pronunciation at the beginning and end of the course. Like David, the other students had all been in Canada for extensive periods of time; the average length of stay was ten years. They were all well educated and ranged from high intermediate to very advanced in terms of English proficiency. Each student agreed to record speech samples in the first and last weeks of the course; they were offered an honorarium at the end of the study.

What the Research Says

What could David, after 16 years of living in an English-speaking city in Canada, realistically expect from thirty-six hours of instruction over twelve weeks? The conventional wisdom about immigrants like David is quite discouraging. A widespread assumption is that he would have fossilized, a term coined by Selinker (1972) to describe the process undergone by a second language (L2) speaker who is unlikely to show improvement in certain forms of the target language, regardless of instruction. Selinker's proposal is supported by a number of early pronunciation studies. Oyama (1976), for instance, examined the pronunciation of 60 Italian immigrants to the United States. Their ages on arrival ranged from six to twenty years, and they had lived in the U.S. for five to eighteen years. Two linguistically trained judges assessed their accentedness on a five-point scale. Oyama found that the immigrants who arrived at later ages had much stronger foreign accents than those who had come at an earlier age. Interestingly, length of time in the U.S. made no significant difference to degree of accentedness. Oyama concluded that pronunciation instruction in an L2 should take place when learners are young. Her finding has often been interpreted as indicating that older learners don't benefit from pronunciation instruction; in other words, they have "fossilized."

Another interpretation of fossilization is connected to the length of time an L2 learner has spent in the target language community. Research on naturalistic development of L2 pronunciation patterns has shown that experience in the second language environment does indeed have some impact on pronunciation, even though it is quite small. Moreover, most changes in the direction of the target language tend to occur within the first year in the second language environment (Flege, 1988; Munro & Derwing, 2008). These findings, along with those of Oyama (1976), suggest that L2 learners' productions will fossilize after even a relatively short period of residence in their new language environment. Thus, fossilization has been tied to both age and length of residence. Older learners are considered to have more difficulty modifying their L2 speech, and learners who have resided in the target language community for more than a year are considered to be likely candidates for fossilization.

Is Fossilization Restricted to L2 Speakers?

Although the concept of fossilization is usually discussed as an unwanted aspect of second language learning, even native speakers of a language often demonstrate a comparable resistance to change, despite extensive exposure to accepted norms. English language prescriptivists are fond of complaining about the "deterioration" of the language, citing mispronunciations that they find egregious. Some of these mispronunciations arise when a native speaker first encounters a word in written form and attributes a pronunciation to it that may conform with other similarly spelled words. For example, some people pronounce epitome with three syllables (/ 'ep tom/), rather than with four (/ 'pit mi/).Another common mispronunciation is the word heroine, the second syllable of which is mistakenly pronounced to rhyme with groin. Reading pronunciations like these seem remarkably resistant to change. Even after hearing the more accepted pronunciation, some speakers persist in using the one they learned on their own. Uneducated speakers may produce innovations such as drownded for drowned and spayeded for spayed. Other non-standard forms become so widespread that eventually they are accepted as alternative pronunciations. Some dictionaries, for instance, list expresso as an acceptable alternative to espresso and heighth as an alternative to the more common height. In the most extreme cases, a word can actually change in response to the new pronunciation. For example, apron came into existence because people interpreted a napron as an apron. The point to all of these examples is that native speakers, in spite of ongoing exposure to accepted forms, sometimes do not notice that these differ from their own productions. Thus, although fossilization is often regarded as a process restricted to L2 speakers, it appears in native speaker speech as well. In the absence of overt correction, native speakers do not necessarily change their mispronunciations.

The main fossilization difference between native speakers and second language speakers is the level of the errors. Native speaker mispronunciations are usually restricted to specific words. The same can happen with L2 learners, but they are also subject to systematic grammatical and phonological fossilization. A Japanese speaker who is unable to produce English / / (the first sound in the word run), for instance, will extend the error across a wide range of contexts.

Why Don't More Teachers Address Fossilized Pronunciation in Their L2 Classrooms?

Many researchers have bemoaned the fact that language instructors tend to shy away from teaching pronunciation. One key reason is the belief that pronunciation teaching is not effective. A discouraging study conducted by Purcell and Suter (1980) reinforced this sentiment. They examined the speech of 61 English learners from a variety of first language backgrounds. The authors collected information on several factors including age of arrival, length of residence in an English-speaking country, amount of English used in conversation, motivation, aptitude for oral mimicry, strength of concern for pronunciation, amount of general English instruction, and number of weeks focused specifically on pronunciation instruction. In their correlational analyses, Purcell and Suter (1980) concluded that four factors accounted for accuracy of pronunciation: first language, aptitude for oral mimicry, residency, and strength of concern for pronunciation. Notably, pronunciation instruction did not correlate significantly with accent. The authors concluded that the contributors to pronunciation accuracy are largely out of the control of a second language teacher, which led to the interpretation that formal pronunciation instruction is largely ineffective.

Second language instructors may also be reluctant to devote class time to pronunciation because of pedagogical theory regarding second language acquisition. During the audiolingual era of the 1950s and 1960s, pronunciation skills were a central aspect of L2 classrooms. Students were taught to mimic native speaker models as accurately as possible as a means of developing good habits of oral language production. The publication of Purcell and Suter's (1980) study coincided with a major shift in second language classrooms across North America from audiolingual and 'designer' methods to communicative language teaching (CLT). A basic premise of CLT was that with enough input, learners would gradually develop acceptable English pronunciation, but that any special pronunciation instruction was not only unnecessary, but unlikely to be effective. The CLT approach emphasized authentic use of language and moved away from repetition and mimicry and minimized corrective feedback, which was seen as disruptive to communication. Without formal pronunciation instruction, students in CLT classrooms were left to their own devices to change their oral productions in the direction of the target language. In the absence of explicit correction, many students exhibited fossilized patterns. It seemed, then, that pronunciation fell outside the responsibilities of the CLT classroom. A decade later, in a detailed overview of what research revealed about classroom teaching, Pica (1994) agreed with Purcell and Suter: "Precise pronunciation may be an unrealistic goal for teachers to set for their students and in their teaching" (p. 73).

A third reason for the neglect of pronunciation in the language classroom is the lack of formal training in pronunciation pedagogy available to teachers (see Murphy, Myth 7, for more on this topic). In the post-audiolingual period, only a handful of researchers and practitioners promoted the view that adult learners could indeed benefit from explicit pronunciation instruction. Acton (1984), for example, described the approach he used in working with "fossilized" adult learners. Similarly, Ricard (1986) outlined strategies and techniques used in an English pronunciation course for adults that she argued were successful. Meanwhile, pronunciation specialists such as Gilbert (1984) developed materials that could be used within the CLT framework. The impact on teaching, however, appears to have been somewhat limited, right into the twenty-first century, in part because second language instructors generally had little or no training in how to teach pronunciation and were therefore uncomfortable using these materials (Burgess & Spencer, 2000; Breitkreutz, Derwing, & Rossiter, 2002; MacDonald, 2002). Furthermore, articles that were accessible to teachers, such as those of Acton (1984) and Ricard (1986), were based on personal experience rather than empirical evidence that could be replicated by others.

In summary, research and pedagogical practice over the last two decades of the twentieth century conspired to marginalize pronunciation instruction. The empirical studies of pronunciation learning had a negative message. Language teaching shifted to an approach that seemed incompatible with pronunciation instruction, and publications with a positive message for teachers were based on the authors' personal experiences rather than verifiable data.

Is Pronunciation Instruction Effective?

When we began our own program of research on second language pronunciation in the 1990s, we were taken aback at the paucity of empirical research on the effectiveness of explicit instruction for adult learners on pronunciation. At that time, we could identify only a handful of published studies, many of which were relatively inaccessible to second language instructors (Derwing & Munro, 2005). One of the very few before and after studies was an investigation by Perlmutter (1989), suggesting benefits of pronunciation instruction for international teaching assistants in their first six months in the U.S. Regrettably, the lack of an uninstructed comparison group meant that the improvement noted in pronunciation may have been due to overall exposure within the English-speaking community rather than to the teaching intervention.

Since that time, other studies that have included non-instructed comparison groups have been conducted. Couper (2003, 2006) investigated the benefits of explicit pronunciation instruction aimed at certain features that he deemed most problematic for listeners. He administered pre- and post-tests to international university students who were enrolled in his pronunciation courses. Couper found in both cases that the overall number of errors was reduced as a result of the instruction. In his 2006 study, he included an uninstructed comparison group; that group showed no improvement over time. These studies suggest that second language learners' speech can be changed in reaction to targeted instruction.

If we consider second language learners' pronunciation to have a propensity to "fossilize" within the first year of residence in target language community (Flege, 1988), then the studies by Couper (2003, 2006) can be understood as applicable to improving fossilized pronunciation. In both studies, the learners' mean length of residence in New Zealand was 2.5 years, ranging from 0–8 years. Derwing, Munro, and Wiebe (1998) assessed the pronunciation improvement of three groups of high-intermediate ESL learners in Canada over a 12-week period. Most of these individuals had been in the country for longer than one year (on average 3.4 years with a range of 7 months to 15 years). One group received suprasegmental (prosodic) training, while a second group received instruction focused only on individual vowels and consonants (segmentals). The third group had no pronunciation-specific instruction. An important aspect of this study was the evaluation of progress in comprehensibility (how easy or difficult listeners perceived the second language speech to be) as well as accent (how much the second language speech differed from the listeners' own native variety of English) and fluency(how smooth and hesitation-free the speech flow was). When the speakers described a picture story, the suprasegmental group showed improvement in both comprehensibility and fluency, whereas the other groups did not; however, the suprasegmental group did not show any change in accent ratings, despite their improvement on the other dimensions. This outcome supports our multi-dimensional approach to L2 speech (Munro & Derwing, 1995; Derwing & Munro, 1997), which regards accent as only partially connected to factors such as comprehensibility and intelligibility (how understandable speech actually is). Table 1.1 illustrates the definitions of these speech dimensions and gives examples of how these dimensions can be measured.

What Really Matters in Pronunciation Instruction?

The findings of this study help us sort out some of the complexities that arise in understanding pronunciation research in general. Despite its possible usefulness for theoretical purposes, the study of accentedness is not very relevant to second language teaching. Far more important are the concepts of intelligibility and comprehensibility, both of which are strongly connected to communicative success. After all, the primary goal of most language learners is to make themselves understood and to understand other speakers. As our research has shown (Munro & Derwing, 1995), it is possible to have even a heavy accent and still be relatively easy to understand. Moreover, comprehensibility and intelligibility can improve even when there is no noticeable improvement in degree of accentedness. Seen in this light, the somewhat pessimistic view held by researchers and theorists who are concerned with the acquisition of native-like accuracy is essentially immaterial to second language teachers. What is important is to help learners to develop a comfortable intelligibility (Abercrombie, 1949), not the elimination of a foreign accent per se.

What Are the Mechanisms That Make Pronunciation Instruction Effective?

Because we do not yet fully understand all the factors that make second language instruction difficult for adults, we also do not have a complete grasp of the most effective techniques for teaching pronunciation. However, research has provided us with several useful insights. Some studies have indicated a relationship between learner perception and production. For instance, when Bradlow, Pisoni, Akahane-Yamada, and Tohkura (1997) trained Japanese speakers to perceive the English / / versus /l/ distinction, their production also improved, even though there was no oral component to the training. Thomson (2011) found a comparable effect of perceptual training on the production of English vowels.

(Continues...)

---

Excerpted from Pronunciation Myths by Linda Grant. Copyright © 2014 University of Michigan. Excerpted by permission of The University of Michigan Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---

Table of Contents

Contents Introduction Prologue to the Myths: What Teachers Need to Know - Linda Grant Myth 1. Once you have been speaking a second language for years, it’s too late to change your pronunciation. - Tracey Derwing and Murray J. Munro Myth 2. Pronunciation instruction is not appropriate or beginning-level learners. - Beth Zielinski and Lynda Yates Myth 3. Pronunciation teaching has to establish in the minds of language learners a set of distinct consonant and vowel sounds. - John Field Myth 4. Intonation is hard to teach. - Judy Gilbert Myth 5. Students would make better progress in pronunciation if they just practiced more. - Linda Grant Myth 6. Accent reduction and pronunciation instruction are the same thing. - Ron I. Thomson Myth 7. Teacher training programs provide adequate preparation in how to teach pronunciation. - John Murphy Epilogue to the Myths: Best Practices for Teachers - Donna M. Brinton Index