Why is Pronunciation So Difficult to Learn ?

In many English language classrooms, teaching pronunciation is granted the least attention. When ESL teachers defend the poor pronunciation skills of their students, their arguments could either be described as a cop-out with respect to their inability to teach their students proper pronunciation or they could be regarded as taking a stand against linguistic influence. If we learn a second language in childhood, we learn to speak it fluently and without a ‘foreign accent’; if we learn in adulthood, it is very unlikely that we will attain a native accent. In this study, the researchers first review misconceptions about pronunciation, factors affecting the learning of pronunciation. Then, the needs of learners and suggestions for teaching pronunciation will be reviewed. Pronunciation has a positive effect on learning a second language and learners can gain the skills they need for effective communication in English.


Introduction
General observation suggests that it is those who start to learn English after their school years are most likely to have serious difficulties in acquiring intelligible pronunciation, with the degree of difficulty increasing markedly with age.This difficulty has nothing to do with intelligence or level of education, or even with knowledge of English grammar and vocabulary.Of course there is no simple answer to why pronunciation is so difficult to learnindeed there is a whole range of theoretical perspectives on the question.What is generally accepted among psycholinguists and phonologists who specialized in this area, is that the difficulty of learning to pronounce a foreign language is cognitive rather than physical, and that it has something to do with the way 'raw sound' is categorized or conceptualized in using speech.Many learners of English as a second language have "major difficulties" with English pronunciation even after years of learning the language.This often results in them facing difficulties in areas such as finding employment.Hinofotis and Baily (1980, pp. 124-125) notes that "up to a certain proficiency standard, the fault which most severely impairs the communication process in EFL/ESL learners is pronunciation", not vocabulary or grammar.This is true despite the fact that research by the likes of Davis (1999), for example, reveals that an area of concern and indeed one of the top priorities of ESL students after completing elementary English courses is pronunciation.It is important at this point in time to make a distinction between speaking and pronunciation as it is sometimes wrongly applied interchangeably.Pronunciation is viewed as a sub-skill of speaking.Generally, if we want to change the way a learner pronounces words, we have to change the way they think about the component sounds of those words.This goes not just for individual sounds, but for bigger elements of speech, such as syllables, stress patterns and rhythm.Despite this, the teaching of pronunciation remains largely neglected in the field of English language teaching.In this study, the researchers discuss common misconceptions about pronunciation, factors affecting the learning of pronunciation.Then, they review the needs of learners, suggestions for teaching pronunciation.

Some common misconceptions about second language pronunciation
It is widely believed that pronunciation skills are related to musical skills.However no link between musical ability and pronunciation ability has been demonstrated, and there are large numbers of people who have one of these 'natural talents' but not both.Second language pronunciation is a cognitive skill for which some people may have more natural aptitude and / or interest and motivation than others, but which everyone can learn to a certain degree if given appropriate opportunities.The main problem that second language learners have with pronunciation has to do with their need to change a conceptual pattern appropriate for their first language that they have internalized in childhood.It is not the case that learners are best helped if they are able to 'see' speech, whether in articulate or acoustic form.Learners need help in categorizing or conceptualizing sounds in a way appropriate to English.Simply seeing a speech-wave or a diagram of the articulation of a sound, however 'animated' and however accurate, will not help them unless they are also helped to understand what features of the sound are significant and given appropriate ways of thinking about the sound so that they can reproduce it.In fact it will be difficult for most learners -indeed for most teachers -to relate a speech wave or articulate diagram to the auditory quality of the sounds -for exactly the same reason that instruction in terms of the detailed physiology of required shoulder movements is unlikely to help an aspiring tennis player perfect her stroke.In the case of the tennis player, what helps is instruction in how to think about the actions, e.g.'think about hitting it beyond the baseline', 'keep your eye on the ball' (Baker, 1981).Since people generally think about sounds in terms of their auditory quality, rather than directly in terms of their articulation or acoustics, the key is to find ways of describing the auditory quality of sounds that makes sense to the learner.There is a major role for the use of computers in helping learners with pronunciation -but it is not the role of displaying speech-waves with no guidance as to how they should be read.It is not the case that learners have an accent primarily because they 'transfer' the sounds of the native language to English.The notion of 'transfer' as the 'cause' of accents and the key to helping learners with pronunciation has been seriously questioned by specialists for at least two decades.Though there is some validity to the 'transfer' idea, it is only useful in an elaborated form which requires a good understanding of its limitations and ramifications.A simplistic idea that learners are transferring sounds from their native language to the new language is a hindrance rather than a help.It is unfortunate that so many teachers still hold so strongly to a simple notion of transfer (Gass et al., 1989;Bohn, 1995).
It is not the case that accent is caused by an inability of speakers of other languages to produce the sounds of English.This is not to say that there are not individual sounds in English, or more especially combinations of sounds, that are difficult for learners from different backgrounds to produce.It is to say that this difficulty is a relatively minor aspect of intelligibility, and certainly not the main cause of the accent.Firstly, individual sounds are not in themselves very important to intelligibility.After all, many native speakers, or fluent non-native speakers (NNSs), pronounce individual sounds differently from the norm, with no problems for intelligibility.A learner with good stress and intonation and poor pronunciation of, say, 'th', is very easy to understand.Secondly, in many of the cases in which a learner seems to have trouble pronouncing a particular sound, it is easy to demonstrate that the learner commonly pronounces a perfectly acceptable version of the sound in another context.Consider for example a German learner of English who has difficulty with the 'v' sound in 'very' etc -and yet the sound of the German 'w' is virtually identical to English 'v'.The same goes for the classic case of 'r' and 'l': it is certainly not the case that learners cannot produce these sounds; in almost all cases, they can produce perfectly acceptable versions of both sounds.The problem is that they do not have concepts of them as separate sounds, but rather think of them as indistinguishable variants of a single sound.Another classic case is the English vowel sound of words like 'bird', 'term', with which Japanese and other learners often have a lot of difficulty: that difficulty is not in producing the sound which they can easily do if thinking about it as a non-speech sound.The difficulty is in developing a concept of the sound that they can use as a vowel in words.As a final example, consider the notorious 'final consonant problem'.Even this is not primarily a problem of articulation.Consider a sentence like 'Put it back up' -bound to be difficult for speakers of languages like Thai which have a limited range of consonants in word final position.The difficulty such a learner has in imitating an English pronunciation of this sentence is caused by the cognitive interpretation of the relevant sounds as being 'word final'.The sentence as it is produced is a continuous flow of vowels and consonants.The pronunciation of the last three sounds of this sentence, '…ck up,' is highly similar to that of the word 'cup' -and yet such a learner will have great difficulty with the former and little difficulty with the latter (Beebe, 1987).

Factors affecting the learning of pronunciation
In this section, the researchers mention some of the important factors that affect the learning of pronunciation.They are as follows:

Accent
An accent is "the cumulative auditory effect of those features of pronunciation that identify where a person is from, regionally or socially" (Crystal, 2003, p. 3).Accentedness, a "normal consequence of second language learning" (Derwing & Munro, 2005, p. 383), is a "listener's perception of how different a speaker's accent is from that of the L1 community" (p.385).Many adult learners of English have foreign accents that identify them as nonnative speakers.Some linguists support the idea, known as the Critical Period Hypothesis, that a learner needs to begin learning the language before age 7 to develop native-like pronunciation (Lenneberg, 1967).However, more recent research suggests that environment and motivation may be more important factors in the development of native-like pronunciation than is age at acquisition (Marinova-Todd, Marshall, & Snow, 2000).An understanding of the features of learner accents and their impact on intelligibility can help teachers identify and address characteristics of learner pronunciation (Derwing & Munro, 1997).The primary aim is that students be understood.Good pronunciation is needed for this, but a "perfect accent" is not (Harmer, 1991).

Stress, intonation, and rhythm
Munro and Derwing (1999) observed that even heavily accented speech is sometimes intelligible and that prosodic errors (i.e., errors in stress, intonation, and rhythm) appear to affect intelligibility more than do phonetic errors (i.e., errors in single sounds).For this reason, pronunciation research and teaching focus both on the sounds of language (vowels and consonants) and on supra-segmental features-that is, vocal effects that extend over more than one sound-such as stress, sentence and word intonation, and speech rhythm (Crystal, 2003;Low, 2006;Munro & Derwing, 1999).Languages have been classified as either stress timed or syllable timed.In stress-timed languages (e.g., British and American English, German, Dutch, Thai), "stressed syllables fall at regular intervals throughout an utterance" (Crystal, 2003, p. 245), and rhythm is organized according to regularity in the timing of the stressed syllables.That is, the time between stressed syllables is equal because unstressed syllables are spoken more quickly and vowel reduction occurs.For example, the sentence "Tom runs fast" is made up of three stressed syllables, as indicated by the letters in boldface.The sentence "Meredith can run fast" is made up of six syllables, but only three of them are stressed.The unstressed syllables -e-, -dith, and can are spoken quickly and vowel reduction occurs, so the time between the stressed syllables tends to be equal, and both sentences take approximately the same amount of time to say.In syllable-timed languages (e.g., some nonnative varieties of English, such as Singapore and Malaysian English, and languages such as Tamil, Spanish, and French), syllables are said to be equal in timing (Crystal, 2003).All syllables are nearly equally stressed, vowel reduction does not occur, and all syllables appear to take the same amount of time to utter.
Recent phonetic research has shown that languages cannot be strictly classified as syllable timed or stress timed.A more accurate description is that they are stress based or syllable based; that is, they are not completely in one category or the other, but tend to have more stress-timed or syllable-timed features (Low, 2006).Stress-based rhythm is achieved through the presence of reduced vowels for unstressed syllables in a sentence.Function words (e.g., articles, helping verbs, prepositions) typically have reduced vowels instead of full ones, and the reduced vowel version is known as a weak form.For example, in the sentence "Bob can swim," the words Bob and swim have the major stress, and can, which is unstressed, is pronounced [kin]-its weak form.The distinction between stress-and syllable-based languages is important, especially if an adult English language learner speaks a first language that is different rhythmically from stress-based British or American English.An understanding of whether a learner's first language is stress based or syllable based will help a teacher plan appropriate pronunciation exercises.In examining the role of stress-"the degree of force used in producing a syllable" (Crystal, 2003, p. 435)-in intelligibility, Field (2005) asked trained listeners to transcribe recorded material when the variables of word stress and vowel quality were manipulated.He determined that when word stress is erroneously shifted to an unstressed syllable, without a change in vowel quality, utterances are significantly less intelligible than when only vowel quality is manipulated.Native and nonnative English speakers responded similarly when judging the intelligibility of words with misplaced word stress.O'Brien (2004) reported the results of research on the importance of stress, intonation, and rhythm for a native-like accent in German.Native speakers of German were asked to rate American university students reading aloud in German.It was found that the native speakers focused more on stress, intonation, and rhythm than on individual sounds when rating speech samples as native-like.Implications of this research for classroom instruction are that teachers need to spend time teaching learners the rules for word stress, intonation, and rhythm in English, as well as focusing on individual sounds that may be difficult for the learners in their classes.

Motivation and exposure
Along with age at the acquisition of a language, the learner's motivation for learning the language and the cultural group that the learner identifies and spends time determine whether the learner will develop native-like pronunciation.Research has found that having a personal or professional goal for learning English can influence the need and desire for native-like pronunciation (Bernaus, Masgoret, Gardner, & Reyes, 2004;Gatbonton et al., 2005;Marinova-Todd et al., 2000;Masgoret & Gardner, 2003).The review by Marinova-Todd et al., (2000) of research on adult acquisition of English concluded that adults can become highly proficient, even native-like speakers of second languages, especially if motivated to do so.Moyer (2007) found that experience with and positive orientation to the language appears to be important factors in developing native-like pronunciation.In a study of learners of Spanish, Shively (2008) found that accuracy in the production of Spanish is significantly related to age at first exposure to the language, amount of formal instruction in Spanish, residence in a Spanish-speaking country, amount of out-of-class contact with Spanish, and focus on pronunciation in class.Therefore, in addition to focusing on pronunciation and accent in class, teachers should encourage learners to speak English outside the classroom and provide them with assignments that structure those interactions.

Attitude
It seems as though some learners are more adept at acquiring good pronunciation.Even within one homogenous classroom, there is often a large discrepancy among the pronunciation ability of the students.This phenomenon has lead many researchers to study the personal characteristics of the learners that contribute to their success in foreign language acquisition.In a study on pronunciation accuracy of university students studying intermediate Spanish as a foreign language, Elliot (1995) found that subjects' attitude toward acquiring native or near-native pronunciation as measured by the Pronunciation Attitude Inventory (PAI), was the principal variable in relation to target language pronunciation.In other words, if the students were more concerned about their pronunciation of the target language, they tended to have better pronunciation of the target allophones (Elliot, 1995).This study echoed earlier research done by Suter (1976), which found that students who were "more concerned" about their pronunciation (p.249) had better pronunciation of English as a Second Language (Elliot, 1995).When discussing the attitude of the second language learners in relation to their pronunciation and second language acquisition, it is necessary to note the work done by Schumann (1986) on acculturation and its role in the process of language learning.His acculturation model defines that learners will acquire the target language to the degree that they acculturate (Celce-Murcia, et al., 1996).According to Schumann, acculturation refers to a learner's openness to a target culture as well as a desire to be socially integrated in the target culture.His research (1976,1986) on acculturation examines the social and psychological integration of immigrant students as a predictor of the amount of English language they acquire and use.Schumann maintains that the acquisition and use of English is a measure of the degree to which students have become acculturated to the host culture.Acculturation, according to Schumann (1986), refers to the social and psychological contact between members of a particular group and members of the target culture.The more interaction (i.e., social/psychological closeness) a group has with the target group, the more opportunities will result for the group to acquire and use English.Conversely, less interaction (i.e., social/psychological distance) results in less acquisition and use of English.The group's amount of contact with the target culture has an effect on the amount of English acquired and used.Sparks and Glachow's work (1991) on personality found similar results.They state that students with motivation to learn with positive attitudes towards the target language and its speakers were more successful than were students with less positive attitudes.They refer to Gardner and Lambert's research on motivation wherein two types are highlighted.The first type of motivation is instrumental, which is motivation to learn the L2 for the value of linguistic achievement.Second is integrative motivation, which describes the desire to continue learning about the second language culture.According to Gardner and Lambert students with integrative motivation would be expected to work harder to develop communication skills in the second language because they are more likely than their less interested counterparts to seek out native speakers of the language.

Instruction
Foreign language instruction generally focuses on four main areas of development: listening, speaking reading and writing.Foreign language curricula emphasize pronunciation in the first year of study as it introduces the target language's alphabet and sound system, but rarely continues this focus past the introductory level.Lack of emphasis on pronunciation development may be due to a general lack of fervor on the part of the second language acquisition researchers, second language teachers and students, that pronunciation of a second language is not very important (Elliot, 1995).Furthermore, Pennington (1994) maintains that pronunciation which is typically viewed as a component of linguistic rather than conversational fluency, is often regarded with little importance in a communicatively oriented classroom (Elliot, 1995).According to Elliot (1995), teachers tend to view pronunciation as the least useful of the basic language skills and therefore they generally sacrifice teaching pronunciation in order to spend valuable class time on other areas of language.Or maybe, teachers feel justified neglecting pronunciation believing that for adult foreign language learners, it is more difficult to attain target language pronunciation skills than other facets of second language acquisition.Possibly, teachers just do not have the background or tools to properly teach pronunciation and therefore it is disregarded (Elliot, 1995).
Teachers have taught what they thought was pronunciation via repetition drills on both a discrete word or phrase level, or give the students the rules of pronunciation like the vowel in a CVC pattern, when given an e at the end, says its name.For example, when an e is added to the word bit (CVC) the pronunciation of the "short i", becomes long and therefore "says its name".This type of instruction is meant to help students with decoding words for the purpose of reading rather than pronunciation.For example, students are rarely given information about the differences between fricatives and non-fricative continuants, or the subtleties between the trilled or flapped /r/ between Spanish and English (Elliot, 1995).This particular information is often left up to the students to attain on their own.Researchers have explored the question of whether explicit instruction helps these second language learners.Such studies have generated inconsistent results.Suter (1976) reported an insignificant relationship between formal pronunciation and students' pronunciation of English as a Second Language (Elliot, 1995).Murakawa (1981) found that, with 12 weeks of phonetic instruction, adult L2 learners of English can improve their allophonic articulation (Elliot, 1995).Nuefield and Scheiderman (1980) reported that adults are able to achieve near native fluency and it can be developed in a relatively short time without serious disruption to the second language teaching program with adequate pronunciation instruction (Elliot, 1995).It is necessary to note at this point that even though there seems to be quite a contradiction in the range of results presented, the diversity of those results may be due to the differing designs of the particular experiments.Some pronunciation studies focus specifically on the instruction of supra-segmental.Derwing, Munro and Wiebe (1997) conducted research in which ESL learners who had been studying for an average of ten years, participated in a speaking improvement course that focused on the supra-segmental features of pronunciation (e.g.stress, rhythm, intonation).Thirty-seven native listeners transcribed speech samples (true/false sentences) taken at the beginning of a 12-week course in order to assess the learners' intelligibility.Each sample was rated in order of comprehensibility and degree of accentedness.In the end, there was a significant improvement in the intelligibility, and better ratings over time of comprehensibility and accentedness.They showed that 30 language learners could alter their pronunciation in a reading task (Derwing & Rossiter, 2003).

Age
The influence of age on language acquisition and specifically pronunciation may make adults find pronunciation more difficult than children do and that they probably will not achieve native-like pronunciation.According to the "Critical Period Hypothesis" proposed by Lenneberg (1967) there is a biological or neurological period which ends around the age of 12; after this period it becomes extremely difficult to attain the complete mastery of a second language especially pronunciation.Conversely, Bialystock (1997), and Bongaerts, Planken and Schils (1997), among others have shown that adult learners are capable of achieving native-like in an L2.However, the degree of pronunciation accuracy, according to Avery and Ehrlich (1992), varies considerably from one individual to another.To them, this discrepancy in pronunciation among adult learners means that ESL classroom time can profitably be devoted to improving students' pronunciation.

Personality
Non-linguistic factors related to an individual's personality and learning goals, attitude towards the target language, culture, native speakers, and type of motivation which are beyond the teacher's control (Miller, 2003), all have their share in the development of pronunciation skills.In addition, the degree of exposure to and use of the target language can support or impede pronunciation skills development.For example, learners who are outgoing and confident and get involved in interactions with native speakers are liable to practice their foreign language pronunciation (Avery & Ehrlich, 1992).Conversely, some learners feel uncomfortable trying out new speech rhythm and melody patterns (Miller, 2003), while others feel stupid pronouncing "weird" sounds, and with time, they decide that it is futile and impossible to learn English pronunciation (Laroy, 1995).In this respect, Miller (2003) believes that changing -and not changing -speech patterns is affected by how much responsibility the student takes, how much the student practices outside of class, and how ready the student is.

Mother tongue influence
Avery and Ehrlich (1992) claim that the sound pattern of the learner's first language is transferred into the second language and is likely to cause foreign accents.The mispronunciations of words by nonnative speakers reflect the influence of the sounds, rules, stress, and intonation of their native language.For example, nonnative speakers' production of English rhythm was investigated in several studies (Wenk, 1985;Machizuki-Sudo, Kiritani, 1991).These researchers concluded that the transfer from the learners' native language influenced their production of English-like stress alternation across a phrase.In this respect, Avery and Ehrlich point out that the sound system of the native language can influence the learners' pronunciation of a target language in at least three ways.First, when there is a sound in the target language which is absent from the learners' native sound inventory, or vice versa, learners may not be able to produce or even perceive the sound(s).Second, when the rules of combining sounds into words (i.e., phonotactic constraints/rules) are different in the learners' mother tongue from those of the target language, they cause problems for learners because these rules are language specific as they vary from one language to another.Thirdly, since the rhythm and melody of a language determine its patterns of stress and intonation, learners may transfer these patterns into the target language.In summary, while there are other factors that influence the learner's L2 pronunciation acquisition (Graham, 1994), the factors reviewed above may help ESL/EFL teachers consider what learners from different backgrounds are likely to face when learning English as a second language.These factors would enable the teachers to identify the difficulties in the pronunciation of the target language experienced by non-native speakers in order to help them overcome their foreign accent and consequently improve their pronunciation.In addition, they would also enable teachers to provide efficient pronunciation instruction and design their teaching methodology according to students' needs.

What learners need?
Learners need to be taught pronunciation, as all other aspects of language, in a communicative method.Though communicative methods for teaching vocabulary, grammar and pragmatics have been around for decades, there has been little development of a communicative method for pronunciation teaching.We consider here some aspects of what is involved in a communicative approach to pronunciation.One thing learners need is teachers with confidence to assure them that the communicative methods do fulfill their needs.Some learners believe that what they need is instruction in the articulation of specific English sounds.Some teachers justify a focus on articulation with reference to the learners' desire for this information.As a whole, what learners need are as follows:

Conversation
What learners most want (Willing, 1993;Volkoff & Golding, 1998) -and what will help them most -is plenty of authentic conversation practice (Burns & Joyce, 1997), supplemented by expert guidance on how to understand and correct their mistakes.Teachers can help with strategies for how to initiate and maintain conversation with native speakers outside the classroom, but ultimately this is something that learners have to do themselves.They can be greatly helped or hindered in this by the attitudes of the native speakers with whom they interact.Native speakers who encourage conversation, and are not themselves awkward in cross cultural communication, are one of the biggest boons a learner can have.

Drilling
Learners also need considerable drilling and repetition -but this must directly exercise the speech that they will actually use in real life.Old fashioned drilling of sounds and minimal pairs or more modern practice with chants and tongue twisters is useful only in so far as it is directly related in the learners' minds with the speech that they will actually use outside the classroom.

Expert guidance
Essentially what learners need to do to pronounce a new language in a way that is easily intelligible to its native speakers is to stop thinking about speech in terms of the categories of their first language and start thinking about it in terms that are appropriate to the new language.If they can do this, even though the exact realization of some of the sounds is likely to be a bit 'foreign', their meaning will be evident.Learners are most helped by teachers who themselves can appreciate and imaginatively explore what the sounds seem like to learners, gradually leading them to more appropriate ways of thinking about English pronunciation.Indeed some of the most gifted teachers are probably those who are themselves good at pronunciation and have an openness to hearing sounds in different ways but this in itself is not enough for teachers.It is also necessary to be able to articulate what one does.This requires understanding of cross-language -as well as English -phonetics and phonology, and of speech perception and production (psycholinguistics).It also requires an ability beyond simple reproduction of this knowledge in technical terms which are unlikely to be meaningful to learners.Where a learner has difficulty in pronouncing specific sounds or sound sequences of English, they need appropriate description of how to think about the sounds in terms they can understand and 'latch on to'.They need to be terms based on the way the learner thinks about the sounds, not the way English speakers do.A variety of 'tricks' can be found for most learners (Burns & Joyce, 1997).

Critical listening
Learners need ample opportunity to listen to their own speech and that of fellow learners in comparison with that of native speakers, and to learn to distinguish the aspects of learner pronunciation that make comprehension difficult for NSs.Listening to your own speech as you are speaking is very difficult.So is discussion of particular aspects of the pronunciation of a phrase or sentence which has just disappeared into thin air.For these reasons it is essential for learners and teachers to work with recorded voices so that the speech they are discussing is external to both of them, and can be referred to objectively without distortion.Computer technology makes this type of recording and play back extremely easy.

Suggestions for teaching pronunciation
In looking at recommendations for teaching pronunciation, the researchers describe them within the context of the communicative method of teaching English that is employed in most ESL classrooms.

Curriculum design
Referring to curriculum and syllabus designs, Morley (1998) submits that ESL programs should start by "establishing long range oral communication goals and objectives" that help identify pronunciation needs as well as speech functions and the contexts in which they might occur.These goals and objectives should be realistic, aiming for "functional intelligibility (ability to make oneself relatively easily understood), functional communicability (ability to meet the communication needs one faces), and enhanced confidence in use" and they should be the result of a careful analysis and description of the learners' needs.Bray (1995, p. 3) observes that beginning in the late 1970s, several teachers/theorists took a stand by suggesting that at a very basic level if communicative competence was the goal of language learning, then it would have as one of its essential components, intelligible pronunciation.Intelligibility rather than the native-like competence valued in traditional approaches became the goal of phonological instruction.Therefore, teaching speech from the perspective of supra-segmental seems indispensable within the communicative approach to teaching ESL.However, Bray (1995, p. 3) adds that although many theorists began to make a case for the role of supra-segmental phonology in communication on paper, "many teachers continue with a limited conception of the role of explicit phonological instruction in the language classroom."We can say that curriculum and syllabus designers need to focus on the supra-segmental features of pronunciation.In making their case for emphasis in teaching rhythm to ESL learners in China for example, Chen et al., (1996) discovered that Chinese students were not aware of the difference between the rhythm of the syllable-timed Chinese language and the stress timed English language and therefore drawing their attention to this supra-segmental feature helped significantly in improving their communicative ability.

Academic research and classroom experiments
Fraser (2000b, p. 5) notes that there is currently "a dearth of reliable research-based information about what works and what doesn't in pronunciation teaching."She adds that there is a need to increase the amount of academic research on these topics as well as to "increase the research orientation of teachers and their opportunities to contribute to serious research because teachers are in a position to provide essential information to linguistics" (Fraser, 2000b, p. 5).One particular area that needs attention is in the area of assessing ESL pronunciation.Without reliable assessment and diagnosis tools, it is very hard to quantify the effectiveness of methods or materials and get beyond opinion-based debate.Teachers are always experimenting with a variety of teaching methods and improvise and even improve on previous techniques.This tradition has to continue especially in relation to teaching pronunciation.Methods that have been tried and tested should be converted into research papers that can be shared with others.Through such methods, Bray (1995) for example describes the use of limericks in the ESL classroom and provides an effective technique for using limericks to help address problems related to supra-segmental features such as stressed, unstressed, and stress-timed rhythm.Chen et al., (1996) present several techniques and tools for teaching word rhythm including the use of visual effects for teaching word stress, auditory techniques such as clapping to differentiate between stressed and unstressed syllables, the use of rubber bands as a visual image for length variation in syllables and stress matching games.Makarova (1996) addresses the question of teaching pronunciation to large groups of students when ideally it requires close individual interaction between teacher and students.She contends that it is possible to get feedback from large number of students and enhance student motivation by applying some less traditional techniques like using phoneme cards, pronunciation-based quiz games, utilizing "sign language" and employing materials prepared by students such as tongue twisters and limericks.

Improved training for teachers
Existing teachers should be able to receive professional development in pronunciation teaching (on a voluntary but properly funded basis), and trainee teachers should receive such tuition as part of their teaching courses.Teachers should receive a range of different stories from different pronunciation specialists.There is a need for a more coordinated approach.It is recommended that there should be a conference of people interested in the area of ESL pronunciation to discuss a range of issues.It seems likely that it would be possible to integrate a range of existing practices into a coordinated approach to pronunciation teaching, or at least to identify opposing schools of thought which can be articulated and evaluated so as to allow teachers and teacher trainers to choose which they wish to adhere to (Forman, 1993).

Provision of materials and courseware for teachers and learners
The lack of suitable materials for teachers, teacher trainers and learners has been commented on at several points above.Though there are some highly commendable materials, there is need for much more to cover the wide range of needs.Ideally this should be based on a sound foundation of well-documented research on 'what works' in pronunciation teaching.Computer disks are particularly well suited as a medium for imparting information about pronunciation and pronunciation teaching, and are also a useful way of up-skilling teachers and learners in computer use.Of course there will always be a place for written materials.It would be good to develop some informal endorsement scheme, whereby books and computer disks which adhered to basic principles of linguistics and language teaching could be easily distinguished from those that do not -as an aid to learners and other consumers.It would also be good to publicise information in the media about what sorts of materials are helpful and why.This would serve useful functions of gradually dispelling myths, and attracting students and scholars in related fields to the topic of pronunciation.

Increased research on pronunciation teaching methodology
It will be clear that there is a burning need for an increase in the amount of serious research at all levels into a wide range of issues to do with ESL pronunciation teaching.The first priority is development of a range of assessment tools to allow methods and policies to be assessed for their effectiveness.While the reasons people have for opposing the objective assessment of learners' pronunciation are appreciated, it is really impossible to improve a system that allows no proper benchmarking or analysis.With an appropriate assessment tool, an early priority would be a set of benchmarking studies, to provide answers to such questions as 'How much improvement is possible, or realistic, to expect from a pronunciation class over a given period?' Next a series of studies should investigate the relative effectiveness of different methods and materials with different types of learners (Macdonald et al., 1994, Munro & Derwing, 1995).An early opportunity should be taken to investigate empirically a range of questions which are the topic of debate among teachers, but have never been properly tested, such as: is it better to focus on teaching stressed syllables before teaching unstressed syllables; is it better to represent pronunciation for learners with symbols of the international phonetic alphabet, or with ordinary English spelling conventions.Similarly, experiments to investigate the relative effects of common pronunciation errors on ESL English listeners would allow proper planning of pronunciation curricula.One particular issue that needs urgently to be addressed is that of the relationship between generic pronunciation tuition and integrated training.There is some inconsistency about the relationship between these -on the one hand there is a call for integrated training to provide workers with the specific skills needed in their job.On the other there is a call for generic skills development, allowing employers to hire highly adaptable staff.On the face of it, it would seem that pronunciation would be an ideal candidate for generic training, as improved oral communication skills in one area are very likely to translate into improvements in other areas.It would be useful to demonstrate this -as part of a more general demonstration of the value of ESL pronunciation training to the workplace.

Methods and materials development
Teachers need a greater appreciation of the pronunciation difficulties faced by learners of ESL and the reasons for these difficulties, and a simple framework for understanding the situation of the second language learner.Lambacher (1999, p. 138) notes that research in pronunciation has revealed that "difficult L2 contrasts (not just suprasegmentals) can interfere with intelligibility and a need therefore exists within the pronunciation curriculum to address the problems of L2 learners in identifying and producing difficult L2 speech contrasts."Computer Assisted Instruction (CAI) for pronunciation can be an effective tool as it "provides electronic visual feedback (EVF), which can help meet this essential need by showing the exact sound features that learners produce and thereby drawing attention to changes that they need to make" (Lambacher, 1999, p. 138).

Conclusion
Pronunciation can be one of the most difficult parts for a language learner to master and one of the least favorite topics for teachers to address in the classroom.There are many reasons why the teaching of ESL pronunciation is currently less than optimally effective, and certainly it is wrong to blame anyone group, whether teachers, pronunciation specialists, or academics.Second language pronunciation is a topic of great theoretical interest and practical relevance which unfortunately has been out of fashion for some decades.It seems that a few well-publicized interesting developments would help to swing this topic back into fashion among a range of people with relevant skills and interests and put EFL learners in a position to contribute impressively to worldwide developments in this area.EFL/ESL teachers should focus on the students' needs, level and ability, incorporate pronunciation into their oral skills and other classes and focus on both segmental and supra-segmental features whenever there is opportunity and time.Pronunciation must be viewed as more than correct production of individual sounds or isolated words.Instead, it must be viewed as a crucial and integral part of communication that should be incorporated into classroom activities.Teachers can help students by highlighting elements such as sounds, syllables, stress and intonation.Once the students understand the functions of these elements, they will know what to focus on and can build upon this basic awareness.Teachers can actively encourage the students' actual production, build pronunciation awareness and practice through classes gradually building skills in listening and speaking in both formal and informal situations.It can be concluded that with careful preparation and integration, pronunciation can play an important role in supporting the learners' overall communicative power.