Pronunciation Deviations in Second Language Zulu Speakers

COPYRIGHT AND CITATION CONSIDERATIONS FOR THIS THESIS/ DISSERTATION

o Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

o NonCommercial — You may not use the material for commercial purposes.

o ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.

How to cite this thesis

Surname, Initial(s). (2012) Title of the thesis or dissertation. PhD. (Chemistry)/ M.Sc. (Physics)/ M.A. (Philosophy)/M.Com. (Finance) etc. [Unpublished]: University of Johannesburg. Retrieved from: https://ujdigispace.uj.ac.za (Accessed: Date).

PRONUNCIATION DEVIATIONS IN .

SECOND LANGUAGE ZULU SPEAKERS

by

Ntombenhle Rosemary Ncobela

Mini- Dissertation

Submitted in partial fulfillment of the

requirements for the degree of

MASTER OF ARTS.ln

AFRICAN LANGUAGES

at the

RAND-AFRIKAANS UNIVERSITYJOHANNESBURG

Supervisor: Dr. A.C. du Toit

OCTOBER 2002

Dedication

Kumyeni wami engimthandayo nasezinganeni zethuuNkosinathi no Bonginkosi Ncobela.

I'

11

ACKNOWLEDGEMENTSNgibonga kakhulu kuMdali weZulu nomhlaba, uJehova, ongigcine kwaze

kwaba lana, nokuthi ngikwazi ukwenza 10 msebenzi.

Ngibonge abazali bami ngokungiyisa esikoleni nangokungikhuthaza

emsebenzini wesikole.

Ngithanda ukubonga kakhulu kusupervisor u - Dr A.C. du Toit

ngesikhathi sakhe nangokungibonisa ukwenza 10 msebenzi ube

impumelelo.

Ngibuye ngibonge kakhulu umyeni wami uLucky ngokungikhuthaza

nangokungilekelela ekwenzeni lomsebenzi.

Ngingakhohlwa ukubonga u - Dr I.Z. Machi ngomsebenzi awenzile ukufaka

isandla ekugayeni 10 msebenzi.

Ngibonge nomsizi wami ekhaya, uRebecca Mdletshe ngokubheka abantwana

ngishikashikeka izifundo.

Ngibonge nabo abantwana ngokubekezela umama engenaso isikhathi

esiningi nabo.

OPSO~ING r .

III

Korrekte uitspraak van 'n taal is belangrik omdat uitspraakafwykings

die betekenis van 'n woord kan verander of die woord betekenisloos

kan maak.

Hierdie studie stel ondersoek in na die uitspraak van Zulu as tweede

taal deur Engelssprekende leerders. Sodanige afwykings is reeds met

toepassing op ander tale bestudeer en het aanleiding gegee tot 'The

Contrastive Analysis Hypothesis' ; 'The Markedness Differential

Hypothesis'; 'The Error Analysis Approach'; 'The Autonomous System

Analysis Approach' en 'Interlanguage.' Daar word vlugtig na hierdie

navorsing gekyk om raakpunte te identifiseer' en 'n toeretiese

raamwerk vir die studie te skep.

Dit blyk dat veral die ortografiese verkille problematies is. Engels

gebruik ook, in somrnige gevalle, 'n paar verskillende letters en

letterkombinasies om identiese klanke in die skryftaal te

verteenwoordig terwyl Zulu foneties is.

'n Vergelyking tussen die uitspraak van vokale en konsonante van Zulu

en Engels toon dat dit veral klankeOmet soorgelyke uitspraak is wat

verwarrend is. Daarbenewens skep die uitspraak van klanke wat vreemd

is ann Engels ook probleme. Mootlike redes v i.r' die vervanging van

Zulu klanke met die van Engels word telkens uitgewys.

Dit sal tweede taal Zulu onderwysers die moeite loon om bedag te wees

op hierdie afwykings sodat spes i a Le aandag hieraan gegee word om

kwaliteit van uitspraak te verbeter.

iv

! '

ABSTRACT

Correct pronunciation of a language is essential since deviations in

pronunciation may change the meaning of a word or render it

meaningless.

This study investigates the pronunciation of Zulu, as second

language, by English speaking learners. Similar deviations have been

studied with reference to other languages and resulted in theories

such as 'Contrastive Analysis Hypothesis', 'Markedness Differential

Hypothesis', 'The Error Analysis', 'Autonomous System Analysis' and

'Interlanguage'. A brief discussion of this research creates the

theoretical framework for this study.

Differences in the orthography of Zulu and English seem to be

problematic. In certain instances, English utilises a variety of

letters to represent identical sounds in written form, whilst Zulu is

phonetic.

When comparing the pronunciation of vowels and consonants of Zulu and

English, it is clear that learners find sounds with a similar

pronunciation especially confusing. In addition to this, sounds

foreign to English often cause problems. Pqssible reasons for

replacement of Zulu sounds by that of English sounds are indicated by

this dissertation.

The concLus i.on of this dissertation is that second language Zulu

educators should be aware of these possible pitfalls and affords it

special attention in order to improve the quality of pronunc i a t i.on,amongst second language learners.

,i

TABLE OF CONTENTS

CHAPTER 1

INTRODUCTION

1.1 AIM OF STUDY

1.2 SECOND LANGUAGE TEACHING

1.2.1 THEORETICAL FRAMEWORK

1.2.2 THE SOUNDS AFFECTED

1.3 METHODOLOGY

CHAPTER 2

THEORETICAL APPROACHES

1

1

1

2

3

.3

4

5

5

2.1 INTRODUCTION 5

2.2 THE CONTRASTIVE ANALYSIS HYPOTHESIS 5

2.2.1 CRITICISM OF THE CONTRASTIVE ANALYSIS

HYPOTHESIS 10

2.3 MARKEDNESS DIFFERENTIAL HYPOTHESIS 11

2.4 THE ERROR ANALYSIS APPROACH 12

2.4.1 CRITICISM OF THE ERROR ANALYSIS APPROACH 15

2.5 THE AUTONOMOUS SYSTEM ANALYSIS

APPROACH 16

2.6 INTERLANGUAGE 16

2.7 SUMMARY AND CONCLUSION

CHAPTER 3

ZULU AND ENGLISH VOWELS

3.1 INTRODUCTION

3 .. 2 ZULU VOWELS

3.2.1 THE BASIC VOWELS

3.2.2 THE RAISED VARIANTS [e] AND [0]

3.3 ENGLISH VOWELS .

3.3.1 FRONT VOWELS

'3.3.2 CENTRAL VOWEL

3.3.3 BACK VOWELS

3 . 3.4 CENTRAL VOWELS

CHAPTER 4

ZULU AND ENGLISH CONSONANTS

4.1 INTRODUCTION

4.2 ZULU CONSONANTS

4 . 2 . 1 STOPS .

4.2.2 CONTINUANTS

4.2.3 CLICKS

4.3 ENGLISH CONSONANTS

4 .3. 1 STOPS

17

20

20

20

20

21

23,

25

26

28

29

34

36

36

36

36

40

43

44

45

45

4.3.2 CONTINUANTS 47

CHAPTERS 48

PROBLEMATIC SOUNDS 48

5.1 INTRODUCTION 48

5.2 PROBLEMATIC VOWELS 49

5.2.1 THE VOWEL [u] 50

5.2.2 THE VOWEL [i] 50.

5.2.3. THE VOWEL [a] 51

5.3 THE REASONS FOR REPLACEMENTS 52

-5.4 PROBLEMATIC CONSONANTS 53

5.4.1 SPELLING MISTAKES 53

5.5 MISPRONUNCIATION OF SOUNDS FOREIGN TO

ENGLISH 61

5.6 CONCLUSION AND SUGGESTIONS 62

APPENDIX A 64

INCORRECT 64

APPENDIX B· 65

CORRECTED: GROUP 1 65

APPENDIX·C 66

CORRECTED: GROUP 2 66

BIBLIOGRAPHY . 67

LIST OF FIGURES

3.1 THE VOWEL CHART OF ZULU

3.2 THE VOWEL [i:]

3.3 THE VOWEL [e]

3.4 THE VOWEL [re]

3.5 THE VOWEL [1\]

3.6 THE VOWEL [a:]

3.7 THE VOWEL [0]

3.8 THE VOWEL [0:]

3.9THE VOWEL [u]

3.10 THE VOWEL [u:]

3.11 THE VOWEL [3:]

3.12 THE VOWEL [a]

21

26

27

28

29

30

30

31

33

34

34

35

LIST OF TABLES

5.1 REPLACEMENT OF VOWELS 49

5.2 REPLACEMENT OF VOWEL [u] 50

5~3 REPLACEMENT OF VOWEL [i] 50

5.4 REPLACEMENT OF VOWEL [i] 51

5.5 REPLACEMENT OF VOWEL [a] 51

5.6 REPLACEMENT OF PROBLEMATIC CONSONANTS 54

5.7 REPLACEMENT OF ASPIRATED STOPS 54

5.8 REPLACEMENT OF DELAYED BILABIAL PLOSIVE 56

5.9 REPLACEMENT OF DELAYED VELAR PLOSIVE 56

5.10 REPLACEMENT OF DELAYED VOICED CONSONANTS 58

5.11 REPLACEMENT OF AFFRICATE 59

5.12 REPLACEMENT OF ALVEO-LATERAL FRICATIVE 62

CHAPTER 1

INTRODUCTION

The cultural diversity of South Africa is a valuable asset and is

reflected in the National Language Plan encompassing eleven official

languages.

One of the aims of the Language in Education Policy is building a

non-racial nation by means of promoting communication across the

barriers of colour, language and region, thus creating multilingual

South Africans.

To this purpose, the Policy states that the home language should be

used as medium of instruction, while at least one additional

language should be taken as a subj ect from grade one. Due to this

policy, schools in South Africa, particularly the so-called

multiracial schools, offer at least one African language, which is

taught as a second language. These second languages are taught

mainly for communication purposes.

Zulu is one of the languages that is taught in the schools where

English is the medium of instruction. The pronunciation of certain

sOl.lnds proves to bep:r:--oblematic to the second language learners.

1.1 Aim of study /In this' research we will be looking at Zulu taught as a second

language in schools, with English as the medium of instruction. The

Zulu sounds that prove to be problematic for second language

learners will he identified and possible reasons for the

mispronunciation will be investigated.

1.2 Second language teachingMawasha (1973:5) says:

Teaching the first language is different from teaching

the second language. For the first language, the

teacher can build on whatever the child has already

acquired while for the second language, the child has

to contend with the first language he has already

required.

This research is intended to help second language learners to avoid

the' possible pitfalls when learning Zulu, and may be of help to

second language educators of Zulu as well.

The main cause of the mispronunciation of the target language might

be the transference of the sounds of the mother tongue to the target

language. Transfer is defined by Jordens (1997:57) as: " a

psycholinguistic procedure by means of which second language

learhers activate their native .1anguage in developing the

interlanguage". This happens psychologically when the person is

learning the new language. A learner thinks of what he or she knows-

from his or her home language and substitutes it for the sound of

the new language. Ri.ntjbom (1987:58) says that it .Ls natural that the

procedure in learning. something new is to establish a relation

between the new proposition and what already exists.

Tarone (1974: 223) found, that in second language acquisition, the

ability to understand' syntactic constructions seems to develop

faster than the ability to actually 'produce these syntactic

2

constructions. The two active linguistic skills, namely speaking and

wri ting are interrelated; therefore the mispronunciation of sounds

affects the written language. (The writing system of Zulu is

phonetic;~ that is, its symbols (letters) represent individual

sounds,') Thus the learner tends to write a word the way it is

produced.

This view is also expressed by Piper (1993:55): "If languages were

not written we could all go about learning them without worrying

about bothersome details like spelling, but they are written...-." This

implies that when the pronunciation problem is resolved, the writing

is corrected as well.

1.-2.1- "Theoretical Framework

Chapter 2 focuses on various attempts to discover and explain the

characteristics of second language phonology. These attempts were

made from a number of perspectives.

Various approaches crystalised in the study of the transfer of the

nati:ve language to the target lang~age. Of these the Contrastive

Analysis Approach has been the most popular. Lado (1957), quoted in

Altenberg and Vago, and Flege (1987) used the Contrastive Analysis

Approach and assumed that second language speakers' deviation from

the target language r.orm is due to the interference from the native

language. He attempted to predict where such deviation would occur.

1.2.2" The sounds affectedChapter 3 and 4 include a description of the relevant Zulu and

English vowels and consonants respectively~

3

In'Chapter 5/ .t.he problematic vowels and consonants are identified

and the possible cause of these substitutions is investigated. In

conclusion, a few suggestions on avoiding the pitfalls of second

language pronunciation of Zulu are ~ade.

1 .,3 MethodologyThe subj ects of this research are

14/ thus from grade 3 to grade 7.

scholars ranging from ages 9 to

The learners wrote a number of

tests while passages were' read and recorded on tape. This was done

unobtrusively so as to avoid false results caused by nerves. These

tests are included in Appendixes A/ Band C.

In 'addition to this / some everyday occurrences of mispronunciation

by second language speakers were noted as well. Input of other

educators was included as well as general pronunciation difficulties

off for example, SABC reporters and newsreaders.

This study does not claim to exhaust all possible pronunciation

mistakes.

4

CHAPTER 2

THEORETICAL APPROACHES

2.1 IntroductionThe. problems concerning second language pronunciation were studied,

interalia, by the Contrastive Analysis of Lado (1957), quoted in

Altenberg and Vago (1983), and Flege (1987), the Markedness

Differential Hypothesis of Eckman (1977), the Error Analysis of

Corder (1961) and the Autonomous System Analysis of Eckman(1981).

2.2 The Contrastive An~lysis HypothesisThe contrastive analysis approach assumes that all speakers'

deviation from the target language norm is due to the interference

from the native language and, it attempts to predict where such

deviat~on will occur.

Mawasha( 1973: 123) says that contrastive analysis provides separate

detailed phonoLoq i ca l, and syntacti,c information on different

languages according to the principles of structural linguistics. It

also highlig~ts the differences and similarities by comparing

resul ts . Contrastive analysis assumes that where there are maj or

structural differences between the native and the target language,

chances are that there will be high interference in the learning

process.

Archibald [1998:50) states that there are four steps to conduct

Contrastive Analysis. These steps include description, selection,

5

comparison and prediction. Thus, a phonetical and phonological

description of the two languages involved is made. Secondly, a

particular part of the ·language is selected for analysis such as

segments, consonants or vowels. Thirdly, the two systems are then

compared. Lastly, the areas of 'difficulties are predicted. By

comparing the two languages the difference will be noted. The

Contrastive Analysis' phonetic and phonological discussion is based

on two important issues, namely similar sounds and new sounds

(sounds foreign to the first language).

New and similar sounds have a great impact on learning a second

language. Flege (1987:9) used this approach in studying Arabian and

English-speaking people. The assumption from many studies is that

the learners may substitute a native language sound for a second

language sound if they hear second language sounds that are similar

to native language sounds. Flege's prediction, however, is that new

sounds would be produced more authentically than the similar sounds.

The finding was that even the experienced second language learners

might ~mly partially approximate the phonetic norm of the second

language in producing the similar sounds.

FI~ge (1983) believes that an accurate pronunciation of foreign

language sounds often requires a language learner to make subtle

articulatory adjustments even when cognate native language and the

target language share a similar phonological sound system. In his

research, it was found that the values for several correlates of

stop voicing produced by Saudis in English, resemble values for the

same phonetic dimensions found in Arabic.

Archibald (1998: 37) says that there are four possible difficulties

in learning! a second language. Firstly, learning a new sound.

Secondly, some sounds are naturally moreaifficul t to produce than

6, . f

others. Thirdly, sounds that are found in many of the world's

languages are easier to learn than the rare sounds. He referred to

the sounds that are common and easier as unmarked and the rare

sounds as marked. Lastly,

relationship between the

systems.

the degree of difficulty depends on the

native 'and the target language sound

Ringbom (1987:58) states that the second language learner tends to

facilitate his or her task by making use of previous knowledge. This

knowledge .i s the linguistic knowledge, which consists of what he

already knows about the target language and of his knowledge of the

native language and possible other languages. He says that the

natural procedure in learning something new is to establish a

relation betw~en the new proposition and what already exists in the

mind.

Ringbom (1981:86) furthermore noted the tendency of learners

transferring function words from one language to another. This

happen~ even if the correct and incorrect words are dissimilar in

sound. He based his viewpoint on research done on Finland-Swedish

candidates learning English, which has shown that their English was

influenced by their mother tongue.

Avery & Ehrlich (1992:XV) state that the nature. of a foreign accent

is determined by a learner's native language. This shows that the

sound patterns of the-native language are being transferred onto the

second language. Every language has a different inventory of sounds,

different phonological rules and different stress and intonation

patterns.

Avery and Ehrlich, in their research, found that the native language

was seen to influence their students' pr6hunciation in three ways.

7

Firstly, difficulty arose when a learner encountered sounds in the

target language that were not part of the sound inventory of his or

her native language. Secondly, the rules for combining sounds into

words are different in the learner's native language. Thirdly, the

problem was in the patterns of stress and intonation. They found

that the native language not only affected the ability to produce

the target language sounds but also the ability to hear the sounds

of the target language.

Transfer is assumed to be the cause of the difficulty in learning

the sounds of the target language. Learners tend to transfer what

they already know to the target language. Smith (1994: 13) defines

transfer as the influence of the mother tongue on the learner's

performance in the development of a given target language. Smith is,

however, against the idea that it is only the influence of the

mother tongue, but is of the opinion that all other languages known

to the learner influence the language.

Littlewood (1984:17) came up with two types of transfer, namely,

positive .and negative transfer. Positive transfer refers to cases

where the first language is helpful in acquiring the second

language, while negative transfer pertains to cases where the first

language is misleading when acquiring the second language. The

difference between the two languages leads to interference, which is

termed negative. transfer. This interference causes learning

dif:ficultiesbecause the learner uses the previous experience as a

means~f organizing the 'second language data. Ringbom (1982:93) sees

transfer as the learner's assumption that the two languages work the

same or.that there·is some translation equivalence between them.

Hatch (1983:20) observed Norwegian learners acquiring English

utterances. He noted that they did not hesitate to use Norwegian

8

words if they were unable to produce the English. They simply used

their language.

Richard (1985:21) agrees by saying that the foundation for second

language is built largely on a transfer of the rules of first

language and that second language is constructed from the learner's

conceptual knowledge. The dependence on first language seems to

occur in the beginning stages. The learners will use not only the

first language structures with second language words, but· first

language words as well. Therefore, first language is responsible for

some errors, especially in cases where specific linguistic items are

similar.

Fisiak{1980: 2,0) says that the basic assumption of the Contrastive

Analysis was that the learner who comes into contact with a foreign

language would find some features of it quite easy and others

extremely difficult.

Fisiak (1981:2) concentrated on contrastive linguistics. He defined

it as a sub-discipline of linguistics concerned with the comparison

of two or more languages in order to determine both differences and

similarities between them. According-to Fisiak, there are two types

of contrastive studies, namely, theoretical and applied studies.

Theoretical contrastive studies give an exhaustive account of the

differences and similarities between two or more languages. In

phonology, this study' operates with phonological' primes or features

and specifies how these features function in the two or more

languages being compared.

Applied linguistics provides a framework

languages selecting whatever information

specific purpose, e.g. translation.

9

for' the comparison of

is necessary for the

2.2.1 Criticism of the Contrastive AnalysisHypothesis

Towell and Hawkins (1994:18) criticize the Contrastive Analysis

Hypothesis saying that it appeared to make good predictions about

transfer but also made inappropriate predictions, because not all

areas of contrast between the first language and second. language

lead to automatic learning difficulties. The second problem with

Contrastive Analysis is that not all areas of similarity between a

first language and the second language lead to immediate positive

transfer. Thirdly, when researchers began systematically to classify

errors in the second language learner's oral and written production,

they found that only a small proportion of errors could be

unambiguously attributed to constructive properties between the

first and second languages.

Another criticism is from Whitman and Jackson (1972), quoted by

Freeman and Long (1991:56), saying that Contrastive Analysis was

inadequate to predict the interference problems of a language

Lea rne r., They continued by saying that interference plays a small

role in language learning performance. This means that no

Contrastive Analysis could correlate highly with performance data.

Norish (1983: 28) r ecommends Contrastive Analysis to educators who

can use it as a device in their study of learner's errors, but it

has its own problems. The important problem is that the educator

should know both the mother tongue .and the target language. Thus,

Eckman . (1977:440) replaced Contrastive Analysis Approach with

Markedness Differential Hypothesis.

10

2.3 Markedness Differential Hypothesis

Eckman (1977: 316) proposes an alternative to the Contrastive

Analysis Hypothesis. He feels that comparison is not sufficient and

says that a notion of relative "degree of difficulty should be

incorporated into the Contrastive Analysis Hypothesis. This notion

of difficulty should be independent of any given language and should

thus be universal.

Markedness is defined by Eckman (1977:320) as: a phenomenon A, that

in some languages is more marked than phenomenon B if the p~esence

of A in a language implies the presence of B, but the presence of B

does not imply the presence ofA. He gives an example of the

phenomenon of languages with only voiceless obstruents as opposed to

languages with both voiced and voiceless obstruents. There are no

languages, however, which contain the phenomenon of just voiced

obstruents. Therefore, the presence of voiced obstruent phonemes in

a language· implies the presence of voiceless obstruents phonemes,

but not the reverse.

He argues that the area of difficulty that the language learner will

have~ can be predicted on the basis of the systematical grammars of

the native language, the target language and the markedness relation

stated in the universal grammar. This can occur in three ways.

Firstly, the areas of the target language, which differ from the

native language, are more marked than the native language, and are

more difficult. Secondly, the relative degree of difficulty of the

areas of the target language, which are more marked than the native

the,

language, will correspond to relative degree of markedness.

Thirdly, those areas of the target language that are different from

the native language, and are not more marked than the native

language, will not be difficult.

11

He says that the Markedness Differential Hypothesis is superior to

Contrastive Analysis because it predicts the areas of difficulty a

language learner will have. He continues by saying that most areas

could be attributed to intralihgual ~ather than interlingual

factors, meaning that many errors were similar to those made by

children acquiring the target. language as first language. He states

that the areas of target language that are difficult are those' that

are different from the native language and relatively not marked in

comparison with the native language. This means that second language

learners will have difficulty with precisely those areas that are

more marked in the target language than in the native language.

2.,4 The Error Analysis Approach

Corder (1993:28), the father of the Error Analysis Approach, agrees

with Ringbom ,in saying that previous knowledge and skills are

intimately involved in the acquisition of new knowledge and skills.

He says that the mother tongue plays a part in the process of

learning and in the use of target language in communication.

Al t.houqh Corder found that there is a phonetic and phonological

transfer, he also found that there are limitations to what can ve

transferred from the native to the target language.

Corder (1967: 427) introduced the Error Analysis Approach as an

alternative to the Contrastive Analysis Approach. This approach

examines the speakers' second language errors without any

preconceived notions as to their cause. Error analysis was done on

the second language English of native speakers of Hungarian. In his

study, Corder found four types of errors. Firstly, he found that

Hungarian sounds were substituted for English sounds that do not

12

occur in Hungarian. Secondly,~und errors due to the application

of unmarked rules. These cannot be linked to interference from the

native language. They are part of neither the native language nor

the target language. There are three unmarked processes found in the

results, namely, word final devoicing, affricating and stressed

vowel lengthening. Subjects were occasionally devoicing a word final

obstruent, which was unnecessary. Furthermore, they affricated /s/

in word final position and also lengthened vowels under stress.

These mispronunciations were not due to their native language.

Thirdly, Corder found spelling pronunciation errors where the

subjects were looking at the English spelling to help them in

pronunciation. Lastly, he found idiosyncratic pronunciation that

refers to .pronunciation that was inconsistent but extensive. This

was not due to transfer, unmarked rule application or spelling

pronunciation. It can be explained as confusion with other

morphophonemically or phonetically related words.

In his results, Corder found that none of the higher-level Hungarian

rules were transferred. It appeared that phonological transfer is

constrained in such a way that only the low-level rules may be

transferred to a target language. For example, in his study with

Hungarians as subjects, he found that it was only the Hungarian

voicing assimilation rule that was transferred ~o English. None of

the higher-level rules was transferred.

Norish (1983: 80) states that· there are two main approaches to the

mechanics .of an Error Analysis. The first approach is to set up one

category. of the learners' most common problems. The second approach

is to group errors into particular areas of grammatical and semantic

problems.

13

Freeman and Lon,g .(1991:56) quoted Corder saying that by classif~ing

the, errors made by learners, researchers could learn a great deal

about the second language learning process by inferring the

strategies that the second language learners were adopting. Some

errors were classified as overgeneralization caused by learning

failure in observing the boundaries of a rule. Other errors were

attributed to simplification, for example, when the plural marker

was omitted from a noun preceded by a cardinal number larger than

one. Still other errors were labeled as communication-based errors,

which resulted when speakers invoked communicative strategies.

Norish (1983:35) states that some errors are due to linguistic

creativity. Learners tend to create new utterances but, with the

limi ted experience of the target language, they may make mistakes.

He 'says that there are two types of creativity in language use. The

first one is the abili ty of the learner to use the parts of the

language that he or she has learnt in order to say something that he

or she may not have heard before. The learner draws a conclusion

about how the language behaves. The second type of creativity is

known as the creative arts. This is rare. In this type of creativity

people are able to create works of literature in a language other

than' their own. For example, a Hungarian writing a novel in English.

Gass and Selinker (1994:59) say that Error Analysis provides a range

of possible explanations other than that of Contrastive Analysis for

researchers. There ar'e'two main error types within an Error Analysis

framework, namely, interlingual and intralingual. Interlingual

errors are the errors that can be attributed to ,the native language.

They in~olve cross-linguistic comparisons. They define intralingual

errors as those that are due to the language being learned,

independent of the native language.

14

2.4'.1 Criticism of the Error AnalysisApproach

Gass and Selinker (1994:60) criticize the Error Analysis Approach by

saying that the major problem is the total reliance on errors in the

absence of other information_ According to their view, one needs to

consider errors and non-errors to get the entire picture of a

learners' linguistic behaviour. The second criticism is the

determination of what an error is an error of.

According to Norish(1983:81), there is a danger with this approach.

because it gives too much attention to trying to blot out certain

characteristic errors and not enough attention to the language. To

eradicate a set of errors does not help the learners to improve

their capacity to use the language for communication.

Freeman and Long (1991:56) are of the opinion that by doing an Error

Analysis and thus focusingon~y on errors, researchers were denied

the whole picture. They studied what the learners were doing wrong

but not what made them successful. It was also difficult to identify

the unitary source of an error. Another charge against Error

Analysis is that it fails to account for all the areas of the second

language in which the learners have difficulty.

Ringbom (1981: 86) acknowledges the criticisms of Error Analysis and

agrees that analysi~·of learners' errors does not give a very

complete view of the learners' performance. However, he maintains

that if a person is interested in underlying mental processes, an

ana.lysis of error data can provide educators with the best possible

window to the learner's mind. To him, the influence of the learner's

first language is the obvious cause of errors.

15

2.5 The Autonomous System Analysis

Approach

The Autonomous System Analysis Approach analyses the phonology of

the second language speaker as a system unto itself and then

attempts to account for the characteristics of that system.

Altenberg and Va go (1983:472) used the same approach with the belief

that not all aspects of native language phonology transfer to the

target language. Their aim was to investigate factors other than

transfer.

2.6 Interlanguage" .

In . addition to the hypotheses discussed above, there is another,

which is well known in the field of second language acquisition,

known as Interlanguage.

McLaughlin (1987:60) says that the term 'Interlanguage' was coined

by Selinker (1980) and refers to interim grammars constructed by

second" language learners on their way to mastering the target

language. It was seen asa separate linguistic system resulting from

the learners' attempted production of the target language norm.

Eckman (1981:195) explains this hypothesis as stating that when

acquiring a second language' the learner internalizes a system of

rules, which may be distinct from both the target language and the

native language. This suggests a creative construction in second

language acqu.t s i.t i.on rather than a process of simple transfer from

the native language.

16

The ,utterances of the second language learner do not belong, to

either the target language or to the native language utterances. The

system thus means that it is independent of both native language and

target language.

McLaughlin (1987: 60) states that Interlanguage is the product of

five central cognitive processes involved in second language

learning:

a) Language transfer: some items, rules and subsystems of

Interlanguage may result from transfer from the first language.

b) Transfer of. training: some elements of Interlanguagemay result

from specific features of the training process used to teach

the second language.

c) Strategies of second language learning: some elements of the

Interlanguage may result from a specific approach to the

material to be learned.

d) Strategies of second language communication: some elements of

the Interlanguage may result from specific ways people learn to

communicate with native speakers of the target language.

e) Overgeneralization of the target language linguistic material:

some elements of the Interlanguage may be the product of

overgeneralization of the rules and semantic features of the

target language.

2.7 Summary and Conclusion

Berkoff (1981: 6) says that there are a number of hypotheses that

were suggested for the errors made by learners of a foreign

language. The first hypothesis is the transfer hypothesis whereby

errors are made based on the interference "f r om the native language.

17i .

Hecohtinues by saying that the problem of learning a new langt;age

lies not ii.n the new language but . in the native language. The second

hypothesis is that some of the strategies 'adopted by the learner of

a second language are substantially the same as those by which a

first language is acquired. Errors are therefore made due to

overgeneralization. The third hypothesis is an attempted synthesis.

It suggests that transfer is more common wi thelementarylearners

and overgeneralization is more common with intermediate learners.

Error analysis addresses four related issues of second language

pponology. The first goal is to consider the constraints on transfer

in second language phonology. Secondly, the data allows an

investigation of those factors other than transfer, which are

involved in the target language production of the subject. Thirdly,

it is a description of factors contributing to second language

phonology. This description will enable linguists to define foreign

accent. Fromkin and Rodman (1988:255) define accent as referring to

phonological differences or "interference" from a different

language. They continue to say that the term also refers to the

speech of someone who speaks the language non-natively.

Comparing the Autonomous System Analysis Approach and the Error

Analysis Approach, Corder and Eckman found both approaches to have

weaknesses and strengths. The Autonomous System Analysis Approach

can account for the rules that generate correct second language

utterances, which Error Analysis cannot do. On 'the other hand, the

Error analysis approach can relate Interlanguage to the target

language as well as the native language that the Autonomous Approach,cannot qo. They carne to the conclusion that second language speakers

apply independent rules which do not form part of either the native

or the target language and that both approaches should be used

]8

concurrerit.Ly to arrive at a more complete understanding of second

language usage.

It seems that t~ansfer of the native language sounds to the target

lang~age sounds is, at least partially the cause of pronunciation

failures of some sounds.

Since this study focuses on pronunciation difficulties, it was

decided to use the Contrastive Analysis Approach in studying Zulu as

the target language of English native language learners.

Chapters 3 and 4 will identify and focus on the Zulu sounds that are

problematic to second language learners.

19

CHAPTER 3

ZULU AND ENGLISH VOWELS3.1 Introduction

The comparison of Zulu (second language) and English (first

language) will possibly shed light on the reasons for replacing Zulu

vowels by similar sounding English vowels. This chapter focuses on a

description of Zulu and English vowels. The phonetic symbols used

are that of the International Phonetic Alphabet (IPA).

3.2 Zulu'VowelsPosthumus (s.a:37) explains that the shape and size of the oral

cavity can be modified to pronounce the sounds of different vowels.

Vowels are also described by the position of the vowel relative to

~pe cardinal vowels. For example, [£j is described as being between

cardinal position 3 and 4, but closer to 3, which will be seen in

the vowel chart. The cardinal positions are the positions where

horizontal and vertical lines meet on the outline of the figure

below.

A vowel is described in terms of the vertical modification,

horizontal modificatinn, the extent of lip rounding and its position

relative to the cardinal vowels. The horizontal modification of the

oral cavity is expressed relative to th~ front or back of the mouth.

Thus the vowels are described as front, central or back vowels.

The vertical modification is expressed in terms of the relative

height bf the tongu~ in the oral cavity. Thus, vowels are described

20

'11 '

as low, mid-low, mid-high and high. Sometimes vowels have length

shown by the colon (: ). This colon indicates the length of 'the

immediately preceding vowel. For example; in the word [ik'a:t'i] (a

cat), the length is on the preceding vowel fal. These vowels may be

schematically represented on a vowel chart .

[i]l

[e] 2

• 8 [u]

• 7 [0]

6 [0]

5

[a]

Figure 3.1: The vowel chart of Zulu

3.2.1 The Bas~c vowe1sZulu has five basic vowels,Posthumus (s.a:41) gives thevowels of Zulu:

namely:following

[ a] , [E], [ i] , [0] and [u]description of the basic

3.2.1.1 The Front vowels [i] and [E]

a) [i] is a high, front vowel, unrounded and it lies between the

cardinal positions 1 and 2 but closer to 1. It is found in

words such as:

[usi:si]

[isi:su]

[imifi:nO]

(sister)

(stomach)

(vegetables)

21

.,b) [£] is a mid-low front vowel, unrounded and it lies between the

cardinal positions 3 and 4 but closer to 3.

[ esi~£fje:la]

[ is£l£S£:I£]

(at the hospital)

(frog)

3.2.1.2 The Back vowe1s[O]and [u]

a) [~ is a mid low, back vowel, rounded and it lies between the

cardinal positions 5 and 6 but closer to 6. It can be found in words

such as:

[ulsuDo,:na] (to see)

(a granny)

b) [u] is a' high back vowel, rounded and it lies between the

cardinal positions 7 and 8 but closer to 8. It can be found in

words such as:

[ukufu.na]

[ukul.u.ma l

(to want)

(to bite)

3.2.1.3 The Centra1 vowe1 [a]

[a] is a low, central vowel, unrounded and it lies between the

cardinal positions 4 and 5 but closer to 5. It can be found in words

such as:

[abafiana] (boys)

22

[ukula.La l (to sleep)

3.2.2 The raised variants [e] and [0]

3.2.2.1 The mid-high front raised variant [e]

[e] is a mid-high front variant, unrounded and it lies between the

cardinal positions 2 and 3 but closer to 2. It can be found in words

sl.lch as:

[isikhe:thi] }

[ip'ule:t'i] }

(a skirt)

(a plate)

3.2.2.2 A mid-high back raised variant [0]

[0] is a mid-high, back variant, rounded and it lies between the

cardinal positions 6 and 7 but closer to 7. It can be found in the

.following examples •.

[umgoQo:j i]

[umlo:f>i]

(a dog)

(a writer)

Posthumus (s.a:l7) explains that these variants of the mid-low

vowels are phonologically determined. The originally low vowels are

raised in the following circumstances;

a) The mid-low vowels are raised when followed by the high vowels

[i] or [u] in the immediately following syllable. For example,

23~ , ,I

in the word [Uls:uDE:ka] (to put). the mid-low vowel [E] is not

raised because the immediate syllable has a low vowel [a]. In

the word [siDeki:le] (we have put) the mid-low vowel [E] is

raised to [e] by the high vowel [i] in the immediately

following syllable. Similarly, the vowel [0] is raised when

followed by the high vowels [i] or [u] in the immediately

su~ceeding syllable. The mid-low back vowel [0] in [ikotfana] is

raised to [0] in [Lk'op'Ll (tin). In the latter example, the vowel

[0] is thus raised by the high vowel [i] in the succeeding

syllable.

b) The mid-low vowels are also raised when the syllable following

them contains a syllabic [m], for example, in [uQut'j omda.La] (big

brother) the vowel [0] is raised to [0] by a letter [m] which

represent the syllable Imul and is thus syllabic.

c) The mid-low vowels are also raised when used simultaneously

with the .locative suffix, for example, [endf)i:nij (in the house)

or [ottfani:ni] (on the grass].

d) The mid-low vowels are raised when they are at the end of

monosyllabic words and are pronounced with half-length, for

example, [Pd3e] (just) .

e) The vowel raising process takes place retrogressively when

vowel raising,has been triggered by one of the above-mentioned

factors. For example, the word 'iselesele' (frog) is

phonetically transcribed as [isEIEsEIE] but in the locative form

24

it is [eseleseHmi] (in the frog). The [i] in the locative

suffix raised the final vowel le l of the noun to [e]. This

raised vowel has in turn Lnr Luenoed :the immediately preceding

vowel as well as the other preceding vowels in the word

resulting in [eseleseleni] . This retrogressive vowel raising

can only be stopped by the low vowel [a] , for example in

[ent'ombazaneni] (in the girl).

3.3 English Vowels

In addition to describing the English vowels in terms of horizontal

and vertical modification, Cruttenden (2001: 105) utilises the terms

"open" and "cIose" positions.

High vowels are known as closed vowels while low vowels are referred

to as open vowels. The mid high vowels are half closed and half

opened because the tongue is . lowered from the close position to

about one-third and two-thirds of the total distance from the closed

position to the open red position.

Due to many vowel combinations such as digraphs and dipthongs and

variations in the pronunciation of the vowels in English, it will be

impossible to include all the vowels in one chart . Thus, after the

description of each :~owel, it will be indicated on a vowel chart.

Furthermore, only vowels relevant to this study will be included.

25

3.3.'1 Front vowels

a) [i] is a high front vowel that has length, is unrounded and

lies slightly below the front closed position. Cruttenden also

notes a reduced variant [i] as is found in words such as 'feet'

and 'beef'. [i:] can be represented by the following letters in

the orthography:

leel in tre~s, cheese

lei in complete, be

leal in leaf, reason

Iii in machine, police

liel in piece, field

leil, leyl in seize, key

i:

Figur~ 3.2: The vowel i:

The problem anticipated here is the substitution of the letter leifor Iii due to the fact that the letter lei may be used to represent

the sound [i:] in English. Crutten?en (2001: 105-107) furthermore

indicates that the long vowel [i:] has a reduced variant, which

occurs as the letter Iyl in word final position, for example, city,

26

letter lal for [a] while English may also utilize the letter lal

with reference to [<:e].

Figure 3.4: The vowel~~

3.3.2 Central vowelCruttenden (2001: 113) described· [1\] as a low vowel pronounced with

the lips neutrally open, with the centre of the tongue raised just

above the fully open position. He continues to say that the quality

of the sound is that of a centralized and slightly raised [a]­

sound. He mentions the Cockney pronunciation of this sound as being

further forward and more open, 'approaching' [a]. It is evident that

this sound is very similar to that of the Zulu sound [a], and that

the 'substitution of this sound for [a] will not be easily detected.

The sound [A] is represented by the following letters:

lui in cut, drug

101 in son, among

loul in country, couple

1001 in blood, flood

loel in does

28

A

Figure 3.5: The vowel[~

Keeping the closeness in pronunciation between [A] and [a] in mind,

it is anticipated that second language Zulu learners may substitute

tbe letter /u/ (that represents this sound in English) for /a/ which

represents the sound [a] in Zulu~

3.3.3 Back vowels

a) Cruttenden (2001: 115) describes [a:] as a low, long vowel

pronounced with the lips neutrally open between the centre and

the back of the mouth. It has a reduced form of [a] and is

represented

orthography:

by the following letters in the English

/a/ in pass, after

/a(r)/ in part, car

/ae(r)/ in ~~art, clark

/a(l)/ in calm, half

/au/ in aunt, laugh

Although Cruttenden explains that the English sound [a:] as being

"closer to the cardinal sound [a]", he cites the Zulu word 'inkatha'

as an example and is therefore confused. -Th i.s , however shows that

29

the distinction between [a:] and [a] is very slight and will probably

only be picked up by sophisticated equipment.

a:

Figure 3. 6: The vowel [0:]

b) Cruttenden describes [0] as a low, back, rounded vowel, which

lies in the open position. It is represented by the following

letters

101 in dog, dock

loul in cough, through

laul in because, sausage

o

Figure 3.7: The vowel[~

30

Since the sound [0] may be r-ep.re s ented by the letter faf in

. English, replacing the vowel [aJ in Zulu by this sound will not

be reflected in the spelling of t.he learners. It furthermore

differs to such a slight degree from the pronunciation of both

[aJ and [0] in Zulu, that mispronunciation of these sounds will

not be easily detected.

c) Cruttenden (2001:119) describes [~] as a back, mid-low, rounded

vowel that has length. It has a reduced variant [0] that·

correlates with the [0] sound of Zulu. It may be represented by

the following letters in English:

fa) in war, quart

fo(r) in before, more

fou(r) in court, four

foa(r) in oar, board

foo(r) in floor, door

fau! in fault, cause

faf in all, talk

0:,-------~+-----;

Figure 3. 8: The vowel [0:]

31

Lfreplacement of the vowel [a] in Zulu by the sound [~] does

occur due to the letter la/ that may represent this sound in

English, it was not detected in the spelling tests. It seems

that second language learners are not confused by this.

Cruttenden (2001:121) describes [U] as a mid high (close-mid) vowel

pronounced nearer to the centre than to the back, the quality of

which is that of a centralized [0]. If rounding is involved, it is

closed but loose.

Her furthermore indicated that in some English words there is a

variation between [U] and [u:], for example in room, groom and tooth,

the COmmoner phoneme being lu:/.

This sound may be represented by the following letters in

English:

luI in butcher, cellular

1001 in book, wood

101 in bosom, wolf

loul in could, courier

When replacing the Zulu sound [u] represented by the letter luIin Zulu with the letter 101 in spelling tests as indicated in

paragraph 5.2, the learner might have this sound in mind.

32

u

Figure 3.9: The vowel [u]

It might, however, also be possible that the spelling

deviations are an indication of the replacement of the Zulu

sound [0] by the English sound [u] as will be seen in the

following'paragraph.

Cruttenden (2001:122) describes [u:] as a high, back, rounded vowel

that has lengt~. He also mentions a reduced variant [u] and makes

mention if the similarity in pronunciation between [u:], [u] and [U].

The sound lu.L may be represented by the following letters in

English.

luI in rude, June

1001 in food, soon

101 in do, who

loul in group, soup

lewl in chew, . flew

fuel in blue, sue

33

The replacement bf the Zulu letter lui by the letter 101 may be

due to the fact that the letter 101 may represent the sound" Iu.l

I[u] in English words .

..----------.,,.----,....-----'----,. u:

Figure 3.10. The vowel [u:]

3.3.4 Central Vowels

a} Cruttenden (2001: 125) describes [3:] as a central, mid-low,

vowel with length. It exhibits a reduced variant [3] as well. It

is represented by the following letters:

. Ie (r) I in her, serve

lu(r)1 in turn, church

li(r}! in sir,bird

l(w+}o(r}1 in word, work

lea(r)1 in earth, heard

lou(r) in journey, courtesy

Since this sound is, without exception, followed by the Irl,

Irl is only to be 'found in a very limited set of words in Zulu.

34

3:

Figure 3. 11: The vowel [3:]Cruttenden (2001: 127) describes [a] as a central unrounded vowel

that lies between the half-closed and half open positions. It may

be' spelt with most vowel letters and their combinations, for

example:

/e(r)/ in 'mother,

/o(r)/ in doctor

/i/ in'possible

/a(r)/ in particular

/a/ in suppose

I. <!

a

Figure 3. 12: The vowel [a]

This sound is foreign to Zulu

mispronunciation are anticipated.

35

and no problems in

CHAPTER 4

ZULU AND ENGLISH CONSONANTS4.1 IntroductionThis chapter will focus only on consonants that are perceived as

problematic in that they are commonly replaced by similar English

sounds. The International Phonetic Alphabet (IPA) is used in the

phonetic descriptions of the sounds.

4 .. 2 Zulu ConsonantsPosthumus (s.a:59) states that Zulu consonants are classified into

two large groups, namely, stops and continuants. Stops are

characterised by a single momentary total stopping of the airstreams

by a definite obstruction. This obstruction may occur repeatedly or

contimiously. In pronouncing continuants no partial obstruction of

the airstreams occurs.

Consonants are described in terms of six criteria, namely:

a) Manner of articulation:

This criterium refers to modification of the airstream during

articulation. According to the manner in which the airstream is

obstructed or modified, four types of speech sounds may be

distinguished, namely explosives, implosives, fricative,

resonants and affricates.

36

Explosives are sounds where the airstream moves to the outside

of the mouth when the closure is released. Examples of plosives

are: [p'], [t'] and [k'].

Implosives are sounds where the airstream moves into the mouth

once closure in the mouth is released. There is one implosive

in Zulu, namely [0].

Fricatives are marked by a narrowing of the opening between the

articulator and place of articulation so as to produce a

turbulent airstream. Examples are [f], [s], UJ, [I I], [h], [v]

and [z].

Resonants is the term used to refer to all sounds that do not

have a definite occlusion in the speech tract at the end of the

pronunciation. Examples of resonants are [1], [w], [m], [11],

[I'll, [f1J] and [n].

Affricates are sounds that begin as an explosive and end as a

. fricative. Examples of the affricatives are [tf]. [d3l, and [Is'].

b) Direction of the airstream:

In articulating sounds, the airstream may move outwards of or

inward into the ,oral cavity. Speech sounds that are produced by

the airstream moving into the oral cavity are the click sounds

[!], [/J, and [II] (and the different variants) and the

implosive sound [0]. All other speech sounds of Zulu are

egressive.

37

c) Origin of airstream:

The origin of airstream used may be oral and glattal. Example

of an oral is [11 and the example of the glottal is It;'}, If the

airstream is initiated by moving the larynx. Upwards, it is

called glottalic initiation and is referred to as ejection by

many phoneticians.

The origin of the airstream can also be glottalic or pUlmonic.

It may be glottalic if the sound is pronounced bY a narrowing

of the glottis. The examples of pulmomic sounds are [h] and [h].

It can be pulmonic when the air used to produce a speech sound,

originates from the lungs. Most sounds are produced with the

air originating from the lungs .

. d) Channel of air release:

This criterium refers to the manner in which air is channeled

when producing sounds. There are two channels of airstream

release, namely, oral and nasal. The sound may b~ oral if it is

articulated by releasing the airstream over the centre of the

tongue or over the side of the tongue. All sounds are oral

except the nasals. Nasals are sounds articulated by an

airstream moving through the nose. An example of an oral sound

is [1] and an example of a nasal sound is [n].

e) State of glottis:

Posthumus (s.a:54) explains that this criterium refers to the

state assumed by the openingb.etween two vocal cords in the

articulation 'of a sound. Using the state of glottis as a

criterium results in different types of speech sounds, namely,

38

voiced, voiceless, breathy voice, devocalised and aspirated

consonants. , '

A vibration on the vocal cords marks the voiced speech sound

when the sound'is articulated.. Examples are [g] and [d]. The

voiceless speech sounds are those sounds that are articulated

when' the glottis is open and the vocal cords do not vibrate.

Examples of voiceless sounds are [s] and [k]. Breathy voice

sounds are the sounds produced with the breathy voice. There is

only one sound, that is, [h]. Devocalised sounds are the sounds

with delayed. voicing. Examples of devocalised sounds are [g] and

[Q]. Aspirated sounds are produced with the extra pulmonic air.

Examples of aspirated sounds are [ph], [th] and [kh].

f) Place of Articulation

Speech sounds are described and classified in terms of the

tract where articulation takes place.

Velars are sounds that are articulated by raising the back of

the tongue to touch the soft palate (velum). Alveolars are

sounds that are articulated by raising the tip of the tongue to

touch the tooth ridge. Bilabials are sounds that are

articulated by employing both lips. Palatals are sounds that

are pronounced at the oral roof region situated behind the

alveolar ridge.~Whenthe middle tongue. is raised to touch the

hard palate, palatal sounds are formed. Alveo-laterals are

sounds that are produced when the tip of the tongue is placed

against the alveolar ridge and the back of the tongue is

against the velum. Labio-dentals are sounds that are pronounced

by raising the lower lip to touch the upper front of the teeth.

39

4.2 .. 1 Stops

4.2.1.1 Velar stopi) [k] is an eggressive, ejective, oral central, voiceless, velar

stop.

Examples:

[ik'a:t'i]

[isik'a: Li.]

(a cat)

(a scale)

ii) [kh] is an egressive, aspirated, oral voiceless velar stop ..

Examples:

[ukha:mba]

[ikhe:li]

(a clay pot)

(an address)

iii) [g] is an egressive, oral voiced velar stop. The voiced

consonants are found only in combination with nasals.

Examples:

[ir]ga:nE] (a child)

[ir]goQu:sO] (fiance)

@]is an egressive, oral delayed voiced velar stop.

Examples:

[u~ucja:r]ga] (be naughty)

[u~ucja:Qa] (to look after)

40

4.2.1.3 Bilabial stops

i) [i?] is an egressive, pulmonic delayed voiced bilabial.

Examples:

[uJsui?a:la]

[uJsui?E:k'a]

(to write)

(to look)

ii) [mb] is an egressive voiced nasalised bilabial. This voiced

sound is combined with the nasal.

Examples:

[imba:li] , (a flower)

[imbu:zi] (a goat)

iii) [0] is an ingressive implosive

Examples:

[ukubona] (to see)

[uJsubopha] (to arrest)

iv) [p'] is an egressive ejective bilabial sound.'

Examples:

[p'op'o:la]

[is-ip'o:k'i]

(to examine)

(a ghost)

v) [ph] is an egressive aspirated bilabial stop.

42

4 .2 .. 1.3 Bilabial stops

i) [9] is an egressive, pulmonic delayed voiced bilabial.

Examples:

[ul,}u9a: l a]

[ulsuQE:k'a]

(to write)

(to look)

ii) [mb] is an egressive voiced nasalised bilabial. This voiced

sound is combined with the nasal.

Examples:

[imba:li] . (a flower)

[imbu:zi] (a goat)

iii) [b]is an ingressive implosive

Examples:

[ukubona l (to see)

[ul,}ubopha] (to arrest)

iv) [p'] is an egressive ejective bilabial sound."

Examples:

[p'op'o.La l

l Lsd.p'o.k'd l

(to examine)

(a ghost)

v) [ph] is an egressive aspirated bilabial stop.

42

Examples:

[ukuphd inda]

[ukuphe.pha]

(to repeat)

(to be safe)

, '

4.2.1.4~Palatal stop

i) [~Q3] is an egressive, pulmonic voiced nasalized palatal.

Examples:

[i:~Q3a]

[ma:~Q3e]

(a dog)

(now)

ii) [d3]is an egressive devocalised palatal stop.

Examples: '

[d3af>:ula] (to be happy)

[id3ok'a] (a yoke)

4.2.2 Continuants

4.2.2.1 Alveolar

[s] is an egressive palrnonic voiceless alveolar fricative.

Examples:

[is.i:su] (stomach)

[usu.k'u l (a day)

43

4.2.2.2 Labio-dentalt,

[f] is an eggressive, pulmonic voiceless labia-dental fricative.

Examples:

[ukufLka l

[uJsufu:nda]

(to arrive)

(to read)

4.2.2;3 Alveo-lateral

Alveo-lateral [1] represented by letter Ihll is foreign to English.

We find this sound in words such as [uJsu1aba] (to slaughter).

4.2.3 ClicksClick sounds [I], [!] and [II] are also problematic sounds to

second language learners. We are not going to look at them, however,

because they are totally foreign to English speakers and will

therefore enjoy continued attention in the classroom. The click

sounds are [I], [//] and [!] represented by letter c, x and q

respectively in Zulu.

For example: iculo [i/u:lO]

ukuxoxa Iuku/ 10:1 fa]

ukuqala .Juku l a.La I

(a song)

(to chat)

(to begin)

The learners' attention should be, however, drawn to this since in

the English orthography, the letter lei represents the sounds [k] or

[s], for examp.le ; in 'complete' and 'special'; the letter Iqlrepresents the sound [k] in 'queen' or 'quick' and the letter Ixl

represents the sounds [ka'] in words such as 'extra'.

44

r .

4.3 English ConsonantsEnglish consonants are described in the same way as in Zulu. The

difference is that these syrnbo~s can be represented by many

different letters in the orthography.

4.3.1 Stops

4.3.1.1. VeJ.ar stopsi) Cruttenden (2001: 151) describes [k] as a voiceless velar

plosive. If it is followed by a vowel or is in word initial

position, it becomes aspirated. It may be represented by the

following letters:

Ikl in king, kept

Icl in ·carpet, cat

Iccl in accused, accomplish

Iql in cheque, unique

Ichl in stomach, back

Ick/in neck, rock

ii) [g] is a voiced velar plosive, which can be represented by

the following letters:

Ig/in go, gut

Iggl in egg, aggressive

Ighl in ghost, spaghetti

Igul in guilty, guitar

Iguelin league, vague

45

, >

4.3.1.2 Al.veol.ar stopsL) [t]is a voiceless alveolar plosive. If it is in word initial

position or when followed by a vowel in a stressed syllable as

in 'taken', it is aspirated.

ii) £d3] is a voiced peLat.o-aveoLa.r affricate. It appears as

follows in the orthography:

Ijl in jam, jaw

Igl in gem, magic

Idgl in judge, fridge

Idjl in adjacent, adjective

4.3.1.3 Bil.abial. stopsi) [p]is a voiceless bilabial stop. If this sound is followed

by a stressed vowel or is in word initial position, it is

produced with a considerable puff of air or aspiration. For

example, in a verb 'push'.

ii) [b] is a voiced bilabial plosive, which is represented by the

following letters.

Ibl in rib, rub

Ibbl in rubber, rubbish

Cruttenden (2001:151-168) draws attention to the variant

pronunciations of the velar, alveolar and bilabial stops. These may

prove to be problematic for second language learners and will be

discussed in chapter 5.

46

4.3.2 Continuants

4 . 3 •.'2.1 AJ.veolari) According to Cruttenden (2001:185) [s] is a voiceless

alveolar fricative, which can be represented by the following

letters in the orthography:

lsi in saw, sand

Iss I in pass, passive

Icl in receive, reception

Iscl in science, scent

Ixl in axe, six

ii) [z] is the voiced counterpart of [s] and is described as a

voiced alveolar fricative which may be presented by the

following letters in the orthography:

Izl in zoo, zeal

Izzl in dizzy, buzz

Ixlin exact, exit

Is(e)1 in bars, rose

Issl in possess, dessert

iii) [f] is a voiceless labio-dental fricative, which is

represented by the following letters:

If I in fork, friend

Iffl in off, stuff

Iphl in physics, phonetics

Ighl in enough, rough

47

CHAPTER 5

PROBLEMATIC SOUNDS5.1 IntroductionThe meaning of a word may be changed, or the word may be rendered

meaningless by substituting one sound for another.

For example, 'ukuthula' (to be quiet) : 'ukuthola' (to find)

'ukuvala' (to close) 'ukuvula' (to open)

'ukugoba' (to bend) 'ukukoba' (to take a glance)

A number of tests were given to the learners to evaluate their

pronunciation and spelling of Zulu sounds. Since writing and

speaking are interrelated, pronunciation was also affecting their

written work.

Spelling tests' were given in the form of a worksheet with pictures

of people involved in different actions. The learners had to fill in

a suitable Zulu verb next to each picture. Spelling was also tested

in the form of a passage that contained some incorrect vowels (see

appendix A) . This was done to see i~ they could identify incorrect

vowels by reading the words.

Passages were given for reading that would test their pronunciation,

while recording their reading on tape. These tests were aimed at

identifying problematic sounds.

In addition to the. sounds that surfaced in these tests, listening to

everyday conversations identified more problematic sounds and the

language usage in SABC and E TV broadcasts by second language

48

speakers. These programs included sports and news programs as well

as 'soap operas' such as Muvhango, Isidingo and Backstage. I·

This chapter will focus on the reasons why Zulu sounds described in

the previous chapter were replaced by English sounds. This

discussion will be based on the description of both Zulu and English

sounds in chapter 3 and 4. The purpose is to alert educators to

these potentially problematic sounds.

5.2 Problematic VowelsLetters that represent the English equivalent or similar sounding

vowels in English often replace the following letters representing

Zulu vowels. This results in spelling mistakes and incorrect

prQnunciation in some cases. The incorrectly spelt words are

indicated by an asterisk in the table below, while the problematic

vowels are printed in italic.

Table 5.1: Replacement of vowels

The· Zulu sounds ~eplacements<: ',•• '

~,.

u *0

i *e

i *y

a *u

The replacement of., for example, [0] by [1\], [a] by [83], [a] by [u:], [a] by [0]

and [u] by [u] will hot be reflected in spelling since the sounds are

represented by the same letter in both languages. It will only come

to the fore when comparing actual pronunciation and will often be a

49

question of degree. This study does not focus on this kind of

mispronunciation.

5.2.1 The vowel [u]In written form, the letter lui which represents the high back vowel

[u] in Zulu words is often replaced by a letter 10/. This is caused

by the similarity in pronunciation between the sound [u] and the

mid-high rounded central vowel [U] in English. Both these sounds can

be represented by the letter 101 in English. For example, [u: J in

'do'and [UJ.in 'bosom'.

Figure 5.2: Replacement of vowel [u]

Zulu words Replacement

ukuthula (be quiet) *ukuthola

umlomo *omlomo

ukuvuka *ukuvoka..

5.2.2·The vowel [i]The letter IiI which represents the high front vowel [iJ in Zulu

words is replaced by the letter lei and sometimes the letter Iyl as

may be seen in the table below. This appears in words that will be

shown in the table below.

Table~.3: Replacement of vowel [i]

Zulu words Replacement

sikhona (we are fine) *sekhona

phansi *phanse

50

The substitution of the letter Iyl for the letter Iii usually

happens when the vowel is preceded by a nasal[n] and when it lis a

terminal vowel.

Table 5.4: Replacement of vowel [i]

Zulu'; words Replacement

unjani? (how are you?) *unjany

sanibonani (hello) *sanibonany

'l'he high front vowel [i] is represented by' the letter Iii in Zulu

orthography. This English sound is, however, often represented by

the letter lei in words such as 'complete' and the letter Iyl in for

example 'pretty', (thus in word final position). This causes the

letter Ii/ to be replaced by the letter lei by second language

learners, which renders the written words meaningless.

5.2.3. The vowel ra]

Table 5.5: Replacement of vowel [a]

'Zn,l1 [~~'~~~i~~!j\~!.;;i ~epl~c:eJ1l¢t1t.,:;Of.", ., . .,iii,,'/; .;

thula (be quiet) *thulu

ngenani (come in) *ngenuni

sanibona (hello) ..*s unibona

nami (and I) *numi

unjani? (how are you *unjuni?

izandla (hands) *izundla

isifuba (chest) *isifubu

amehlo (eyes) *umehlo

51

The low central vowel [al represented by the letter lal in the Zulu

orthography, is commonly replaced by the English sound [~ In English

this sound is represented by the letter lui in words such as 'but'

and 'cut'. The replacement is due tb the closeness in pronunciation

between the two sounds. The incorrect pronunciation is also

reflected in the spelling.

5.3 The reasons for replacementsThe replacement of the vowels is evident in the tests that were

given. These tests and responses are included as Appendixes A, Band

C. The text in Appendix A contains spelling mistakes, which the

learners were asked to correct. The responses of the learners are

included in appendixes Band C. The letters in brackets indicate the

corrections made by the learners. As can be seen in Appendix B, some

learners did not recognise the letter 101 in 'suwobona' as

incorrect. Only the letter lui was corrected. The letter 101 in

English is used to represent the high back vowel [u] as for example

in the word 'soon'. The sound is, however, represented by the letter

lui in Zulu. The learners failed to see that.

Secondly, they did not recognize that the word 'ngikhona' is spelt

incorrectly as 'ngekhona'. The letter Iii in the Zulu word is

replaced bya letter leI because in English the high-front vowel [i]

may be represented by-the letter lei in words like 'complete'.

As is evident in the first 'suwubona' in Appendix C, a group of

learners confused the letter lal which represents [al in Zulu words

with the letter lui. In English, the letter lui may represent the

similar sounding, [1\] in words such as 'cut'.

52

I .

5.4 Problematic ConsonantsIn analyzing the spelling and pronunciation difficulties of second

language Zulu learners, two main c~tegories of mistakes come to the

fore, namely: spelling mistakes (but correct pronunciation) and

pronunciation mistakes that are or are not reflected in the written

form. The latter is due to zulu sounds that are foreign to English.

Second language speakers tend to replace these by similar English

sounds. Educators should give special attention to the correct

pronunciation of these sounds.

5.4.1Spellinq mistakesThis category pertains to words or sounds that are pronounced

correctly but spelt incorrectly. The spelling mistakes are due to

differences in the orthographies of Zulu and English - identical

sounds may be ~epresented by different letters.

The denti-labial (or labio-dental) voiceless fricative [f] is

represented by the letter If I in Zulu. This sound may, however, be

represented by the letters Iphl in words like 'physics' in English.

The voiceless alveolar fricative Is] is represented by the letter

Is/ in Zulu. This sound may, however, be represented by the letter

Ic/ in words like 'receive' in English.

The voiced palata-alveolar (or alveo-palatal) fricative [~l is

represented by the letter Ij/in Zulu, preceded by [Pl. This sound

may be represented by the letter /g/ in words like 'engine' in

English.

53

Table 5.6: Replacement of problematic consonants

Sound Zulu letter Replacement

[f] f ph

[s] s c

[<1.3] (n) j (n ) g

For example:

[f]: isifuba (chest): *isiphuba

[s]: sikhona (we are fine): *cikhona

[<1.3] unjani? (how are you?): *ungani?

So~e spelling mistakes are due to the phonological rules pertaining

to the two languages. The phonological rules of English dictate that

the alveolar, bilabial and velar plosive [t], [p] and [k] become

aspirated when these sounds precede a (stressed) vowel. The

aspiration is, however, not reflected in the spelling of the English

words, while aspiration is reflected in the orthography of Zulu.

This difference results in the following mistakes:

Table 5.7: Replacement of aspirated stops

Sound Zulu letters Replacement

[th] th ~*t

[ph] ph *p

[kh] kh *k/cI

54

E;or example:

[th] thula! (be quiet): *!ula

[ph] iphepha (paper) : *iEeEa

[khl sikhona (we are fine): *sikona

[kh] sicona (we are fine) : *si~pna

The last example is even more problematic for the letter leIrepresents the dental click or suction sound [11 in Zulu. The letter

leI may, however, in English be used to represent velar plosive [k]

in words such as 'carpet'.

5.4.1.1 Voiced, delayed voiced and voicelessconsonantsZulu distinguishes voiceless, delayed voiced and voiced sounds. Some

of these delayed voiced sounds are foreign to English. Due to the

similarity in pronunciation, second language Zulu learners tend to

replace delayed voiced sounds with fully voiced sounds.

Cruttenden (2001:152) states that the voiced stops [b], Cd] and [g]

in Eng~ish may have full voice when occurring between voiced sounds.

In initial word positions, that is, following silence, these sounds

-may be devoiced and thus realize as [Q], [~] and [9] respectively.

a) Replacement evident in spelling

i) The delayed voiced bilabial explosive [~

The delayed voiced bilabial explosive [~ is represented by the

letters Ibhl in Zulu words .. Second language learners often replace

this sound with the voiced bilabial explosive [bJ, that is

represented by t he letter Ibl in English words like 'rib'. The

mispronunciation is reflected in the written form as well.

55

Table 5.8: Replacement o£ delayed bilabial plosive

Sound: Zulu letters Replacement

[9] bh b

For Example:

[9] ibhola (a ball) *ibola

ukubhala (to write) *ukubala (to count)

ukubheka (to look) :* ukubeka (to put)

The replacement of [~] may cause a change in meaning as can be seen

in the last examples.

ii) The delayed voiced velar explosive [9]

The delayed voiced velar explosive [9] is represented by the letter

Ikl in Zulu. This sound is often replaced by the voiced velar

explosive [g] that is represented by the letter Igl in English words

like 'go'. Since different letters' are used in Zulu and English,

this mispronunciation is reflected in the written form.

Table 5.9: Repl~~ement of delayed velar plosive

Sou Zulu letters Replacement

[9] k g

For example:

[gJ : ukuvuka (to wake up)

56

*ukuvuga

The replacement of Ikl by Igl renders the ·word meaningless.

iii) The voiceless alveolar fricative [s]

The voiceless alveolar fricative [s] is sometimes replaced by the

voiced alveolar fricative [z]. Thi~ probably happens since the

letter lsi may, in English represent the sound [z] in words such as

'rose' and 'bars', as indicated in chapter 3

b) Replacement not evident in spelling.

i) Delayed voiced consonants

Some delayed voiced consonants of Zulu are represented by the

same letter as that of the voiced counterparts of the sounds.

The sound is changed from the delayed voiced to voiced due to

the phonological rules of Zulu. These rules stipulate that

these sounds will be voiced whenever preceded by a nasal

sound. The delayed voiced sounds are:

• The alveolar delayed voiced consonant [<;I], represented

by the letter Idl in Zulu.

• The alveo-palatal delayed voiced consonant [dB],

represented by the letter Ijl in Zulu.

• The delayed voiced velar consonants [9], represented by

the letter [g] in Zulu

English speaking learners tend not to discriminate between the

[+delayed voice] and [+voice] pronunciation of these sounds and

pronounce both as [+voice]. This mi~pronunciation is not evident in

written. form.

57

Table 5.10: Replacement of delayed voiced

consonants

Sound Zulu letter-.'.

[Q] or Cd] d

[d3] or [d3] j

[g] or [g] g

Examples:

Delayed voiced Voiced

[Q] : udade (sister) Cd] :indoda (man)

fd3]: ukujabula (to be happy) [dz]:injabulo (happiness)

[g] : ukugula (to be ill) [g] :ingubo (a dress)

ii) Affricativization

The vast majority of mother tongue Zulu speakers pronounce

fricative sounds as affricates whenever the fricatives are

preceded by nasal sounds. Although some may regard this

phonological rule as optional, the widespread usage of the rule

should be noticed.

The affricativation ... is not reflected in the orthography of Zulu,

which means that the same letter is utilised to represent both the

fricative as well as the affricative counterpart of the sound.

The fricative sounds affected are:

The voiceless denti-Iabial fricative [f], represented by the letter

If I in Zulu; the voiced denti-Iabial fricative [v], represented by

58

the letter Ivl in Zulu; the voiceless alveolar fricative [s],

represented by the letter lsi in Zulu; the voiced alveolar fricat~ve

[s], represented by the letter Izl in Zulu; the voiceless alveo-

lateral fricative [f]; the voiceless palatal fricative [I] the voiced

palatal fricative.

The affricate counterparts of these are represented by the same

letter as that of the fricative sound. Second language learners of

Zulu tend not to discriminate between the [+affricate] and

[+fricativel pronunciation of these sounds and thus pronounce both

as [+fricative].

Table 5.11: Replacement of affricate

Sound Zulu letter

[f] or [<pf] .f

[v] or [<1>V] v

[5] or [Is] s

,

[z] or [dz] z

[i] or [nti] hI..

Examples:

Fricative

[f] ifu (cloud)

Affricate

[<pf] imfene (baboon)

59

5.5.1 The implosive[b]

The implosive bilabial consonant [b] that is foreign to English is

very often replaced by the voiced bilabial explosive [b]. The

implosive sound is represented by the letter Ibl in Zulu while the

same letter is used to represent the explosive sound in English. The

incorrect pronunciation is thus not reflected in the spelling.

For example [0]: ukubonga (to praise)

ubaba (father)

5.5.2 The alveo-lateral [et]The voiceless alveo-lateral fricative [i] is foreign to English.

Second language speakers of Zulu tend to replace this sound with the

voiceless alveolar fricative Is) in combination with the lateral

continuant [1). This probably happens because [i) and Is] are similar

and only differ with regard to one characteristic, namely that of

[+- lateral]. Learners compensate for the absence of [+lateral) in

Is] by combining the sound with [1). The sound ri] is represented by

the letters Ihll whlie t he. sound [s] is represented by the letter

/5/. The mispronunciation is thus reflected in the written form.

61

Table 5.12: Replacement of alveo-lateral fricative

Sound Zulu· letters Replacement

[i] hl sl

For example:

[i] Hluhluwe (place name) *slusluwe

Hamba kahle (go well) *hamba kasle

5.6 CONCLUSION AND SUGGESTIONS

Incorrect pronunciation may cause a change in the meaning of the

word or may render. the word meaningless. Mispronunciation of the

sound [ph] in for example, uSipho may change the name to the

uncomplimentary *uSifo. Incorrect pronunciation can cause havoc and

must be avoided.

A 'phonetic comparison of the Zulu .and English sounds sheds some

light on pronunciation deviations in Second language' Zulu speakers.

It seems that the most problematic sounds proved to be those of

which the correct pronunciation differs only to a very slight degree

to that of the replacing English sounds. This is not necessary

reflected in the spelling of the learners.

Sounds foreign to English, such as /hl/ ([i]) are replaced by similar

sounds, in this case [s] accompanied by [1]. It seems that in cases

62

where no similar sounds are to be found in Zulu, such as with the

click or suction sounds, pronunciation proves to be a minor problem.

Transfer from the second language learners' native English is the

main cause of the mispronunciation of Zulu.

This study does not claim to be an exhaustive report on all possible

pronunciation deviations but merely serves to alert both second

language Zulu learners as well as educators to possible pitfalls.

Fisiak (1981: 2) says that applied contrastive studies should not

only deal with differences but also attach importance to

similarities. If the educator points out the forms that are similar,

the learner will not have to guess them.

According to Freeman and Long (1991:55), however, Contrastive

Analysis would help to anticipate trouble spots in the target

language. Errors might be prevented or at least held to a

minimum.

The solution to this difficulty, is given by Flege (1980:117) by

s ayi.nq that a learner must slowly learn to articulate unfamiliar

sounds and to extend production of already familiar sounds to the

phonetic context. He also says that the learner of a foreign

language must learn not only how to produce new sounds in second

language, but also how to modify previously established patterns of

production.

This remains a complex and changing issue that depends on the

learners' acquisition of the first language and the similarities

between the target language and the native language.

63

Appendix A

Incorrect

Thembi: S~wQbona Sipho.

Sipho : Yebosawobona Thembi.

Thembi: Unjani?

Sipho : Ng~khona wena unjany?

Thembi: Nami ngikhona.

64

Appendix B

Corrected: Group 1

Thembi: Su(a)w~bona

Sipho : Yebo sawobona

Thembi: Unju(a)ni?

Sipho : Ng~khona wena unjany?

Thembi: Nami ngikhona.

65

Appendix C

Corrected: Group 2

Thembi: S~wo(u)bona.

Sipho : Yebo sawobona.

Thembi: Unju(a)ni?

Sipho : Nge(i)khona wena unjani?

Thembi: Nami ngikhona.

66

BibliographyAl tenberg E. & Vago RM: 1983 Theoretical Imp l i.ce tions of an Error

Analysis of Second Language Phonology Production. Language

Learning. 33.4. University of Mi.chaqan , Canada.

Avery P & Ehrlich S: 1992 Teaching American English Pronunciation.

University Press. Oxford.

Archibald J: 1998 Second Language Phonology. Benjamin's Publishing

Company. Philadelphia.

Berkoff NA: 1981 Error Analysis Revisited in Error, Contrastive

Linguistics and Second Language Learning edited by Nieckel G. &

Nehls D. Julius. Groos Verlag. Heidelberg.

Corder, SP: 1993 A Role of the Mother Tongue in Language Transfer in

Language Learning, edited by Gass S & SeLi.nke r L. John Benjamin's

Publishing Company. Philadelphia.

Cruttenden A: 2001 Gimson' s Pronunciation of English. Arnold

Publishers. London.

Eckman FR: 1977 Markedness and the Contrastive-Analysis Hypotheses.

Language Learning 27:2. University of Michigan. Canada.

Eckman FR: 1981 On the Naturalness of Interlanguage Phonological

Rules. Language Learning 31:1. University of Michigan. Canada.

67

Fisiak J: 1980 The Contrastive Analysis of Phonological Systems in

Theoretical. Issues in Contrastive Linguistics}, edited by Fisiak J.

Benjamin Publishing Company. Amsterdam.

FisiakJ: 1981 Contrastive Linguistics and the Language Teacher.

Pergamon Press. New York.

Flege JE: 1983 Phonetic Approximations in Second Language

Acquisition. Language Learning 30:1. University of Michigan. Canada.

Flege JE: 1987 Effects of Equivalence Classification on the

Production of Foreign Language Speech Sounds in Sound Patterns in

Second Language Acquisition edited by James A & Leather J. Foris

Publication. USA.

Freeman, L.L. & Long, MH: 1991 An Introduction to Second Language

Acquisition. Research Longman, London.

Fromkin V & Rodman R: 1988 An Introduction to Language. Holt,

Rinehart and Winston, Inc. Orlando

Gass SM & Madden CG: 1985. Input in Second Language Acquisition.

Newbury House Publishers. Rowley.

GassSM & Selinker L: 1994 Second Language Acquisition, Lauwrence

Erlbaum Associates PUblishers. London.

Hatch EM:

Cambridge.

1983 Psychol.inguistics. . Newsbu r y House Publishers.

Jordens P: 1997 Studies on Language Acquisition. Mouton den Grutter.

New York.

6X

Littlewood WT: 1984 Foreign and Second Language Learning. Cambtidge

University Press. Cambridge.

McLaughlin B: 1987 Theories of Second Language Learning. Cambridge

University Press. London.

Mawasha AL: 1973 North Sotho-speaking Children in the Junior

Secondary School with Special Reference to Oral Communication.

Unpublished Doctoral thesis. University of the North. Sovenga.

Norish J: 1983 Language Learners and their Errors. McMillan Press

Limited. London.

Piper T: 1993 And Then There Were Two -- Children and Second

Language Learning. Pippin Publishing Limited. Ontario.

Posthumus LC: n.d. Zulu Phonetics Made Easy. Not published. RAU.

Johannesburg.

Richard PA:1985 Making It Happen. Longman. New York.

Ringbom H: 1981 Influence of Other Languages on the Vocabulary of

Foreign Language Learners' in Error, Contrastive Linguistics and

second language learning, edited by Nieckel G & Nehls D Julius Groos

Verlag. Heidelberg.

Ringbom H: 1987 The Role of the First Language in Foreign Language

Learning. Multil~ngual Matters LTD.' Philadelphia.

Smith t-1S:1994 Second Language Learning. Longman. London.

69

Tarone E: 1974 Speech Perception in Second Language Acquisition.

Language Learning. 24:2. University of Michigan. Canada.

Towell R & Hawkins R: 1994 Approach to Second Language. MUltilingual

Matters LTD. Philadelphia.

70