Difficulties For Vietnamese When Pronouncing English: Final Consonants
Difficulties For Vietnamese When Pronouncing English: Final Consonants
Difficulties For Vietnamese When Pronouncing English: Final Consonants
Engelska C Uppsats
Supervisor: Una Cunningham
Spring, 2007
Nguyen Thi Thu Thao
Personal Number: 841020 - 6083
v07ttngu@du.se
Table of Contents
Abstract
1. Introduction
2. General Aim
3. Background research
3.3. How are English final consonants difficult for Vietnamese speakers?
10
4. Hypothesis
13
5. Method
14
6. Results
15
16
16
18
21
25
29
References
31
Appendices
33
Abstract: Vietnamese people have many difficulties when pronouncing English. Among
those, this paper will firstly deal with the hypothesis English word-final consonants are
not pronounced in a native-like way by Vietnamese speakers. Theoretical phonological
research about final consonants in the Vietnamese language and English has been
carried out to characterize the difficulties. Data from Vietnamese informants were
collected and analyzed, then synthesized to the most significant problems. Vietnamese
effort to pronounce English word-final consonants will be towards omitting, adding
schwa or replacing by sounds closer to those existing in their mother-tongue. Results of
native speakers evaluation of Vietnamese-accented final consonants are also concluded
to clarify how comprehensible informants pronunciation is. These findings will hopefully
be useful for those who are interested in the topic and for further research.
1. Introduction
As a developing country, Vietnam currently has an urgent need for the
international language English to attract foreign investment as well as promote the
economy, especially after becoming a member of the World Trade Organization. English,
previously included in the curriculum from the first year of middle school, is now
officially taught for children from six years old. Foreign language centers are
dramatically increasing in number, and English is the most popular foreign language
offered for learners. Although English has become an important demand for schooling
and job opportunities, Vietnamese people cannot pronounce English properly. Like some
other languages, Vietnamese has phonotactic features that keep native learners from
pronouncing English like native speakers. The limitation of Vietnamese word-final
-1-
sounds and the frequency of English word-final consonants errors which are made by
most Vietnamese speakers have caught significant attention to this area of the topic.
Initially, this paper will investigate the difficulties of Vietnamese speakers of English
when dealing with English word-final consonants. This is considered one of the most
significant problematic phonetic features and it attracts much interest from linguists and
teachers of English inside and outside of Vietnam. The findings of this research will
hopefully help Vietnamese speakers become aware of their systematic errors and
carefully try to produce native-like English word-final sounds. Further research by the
same researcher will be carried out to give broader and more sufficient consideration
about the whole topic.
2. General Aim
This research is aimed at offering a systematic study of the articulation of wordfinal consonants of Vietnamese-accented English. Firstly, the differences in word-final
consonants of the two languages are considered in background research. In the second
part of this paper, recorded data and evaluations from native speakers are analyzed.
Difficulties and tendencies in informants final consonants and cluster pronunciation are
synthesized into certain types. The comprehensibility of these productions is then worked
with in order to get a general assessment of how native-like Vietnamese speakers
English final consonants are. These findings are then related to theory and discussed
further. Finally, the conclusion of the whole essay summarizes and comments on the
findings of this research.
-2-
3. Background research
3.1. Vietnamese final consonant:
Vietnamese phonetics and phonology are not sufficiently and scientifically
studied by local as well as foreign linguists. Actually, some serious studies about
Vietnamese are controversial. The problems of which varieties should be chosen as
standard Vietnamese, or which International Phonetic Alphabet (IPA) letters should be
used to transcribe Vietnamese sounds still remain unanswered. The background studies
for this paper are a combination of many sources and points of view with references
given afterwards, but the IPA letters used are mostly adopted from the official textbook
for students of linguistics studies in Vietnam in which phonetics and phonology was
written by Vietnamese leading phonetician on Thin Thut (Nguyen et al., 2006).
None of the varieties is specifically used to transcribe the examples, and transcription for
prescriptive Vietnamese pronunciation is used.
Vietnamese is one of the syllable-timed languages which each syllable is short
and simple in construction. The structure of each syllable is (C) V (C). Taiwan Buffalo
International (2001) characterized syllabic structure in Vietnamese by this table:
Tone
Final (rhyme)
Initial
Onset
Nucleus
Coda
-3-
only the nuclear vowel is compulsory, which is shared by Doan Thien Thuat (Nguyen et
al., 2006) and officially accepted by Vietnamese educators:
Tone
Initial
Nuclear
Final consonant/
vowel
Semi-vowel
Labialization
consonants
(Ngo, 2005: 7)
Numbers of codas available in Vietnamese are limited to a certain degree,
specifically, there are only six consonants and two semi-vowels which can stand in wordfinal position. Following are details about final consonants in Vietnamese. The IPA for
Vietnamese vowels are listed both in Nguyen et al. (2006) and adopted in Ngonngu.net
(2006), consisting of thirteen monothongs and three diphthongs /i, e, , 7, 7<, a, M, , u,
o, O, O<, E<, ie, M7, uo/. The articulation of these monothongs and other consonants is
clearly illustrated
pronunciation of final consonants are rounded /u, O, o/ and front /i, e, E</. Details
about rules of Vietnamese final consonant pronunciation are adopted from on (1999),
and the examples are taken from the Vietnamese dictionary (Hong, 2000).
i) /m/: in words like
em [Em] (I, younger sister/brother)
lm [lM7m] (pick up)
nghim [NIem] (strict)
ii) /n/: in words like
n [a<n] (eat)
-4-
lm [lam] (do)
phin [fien] (turn)
iii) /N/: there are 3 allophones of this phoneme:
[N m]: bilabialised, proceded by rounded vowels /u
, O, o/
xong [sON m] (finish)
sng [SuN m] (gun)
khng [XoN m] (no, not)
[J]: corresponding to letters "nh", proceded by fr
ont vowels /i, e, E</
bnh [beJ] (sick)
tnh [tiJ] (love)
nhanh [JE<J] (fast)
[N]: elsewhere
tng [ta<N] (give gifts)
thing [thieN] (supernatural)
chung [cuoN] (favourite)
iv) /p/: with no air released after pronunciation, in words like
p [up] (up-side-down)
Php [fap] (France)
tip [tiep] (continue)
-5-
v)
/t/: In English, [t] ending is pronounced both as [t] and [?], but in
Vietnamese it is always pronounced shortly and sharply without aspiration such as:
t [7t] (chili)
git [ziet] (kill)
ght [gEt] (hate)
vi) /k/: this sound is produced where the letter c, ch is shown at the final position
of a syllable and it is agreed by many linguists that this phoneme has 3 allophones
which appear in a complementary distribution.
[k p]: this ending sound is bilabialised if the co
nsonant is preceded by rounded vowels /u, O, o/
ngc [NO<k p] (pearl)
cc [kok p] (glass, cup)
nhc [Juk p] (insulted)
[c]: preceded by front vowels
med by letters "ch"
nghch [Nic] (naughty)
lch [lec] (askew)
sch [SE<c] (book)
[k]: elsewhere
nhc [Jak] (lazy)
bc [bMk] (angry)
-6-
Thut
(Nguyen
et
al.,
2006)
transcribed
these
two
sounds
as
/u^/ and /i^/ while some other researchers such as Ngo (2005), Osburne (1996)
and articles of Taiwan Buffalo International (2001) use /w/ and /j/. The involvement of
these two semi-vowels in this research might appear unnecessary due to their weak
relation to the investigation of Vietnamese final consonants. However, they are important
hints for Vietnamese-accented English.
-7-
Syllable
Onset
Rhyme
Nucleus
Consonant (s)
Vowel
-8-
Coda
Consonant (s)
It is widely agreed that syllabic structure of English is (C) (C) (C) V (C) (C) (C) (C).
Almost every English consonant appears word-finally, except for /h/, /w/ and /j/
(Cummins, 1998). Besides, English has a lot of complicated consonant clusters, as
mentioned previously, combining two, three and even four consonants together in the
onset and coda. This is commented as quite unusual for languages to have consonant
clusters of this type (Yule, 2006: 48).
Spencer (1996) shares Yules (2006) ideas about the English coda: The simplest
way of handling these data is to say that English permits any number of coronal
obstruents to be tagged onto the end of a syllable. This would seem to make the coda
rather a complicated animal (Spencer, 1996: 99). He also gave many examples of ending
clusters containing two consonants such as lamp, film, health, grasp; three consonants
such as prompt, against, next and four-membered clusters from plural of past tenses such
as glimpsed, sculpts, twelfths, texts, sixths and thousandths, and added: very few
languages seem to have this sort of property (Spencer, 1996: 86).
For pronouncing English consonants both at onset or coda to the extent that
phonetics is part of an exact scientific discipline (Ladefoged, 2001: 56), nineteen rules
were also listed which might require learners of English rough effort to be adequate
(Ladefoged, 2001: 57-60). Yule (2006: 48), however, suggested that It is also noticeable
in English that large consonant clusters are frequently reduced in casual conversational
speech. Complex consonant clusters are omitted by speakers to simplify the process of
articulation; which is moving towards the idea of using comprehensible rather than
standard English pronunciation that is shared by many linguists and will be discussed in
later parts of this paper.
-9-
What may be concluded from English phonotactics is that English permits any
number of coronal obstruents to be tagged onto the end of a syllable, and for coda
clusters: the order sonorant + obstruent (Spencer, 1996: 86-99). It is apparently
noticeable and widely accepted that English consonant clusters are one of the most
complicated and unusual phenomena of language. Besides many sounds that are not
involved in linguistic cognition of non-native speakers, its consonant clusters as well as
phonetic changes in consonant quality in formal and informal speech prevent non-natives
from accessing and adopting English pronunciation fully. This problem not only belongs
to speakers of Asian origins whose native languages are mostly mono-syllabic, but also
for other parts of the world, even those who share the same language family with English.
Vietnamese is not an exception.
3.3. How are English final consonants difficult for Vietnamese speakers?
Trn (2005) stated that Vietnamese phonology seems to be more complicated
than English, but in the long run, this system becomes simpler and more acquisitive than
English. He added: it may take a non-native speaker of Vietnamese two months to
learn Vietnamese phonology profoundly, and then he will surely be able to pronounce
any Vietnamese word; whilst British or American people during their whole life have to
learn how to pronounce new words continuously with a lot of support from dictionaries
(my translation). Vietnamese is supposed to be easy to phonetically acquire when
speakers have an efficient input, especially of the tones. That is one of the reasons why
Vietnamese native speakers have to deal with many obstacles to learn foreign languages
- 10 -
which are not relatively close to and as easy as their language, for instance, Russian,
French, English or Spanish.
There has been quite a number of studies about Vietnamese difficulties in
pronouncing English consonants and clusters. These have led to important findings,
which become a valuable basis for further studies, most visibly, for this paper. Ha (2005:
35-46) after analyzing her data finally came to these conclusions on the left, in
comparison with the table formed by Center for Applied Linguistics (Neumann, 2007).
Research done by Ha (2005), though data were from numerous informants, is unilateral,
containing only problems of people from the North by the confusion most obviously
between /s/ and /S/, /tr/ and /tS/, /Z/ and /z/. Those findings, which
were mentioned in the table by Center of Applied Linguistics (Neumann, 2007), are
applicable for every Vietnamese learner as it contains all the errors that Vietnamese
speakers of any regions of the country can make. They are also persuasive to linguists
- 11 -
who have certain knowledge about Vietnamese dialects. Taking final consonants into
consideration, for example, /T/ and /D/ at word-finally are commonly confused with
/t/ and /d/ everywhere, whereas /p/ sound in /pOp/ pop is often mispronounced as
/bOp/ Bob by Southern people and /S/ sound in /puS/ push becomes /pus/ puss by
northerners.
Tang (2007: 7) offers a comparison table below:
Studying this table specifically for word-final consonants, there are some points that
conflict with the background of Vietnamese and English consonants described
- 12 -
previously, for example the /N/ sound in sng is bilabialized [N m] and different from
the /N/ in song, so sang (come over) and sang (past tense of sing) seem to be a better
sample couples since they are produced with exactly the same articulation. Furthermore,
Tang (2007) did not list four-membered consonant clusters that are absolutely foreign to
the Vietnamese language. Nevertheless, this material is quite detailed and should be
appreciated. It is easily seen from this data that English has a number of consonants,
especially final consonants and clusters that do not exist in Vietnamese rather than vice
versa. As a result, pronouncing English final consonants and consonant clusters properly
is one of the most difficult things that learners have to face from the very beginning.
Osburne (1996, 164-181) analyzed a case study from her subject - a Vietnamese
native speaker who came to the United States in 1972 - then drew the conclusion that: In
additional to cluster reduction, optional deletion of single syllable-final consonants,
especially fricatives, which is attested for Vietnamese L1 speakers [] was found, and
Consonants omitted, however, were always final consonants not permitted by
Vietnamese (for example /l/ in [k@nthro] control, /z/ in [bIkhV] because)
(Osburne, 169). She also stated that Vietnamese is non-rhotic so there is no /r/ sound at
the end of English syllables spoken by Vietnamese.
What can be concluded from the above studies is that Vietnamese learners have a
tendency to: firstly move strange English ending sounds towards similar sounds which
exist in their mother tongue, secondly omit the sounds that are too difficult for them and
thirdly reduce final clusters. This may make their English very Vietnamese, which
causes some problems for communication with native speakers and others. The rest of
- 13 -
this paper will deal with the analysis of English word-final consonants pronounced by
Vietnamese speakers.
4. Hypothesis
English final consonants are not pronounced in a native-like way by Vietnamese
speakers.
5. Method
Data from five informants of various Vietnamese accents and proficiency in
English were recorded inside and outside Vietnam then sent to the researcher by e-mail.
The first informant is a twenty year-old male from Hue, which is at the centre of Vietnam
with a very different accent from the north and the south. He began learning English three
years ago in high-school classes as well as extra courses. However, his English is still at
pre-intermediate level, and there are some words in the text he does not know. The
second informant has been exposed to English for eight years in Ho Chi Minh City which
is situated in the south of Vietnam. She is also twenty and attends extra communicative
English classes. The next informant is a female in-service bachelor of English. An extra
thing should be mentioned is that there are two kinds of university education in Vietnam:
full time main-stream and night shift in-service for those who are working and feel the
need to study more. The difference between education quality of the two types is quite
visible. In-service students are considered not as good as main-stream ones due to their
age and their lack of time. This informant is thirty-seven years old, from the West
Highland part of the country. She started studying English when she was twelve but does
- 14 -
not have much chance to use her English. The fourth is a twenty-three year-old male from
the capital Hanoi who came to Sweden to study English four months ago. Like the other
informants, he began to learn English from twelve and had English classes continuously
at schools since then. The last informant is a special case, who started learning English
since the sixth grade, became very interested in the subject and won the third prize in a
national competition in English which is awarded to the very few best students
countrywide. He is twenty-three and has left Ho Chi Minh city for California in the
United States for nearly five years. In the hope of creating an objective analysis, the data
were chosen from speakers of different parts in the country and abroad, distinguishing
accents of Vietnamese with various levels of English proficiency. Problematic words and
a short text used for recording are given in Appendix 2. Cool Edit Pro 2.0 and Windows
Media Player were used for cutting and analyzing final consonants produced.
After analyzing the data, an evaluation chart given in Appendix 3 was delivered to
six native speakers of English, in which the evaluators demonstrated their impressions of
Vietnamese-accented English. All of the six evaluators have English as their mothertongue, three of whom are American and others are from England, Ireland and Australia.
Furthermore, they all have experience in communicating orally with non-native speakers
of English everyday. Four of them have a lot of exposure to Vietnamese speakers of
English and the rest do sometimes. This information becomes reliable as the evaluators
have prepared their ears and their minds for understanding and assessing those data
according to their experience. The purpose of collecting this information is to figure out
how native-like Vietnamese-accented English word-final consonants are from the ears of
- 15 -
native speakers besides electronic analysis. Hopefully, the findings will be more exact
compared to acoustic studies and human feelings.
The results are presented mostly in tables and charts, then explained by theoretical
studies. Discussions and applications are to be pointed out on the basis of the findings.
6. Results
The results of the analyzed data are divided into three categories: single final
consonants that exist in Vietnamese or sound similar to Vietnamese, single final
consonants that are foreign to Vietnamese and final consonant clusters. Evaluation by
native speakers also comes as the last part of findings.
6
16
14
5
12
4
10
n
1
p
38
2
12
00
- 16 Total
in Speaker
Speaker11 Speaker
Speaker22 Speaker
Speaker33 Speaker
Speaker 44 Speaker
Speaker 55
Total
in List
Text
Chart 1: Number of errors with final consonants that are closely related to Vietnamese
coda
The most salient feature can be found is that nearly all /p, t, k/ are unaspirated.
Speaker 5, although his English is quite good, still makes systematic errors with aspirated
/p/ but not the same error with /t, k/. The fourth informant, in an attempt to avoid the
mistakes, has carefully pronounced every /p/ with aspiration but /t, k/ are still unaspirated. However, when it comes to the text, he makes all three /p/ unaspirated. The
other three speakers cannot pronounce any /p, t, k/ with aspiration in the final position as
native speakers. Nevertheless, this is not considered a serious problem for mutual
understanding. Native speakers are sometimes also lazy to aspirate in casual speech.
Secondly, all /m/ sounds are predictably pronounced perfectly but not all of /n/
sounds are right. Apparently, down is pronounced as [dau^^] by every informant
regardless of proficiency or accent. The second speaker with a southern accent
pronounces both green and grin as [gri:N], in as [iN] and omits /n/ in govern.
The confusion between /n/ and /N/ is also found by the third informant. /N/ appears only
once in long, and like /m/, there is no difficulty for Vietnamese speakers to pronounce it
correctly.
Thirdly, the results with /t, k/ from the five informants surprisingly introduce
quite a lot of omission and errors in final position. Although they do exist in Vietnamese
codas, /t/ in comfort, about, might, great, thought is not pronounced. Most of them end
with the vowels right before them, some become /k/ or /T/ due to speakers accents and
acquisition. /k/ ending in back and make turn into palatals with a front vowel [bE<c] and
- 17 -
7
6
s
z
f
v
3
'
d
1
0
14
12
s
10
f
v
'
d
l
2
0
Total in
Text
Chart 2: Number of errors with final consonants that are foreign to Vietnamese
coda
As /s/ is one of the popular sounds of English, Vietnamese learners have better
awareness of pronouncing this sound. This can be found in the comparatively small
number of errors made with final /s/ in the word list in careful speech. On the contrary, it
is still a foreign consonant to them so they cannot avoid these errors when it comes to the
text. Four speakers have their house as a homophone of how. The situation is worse with
/z/ in word-final position. His, has, because, places, goes, is final sounds are all vowels
for the first speaker. The second one pays more attention to /z/ sounds but choose, places,
trees, is become [tSu:], [pleI], [tri:] and [i:] several times. The third
informant makes the same errors, and her pronunciation of /z/ in his is confused with
unaspirated /t/ in hit.
What can be seen from the articulation of /f/ by these speakers is generalized in
the movement towards an unaspirated bilabial /p/ sound, so that if becomes [Ip]. The
- 19 -
same phenomenon is found with /v/ when of is reduced to [@p] by most speakers.
Besides, have is pronounced unstably as [h{], [h{p], or [h{u^] by different speakers
while believe is found correctly in three cases, in other two as [bI"li:]. There is no
rule that this foreign sound should move to another similar sound in their mother tongue.
It depends on individuals as well as words followed that sound.
With is pronounced as [wIs], a substitution for [D] in two cases, and others as
unaspirated voiceless /t/ in [wIt]. The same movement commonly happens to /d/ in
could, would. However, there are situations where /d/ in could and would becomes /s/ and
/k/ varying by speakers regional accents. Most significantly, side and sigh are
homophones for four of the informants.
Last but not least, final /l/ is considered as the most stable and systematic error
that all informants face at least once or twice in the data. /l/ in school, small, pull, pool
becomes /n/ as [sku:n], [smon], [pUn] and [pu:n] in most cases and becomes a
semi-vowel /w/ or /u^/ in feel, still, until, will after the vowels /I, i:/ which is very
similar to vocalized dark l in many dialects of English. Once it is heard as [smO:] for
small which is followed by an initial bilabial /b/ in boy. This is an acceptable assimilation
of sounds in normal speech. Similarly, /l/ is pronounced properly in all by every speaker
because it comes before the curled initial in round.
Generally, there is certainly confusion between voiced and voiceless sounds,
which happens fairly commonly between /t/ and /d/, /s/ and /z/. Besides, omission of
problem sounds is popular. A tendency to move foreign sounds towards the first
languages coda is also apparent. These logically lead to a mispronunciation of final
clusters.
- 20 -
2.5
2
-vz, -ldz, -lz, -ndz
-nts, -ts
1.5
1R
-lf
0.5
0
Total in
List
below:
- 21 -
12
10
-vz, -ldz, -lz, -ndz
8
-nts, -ts
-lt, -st, -nt
-ld, -nd
1R
-lf
4
2
0
Total in
Text
Less numerous as the problem final clusters are, it is recognizable that the number
of errors in comparison with the number of all them found in the list and text is of quite a
high proportion. Most significantly, round is pronounced as [raU], adult as [@dVl] or
[@dVn], wouldnt as [wUdn] by all speakers. Three-membered clusters such as /ldz/
in fields, /-ndz/ in friends and /-nts/ in wants are impossible for four informants but not
for the last. Two-consonant clusters that contain /l/ in /-lz/, /-lt/, /-ld/ or /-lf/ are observed
to be really difficult for all speakers since the sound itself never appears in their first
language coda as in Chart 2. When it comes before another consonant, it is totally
omitted, and in most cases the final consonants also disappear so that feels become
[fi:U], world becomes [w3:], himself becomes [hImseu^] and told turns into
[ton] or [toU]. The disappearance of /l/ in such a position is, however, acceptable
in the sense that some accents of English like Cockney, Estuary and African-American
Vernacular English also turn it into dark [l] which is observed as a vowel or a semivowel.
- 22 -
/Nk/ in one and only word found in list as well as text think is another interesting
problem to observe, as four over five speakers pronounced it as thing by the deeply
rooted effect of their mother tongue. Clusters ending with voiceless /s/ and /t/ also
cause quite a lot problems for informants.
The most remarkable error found in final clusters of these informants is the
addition of schwa so that lives turns into [lIv@z], leaves becomes [li:v@z] and believes
[bI"li:v@z]. The reason why schwa only happens in these words may be due to the
spellings that mislead the two female informants. Other speakers just omit the clusters
and end their word with a vowel, or simply try to pronounce one of the consonants.
Generally, final clusters are one of the most difficult categories in English pronunciation
for Vietnamese speakers. They are so different from their mother tongue. The problems
are found in the tendency to reduce the number of members of clusters or even omit them
totally. Schwa is also used as a solution for the ease of speech.
In conclusion, English final consonants are not produced like native speakers by
these informants. The consonants nearer to Vietnamese including nasals and unaspirated
voiceless stops are identified as their first languages own coda. Dealing with consonants
that are new and different from their coda, they either omit them or reduce them to the
sounds that exist in Vietnamese. Clusters are also omitted or reduced, and schwa is added
between consonants to make new syllables in some cases.
Final consonants can convey meanings since English has so many words that
differ only in their coda and they also play important roles in grammar for singular or
plural as well as tenses. The inability to produce exact word-final consonants may
prevent speakers from being understood by others when speaking English, and more
- 23 -
Speaker 1
Speaker 2
Speaker 3
Speaker 4
Speaker 5
16
15
16
12
It is easily seen that the first and third speakers with 16 points have been
considered as very difficult to understand. Referring to data analysis and charts 1, 2 and
3, these two speakers have the most errors with unaspirated /p, t, k/, with ending /z/ and
/l/ as well as clusters. The evaluation is reasonable enough as the first informant has
- 24 -
studied English for only three years in a small city, and the other, in spite of being a
bachelor of English, did her degree at quite an old age with a naturally lower ability of
acquisition and the lack of a communicative educational approach. The second informant
is better with the compromising between very and a little difficult for native speakers to
comprehend with 15 points. This speaker pronounces voiced consonants quite well but
has significant problems with nasal confusion. The fourth, who gets 12 points, is
generally assessed as a little difficult to understand with his final consonants. He makes
quite a lot of errors with word-final /s/ and /l/, but better than the other three with clusters
and aspiration. The last, being one of Vietnams brightest students of English and having
stayed in the United States for some years, is rated 8 points with the best choices among
five speakers. The data analysis also showed that his English final consonants are clear
and comprehensible enough; however, unfortunately, the volume of the recording is quite
low to correctly recognize the sounds. One of the evaluators admits that she just can
guess his final consonants sometimes.
The most important thing which directly gives the answer to the hypothesis of this
paper is that none of native speakers consider any of these five informants produce their
final consonants and clusters like native speakers. This may lead to the conclusion that
Vietnamese speakers of English have to face so many problems to produce native-like
final sounds. It may take learners a long time of extreme practice and improvement to
pronounce them like native speakers.
7. Discussion of results
- 25 -
The findings concluded after processing the recorded data turn out to be
predictable compared with those mentioned from the background research. The
movement towards similar sounds in Vietnamese is best demonstrated by the relation
between Vietnamese phonotactics and the results of data analysis. Those are the
unaspiration of /p, t, k/, the simplification of /l/ into /n/ or /u^/, the replacement of /k/ by
[c] after vowels that are similar to Vietnamese front vowels [e, E<] in make and
back. The confusion between voiced and voiceless final sounds is a result of the
limitation into voiceless coda or nasals in Vietnamese. All word-final consonants that are
not nasals should be voiceless, so that voiced final consonants in could, would, with, of,
have mostly become voiceless. /k/ in like, /d/ in side, /t/ in about and /n/ in down are
reduced into [laI], [saI], [@"baP], [daP] because in Vietnamese diphthongs or
triphthongs, /i^/ and /u^/ must be the last sound of a syllable and no other consonants are
allowed to be attached to them. Other omission of final consonants cases found in the
data can be explained in this way are house, round, told, feels, because, choose, etc.
Finally, as Vietnamese has only one consonant in the coda, final clusters are reduced into
one consonant to be specific: friends [frIen], think [TIN], adult [@"dVt] etc. or adding a
schwa [@] between the members to make two different syllables as a result of the monosyllabic mother tongue. Another thing which should be mentioned is that there are certain
points in the text with natural speech that pulled speakers to correct and acceptable
pronunciation. For instance, and is listed in the number of words containing in the
material but is not counted as errors of Vietnamese informants as most of the time it
appears in its weak form with the reduction of final clusters. Some more can be found are
the assimilation and reduction of final consonants in all round [O:raUnd], round the house
- 26 -
[raUnD@haUs], have friends [h{frIendz], and do [{ndu:] and have to [h{ft@] etc. These
types of pronunciation are not counted as errors in the results as native speakers would do
the same in natural speech. The findings of this research also point out an effort of these
informants, who are objectively chosen among Vietnamese learners and users of English
as a foreign language, to gradually move their pronunciation towards comprehensible
English first, and native-like English later, rather than omit all the sounds that are foreign
to them. The younger generation has more motivation and ability to overcome the
difficulties to improve their English due to the fact that a thirty-seven year-old bachelor
of English gets the same points of incomprehensible accent as just a three-year learner of
English. The older generation will soon be left behind by young beginners, preintermediate and intermediate learners of English. This also shows a promising
development of English among Vietnamese speakers with better and better quality.
The selection of informants from different Vietnamese accents is reasonable
enough due to the fact that their English pronunciation in general can be found to be
affected by elements of their own accents. The most apparent feature found in final
consonants belongs to speaker of Ho Chi Minh city as a southern part whose ending /n/
becomes /N/ in green, grin and even in; and speaker of West Highlands whose ending /t,
d/ becomes /k/ in thought, could and would. Those are not easy to see from the northern
informant because the effect of his accent on initial consonants is more prominent with
the confusion between /tr/ and /tS/ in trees, /s/ and /S/ in shut, ship, sheep. The speaker
from the central region, where pronunciation of Vietnamese is considered clearer but the
intonation is quite flat and difficult to comprehend for people from other regions of
Vietnam, is not an exception. This element also shows its effect on the speakers English
- 27 -
intonation. The fifth who is at an advanced level of English seems to successfully escape
from those features of his southern accent. These would bring some ideas about later
research on the affect of Vietnamese accents into their English. The findings of this
paper, hopefully, will become useful for those who are interested in Vietnamese-accented
English, those who need to communicate with Vietnamese speakers by means of English
and teachers of English for Vietnamese learners. By grasping the tendency towards
omission, movements towards mother tongue and schwa addition of final consonants
might help foreigners understand Vietnamese speakers English more easily. Teachers
might notice the problematic sounds and the unsatisfactory substitutions to pay more
attention in leading learners to the exact or comprehensible sounds.
Native speakers evaluation also brings interesting findings to this research.
Whether Vietnamese-accented English is native-like or not should be judged by native
speakers themselves. Objectively, English final consonants, especially clusters, are
difficult for non-native speakers and, as observed by many phoneticians, not all these
sounds are pronounced properly by natives in casual situations. That is to say the idea of
native-like word-final consonants is quite theoretical, and the evaluation of these
informants ending sounds may be theoretically carried out. None of the six evaluators
assessed any of them as native-like, which strongly illustrated the very real low
probability that Vietnamese speakers will speak English with native-like pronunciation.
Nevertheless, the brighter side of the picture is the proportion of easy to understand and
a little difficult to understand is comparatively much more than very difficult to
understand. Consequently, it is possible for natives and other speakers of English to
- 28 -
omitting the ending sounds that are too foreign or difficult for them,
especially those who contains semi-vowels /i^, u^/ before final
consonants
ii)
reducing and moving the final consonants and clusters towards their first
language like /s, z, l, v, f, d, D, -nd, -Nk, -st, -ld, -ldz,/ to single
unaspirated /p, t, k/ or nasals /n, N/ or semi-vowels /i^, u^/
iii)
- 29 -
beginning stage. The findings are therefore very limited to certain degree. The word list
and text should have been chosen critically so that foreign ending sounds to Vietnamese
like /dZ/, /S/, /T/, /tS/, /g/, /b/ and more clusters would be looked at, and the results would
be more interesting and convincing. Besides, two of the recordings are not of good
quality. One of them is invaded by background noise and the other is recorded from quite
a distance from speakers mouth, so all the final consonants and clusters are diagnosed to
be there but very hard to identify exactly. This experience should be avoided in later
research.
The method of this research, however, is an objective and sufficient approach to
the topic, in which theory cooperates well with analysis of real data from Vietnamese
informants of various accents and levels of English as well as evaluation of native
speakers with much experience in communicating with non-natives around the world.
The results, in this sense, are reliable. These findings can be used for further researches
on the comprehensibility of Vietnamese-accented English for other speakers of English
as first, second or foreign language, or the typical effect of the mother tongue on
Vietnamese speakers English. Non-native speakers evaluation on Vietnamese English
as well as more research on English vowels or initial consonants quality, word stress and
intonation produced by Vietnamese learners and users of English are also suggested.
- 30 -
References
Cummins, Fred (1998) Sound patterns in Human Language: Syllables revisited.
http://cspeech.ucd.ie/%7Efred/teaching/oldcourses/phonetics/syllables2.html
on, Thin Thut (1999) Ng m ting Vit. H Ni: NXB i hc Quc gia H Ni.
Ngo, Nhu Binh (2005) The Vietnamese learning framework Part one: Linguistics.
COTSEAL Journal 2005.
http://www.seasite.niu.edu/jsealt/past%20issues/volume%2010/VN%20LLF%20Part%20
I.pdf
- 31 -
Nguyn, Thin Gip and on, Thin Thut et al. (2006) Dn Lun Ngn Ng Hc. H
Nam: NXB Gio Dc.
Osburne, Andrea G. (1996) Final cluster reduction in English L2 speech: A case study of
a Vietnamese speaker. Applied Linguistics 17.
Yule, George (2006) The Study of Language. Cambridge: Cambridge University Press.
- 32 -
Appendix 1
Vietnamese vowels:
The articulation of /i, e, , 7, a, M, u, o, O/ is
illustrated in the vowel chart beside.
The others /7<, , , O<, E</ are produced the
same as /7, a, O, / but with an extra-short
duration. These four are separate vowels rather
than allophones or weak forms of the above vowels.
(Nguyen et al.,
2006)
They are also performed by different letters in Vietnamese alphabet. The symbols /i^/
and /u^/ used in the research are non-syllabic and appear at the end of Vietnamese
syllables.
Vietnamese consonants:
(Neumann, 2007)
- 33 -
Vietnamese northern dialect does not include the initial /w, j/ but has some more initial
consonants such as bilabial /v/, alveolar /z/ and glottal /?/ according to Ngonngu.net
(2006).
- 34 -
Appendix 2
adult
longer
sheep
feel
because
man
pull
friends
this
places
shut
govern
believe
pool
small
great
boy
country
still
green
choose
quickly
think
thought
comfort
room
become
high
could
run
grin
house
ship
school
through
leaves
day
see
trees
like
very
would
window
unhappy
A small boy lives in this house. There are fields with sheep all round the house.
His room is at the back and he can see his school from the window through the green
leaves of the trees if he wants to pull them to one side.
He feels very unhappy because he has no friends and he believes that if he could
become adult quickly he wouldnt have to go to school. If he could choose, he would like
to govern the country and think great thoughts about the world and have friends in high
places. But he is not yet a man and he must still shut up and do what he is told.
One day he might run away from school and make his way to another country in a
ship. But really, it is not long until he will no longer be a boy. He can comfort himself
with that thought. He starts to grin and goes down to the pool for a swim.
- 35 -
Appendix 3
Evaluation of Vietnamese Accent
Evaluator Information : (Please type or highlight the answers if you do this by computers)
1. Your first language: __________________________________________
2. a. Do you have experience in communicating orally with non-native speakers of
English?
Yes
No
Sometimes
Never
Evaluation Table:
There will be 5 Vietnamese speakers reading a list of words and a text. Please
listen and put an X in the boxes that correspond to your impression in the below table.
Speakers
Speaker 1
Speaker 2
Speaker 3
Speaker 4
Speaker 5
- 36 -