John Benjamins Publishing Company
his is a contribution from English World-Wide 31:2
© 2010. John Benjamins Publishing Company
his electronic ile may not be altered in any way.
he author(s) of this article is/are permitted to use this PDF ile to generate printed copies to
be used by way of ofprints, for their personal use only.
Permission is granted by the publishers to post this ile on a closed server which is accessible
to members (students and staf) only of the author’s/s’ institute, it is not permitted to post
this PDF on the open internet.
For any other use of this material prior written permission should be obtained from the
publishers or through the Copyright Clearance Center (for USA: www.copyright.com).
Please contact rights@benjamins.nl or consult our website: www.benjamins.com
Tables of Contents, abstracts and guidelines are available at www.benjamins.com
Rhoticity in Brunei English
Salbrina Sharbawi and David Deterding
University Brunei Darussalam, Brunei
We might expect Brunei English to be non-rhotic, as the Englishes of both
Singapore and Malaysia are non-rhotic and Brunei has strong ethnic, historical,
economic and cultural ties with those two countries. he current study compares the R-colouring of read data from female undergraduates in Brunei and
Singapore, and it inds that the Brunei data is substantially more rhotic than that
of Singapore. It is suggested that this is for two reasons: the main indigenous
language of Brunei is Brunei Malay, which is rhotic; and Brunei English is at an
earlier stage of development than Singapore English and so it is more susceptible
to outside inluences, particularly from American media.
Keywords: pronunciation, Brunei English, Singapore English, American
inluences, rhoticity
1.
Introduction
Varieties of English can be described as rhotic or non-rhotic. In rhotic varieties,
/r/ occurs wherever there is an 〈r〉 in the spelling, including before a consonant
or a pause. In contrast, non-rhotic varieties only allow /r/ to occur before a vowel
sound (Crystal 2003: 400).
Some people refer to the /r/ in words such as four or cart in rhotic varieties as
“postvocalic /r/”, though we may note that /r/ can actually appear in postvocalic
position in all varieties of English in the middle of words such as marry and as a
linking sound in phrases such as four eggs (Roach 2009: 115), so it is more accurate
to use the term “non-prevocalic /r/” for the occurrence of /r/ before a consonant or
a pause in rhotic accents (Trudgill and Hannah 2008: 11). his is the term that will
be used here. Occurrence of non-prevocalic /r/ is sometimes called “R-colouring”
(Wells 1982: 139), a term that is also used in this paper.
American, Scottish and Irish Englishes are three well-known examples of
rhotic varieties, though the English spoken in New York, Boston and the conservative south tends to be non-rhotic (Wells 1982: 76, 220). Some well-documented
English World-Wide 31:2 (2010), 121–137. doi 10.1075/eww.31.2.01sha
issn 0172–8865 / e-issn 1569–9730 © John Benjamins Publishing Company
122 Salbrina Sharbawi and David Deterding
non-rhotic varieties are Received Pronunciation British English (RP) and the
Englishes of Australia, New Zealand and South Africa (Wells 1982: 76, 220, 542),
though there is some variation. For example, there are many rhotic speakers in
Otago in the south of New Zealand, largely because of extensive migration from
Scotland (Hay, Maclagan and Gordon 2008: 98).
Singapore English is usually described as non-rhotic (Low and Brown
2005: 135; Deterding 2007: 21). However, Tan and Gupta (1992) report that use of
non-prevocalic /r/ is a prestige feature for some speakers, and they suggest it may
indicate a sound change in progress. Poedjosoedarmo (2000a) also documents
evidence for some degree of rhoticity in Singapore English, possibly as a result of
inluence from American media, and she notes that in reading tasks, the percentage of non-prevocalic /r/ tends to be higher when speakers are reading a wordlist
than a passage (which tallies with the classic research of Labov 1966 on New York
City). Of the Singapore Malay speakers in her data, 44% have at least one instance
of non-prevocalic /r/ in a wordlist task as opposed to only 13% when reading a
passage. Eu (2004) similarly reports instances of rhoticity in Singapore English,
though she inds that listeners tend to judge it as pretentious.
Malaysian English is also generally assumed to be non-rhotic (e.g. Rajadurai 2006), and Baskaran (2004: 1 039) transcribes the vowel in word and girl as
/ɜː/ with no following /r/, which similarly suggests that the variety is non-rhotic.
However, some researchers disagree. For example, Hickey (2004: 564) states that
“Malaysian English is also rhotic” and /r/ occurs in words such as art, door, and
worker. Others claim that rhoticity in Malaysian English is a new phenomenon,
just as in Singapore. Although Rajadurai (2006) states that Malaysian English is
non-rhotic because it is derived from British English, she later acknowledges the
increasing inluence of an American accent on the pronunciation of English in
Malaysia, including the use of a lap in words like better, so we might assume
that rhoticity also sometimes occurs. Ramasamy (2005) similarly suggests that the
pronunciation of non-prevocalic /r/ is a new phenomenon in the speech of young
Malaysians. Although that study only considered the English of Malaysian Tamils,
it seems to conirm that Malaysian English is now not exclusively non-rhotic.
To date, there has been little published about rhoticity in Brunei English. Although Mossop (1996) provides a detailed list of many features of the pronunciation of the English of Brunei, including the absence of dental fricatives, consonant
cluster simpliication, and use of a glottal stop in place of inal /k/, he does not
mention rhoticity, though we might note that he transcribes our as [aʊɜ] and the
vowel in board as varying between [ɔː] and [ɔːə] (203), which suggests no rhoticity. In fact, he claims that the close ethnic, historical, economic and cultural ties
between Singapore, Malaysia and Brunei have led to phonological systems “that
are closely related” (Mossop 1996: 189) for the English in the three countries, and
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 123
from this one might expect Brunei English to be non-rhotic. Furthermore, British RP is generally the model of pronunciation promoted in education in Brunei,
something that is reinforced by the continuing presence of expatriate teachers supplied by the Centre for British Teachers (CfBT) in many Brunei schools, and this
adds to the expectation that the variety should be largely non-rhotic.
he current paper investigates the extent of rhoticity in the English spoken in
Brunei using perceptual judgements as well as acoustic measurements, and it further considers some inluences on Brunei English. Comparison is made with data
from Singapore in order to place the research within the framework of a widelyresearched variety of English in South-East Asia and also to facilitate an evaluation
of the current status of Brunei English as an emergent variety of English.
2. Historical development of rhoticity in English
Non-rhoticity in English occurs as a result of phonological change. he loss of
non-prevocalic /r/ was evident in London English from the mid-18th century and
was found in upper class speech in London by the middle of the 19th century,
though some people continued to regard it as “vulgar” for a while (Mugglestone
2003: 87). Rhotic accents such as most varieties of American English can be regarded as conservative as they did not undergo this change.
he date of the loss of rhoticity in southern Britain provides a partial explanation for why some English varieties around the world are rhotic while others are
not. When English was exported to colonial areas before or during the early 18th
century, the resulting variety was a rhotic one, for example in North America.
In contrast, when the export of English occurred ater the mid-18th century, the
variety was more likely to be a non-rhotic one. his is evident with the English
varieties in Australia and New Zealand, as settlers from England irst arrived in
Australia in 1770 and in New Zealand in the 1790s (Jenkins 2009: 7).
However, this is not the only factor. Many people emigrated to America from
Ireland, which continues to be rhotic, while emigrants to New Zealand and Australia mostly were from London and other parts of south-east England which were
the earliest non-rhotic varieties. And, as noted above, there are many speakers with
a rhotic accent in the south of New Zealand as a result of migration from Scotland.
his rationale can be applied to an analysis of the Englishes of Singapore and
Malaysia. English was established in the region in the 19th and early 20th centuries, that is, ater the loss of non-prevocalic /r/ in most parts of England, and few
of the speakers came from Ireland or Scotland.
English was irst introduced into Brunei in the late 19th century, so we might
expect Brunei English to be non-rhotic. However, there is another important
© 2010. John Benjamins Publishing Company
All rights reserved
124 Salbrina Sharbawi and David Deterding
potential inluence: the pronunciation of the indigenous language(s). In Brunei the
most commonly used local language is Brunei Malay, which itself is rhotic. his
issue will be considered further when we discuss the results in the current study.
3. English in Brunei
Brunei was a British protectorate from 1888 till it gained its full independence in
1984 (Hussainmiya 2006: 14, 68). Its colonial history thus started somewhat later
than that of Singapore, which was established as a British colony in 1819.
Malay is oicially speciied as the national language in Brunei, but from 1984
a bilingual education system was implemented (Ożóg 1996). Until recently, Malay
was the medium of instruction for the irst three years of primary school, and then
English was introduced as the medium of instruction from the fourth year of primary school (Jones 2007), though just recently this has changed, and from 2009
English is stipulated as the medium of instruction for maths and science from the
start of primary school (Ministry of Education 2009: 41).
In modelling the emergence of varieties of English around the world, Schneider (2007: 160) claims that Singapore English is in the fourth phase of development of the full ive-phase cycle, currently undergoing endonormative stabilization. hough Schneider makes no mention of Brunei, we might assume that
Brunei English is in the third phase, that of nativization, partly because the language is less widely used as an inter-ethnic lingua franca than in Singapore, as
the lingua franca in Brunei is generally Brunei Malay (Martin 1996). Time will
tell whether Brunei English eventually progresses through to the fourth and ith
phases of the cycle, perhaps spurred on by its recent adoption as the medium of
instruction for maths and science from the start of primary school.
4. Acoustic evidence of rhoticity
Instrumentally, R-colouring is characterized by a low third formant (F3) (Hayward 2000: 167). Boyce and Espy-Wilson (1994) claim that R-coloured American vowels have third formants that usually fall below 2 000 Hz, but Hagiwara
(1995: 118–9) questions this by showing that the syllabic /r/ in his data has an average F3 of 1 995 Hz with a standard deviation of 347 Hz, which means that many
of the tokens have an F3 above 2 000 Hz. Although he concludes that most of the
tokens have an F3 below 2 342 Hz, he observes that specifying a ixed limit for the
F3 of an R-coloured vowel is “an inappropriate way of describing the underlying
facts” (Hagiwara 1995: 118).
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 125
Attempts will be made here to measure F3 and thereby provide instrumental
support for the perceptions of R-colouring. However, we must admit that the correlation between lowered F3 and R-colouring is only approximate, partly because
it is not always possible to derive reliable estimates of F3 even with sophisticated
acoustic sotware such as Praat (Boersma and Weenink 2009). Here, we are limiting the study to speakers of one gender, females, and we can note that a ixed
speciication for the F3 of R-coloured vowels would be even more problematical if
we included data from both men and women.
One other issue to be considered is whether R-colouring is characterised by a
dipping F3 towards the end of the vocalic portion of the syllable, or whether the
whole vowel might have a lowered F3. We may note that, for the American pronunciation of the NURSE vowel (using the lexical keywords of Wells 1982), both
Wells (2008) and Jones et al. (2003) show R-colouring for the whole vowel, so for
example heard is shown as [hɝːd]. In contrast, the American pronunciation of the
START vowel is shown with /r/ ater the vowel, so hard is represented as [hɑːrd].
his seems to suggest that, for some vowels at least, the quality of the whole vowel
may be afected. However, we should also note that one other factor in the diferent representation of words such as heard and hard is that the NURSE vowel can
only occur in a potential R-coloured environment, so in American English there is
no independent /ɜː/ phoneme, while, in contrast, /ɑː/ can occur with no following
/r/ in words in the PALM lexical set such as calm and father. It is therefore unclear whether there really is a diference in the R-colouring of NURSE and START
vowels.
5. Speakers
he speakers consist of 18 Bruneians and 12 Singaporeans, all of them ethnically Malay females. At the time of the recording, the Bruneians (labelled Brun1 to
Brun18) were aged between 20 and 23 with a mean age of 21 (SD = 0.9). All were
undergraduates at the University of Brunei Darussalam (UBD) doing an Englishmedium degree, with eleven of them training to be English teachers. he Singaporeans (labelled Sg1 to Sg12) were aged between 19 and 30 with a mean age of
24 (SD = 3.7), and they were all BA undergraduates training to be teachers at the
National Institute of Education (NIE) who had chosen English language as their
specialty. All the speakers had had more than ten years of experience learning and
speaking English, and all were proicient speakers of English.
It was decided to restrict the study to Malay females to limit the variables. In
Brunei, Malays represent the majority of the population. In contrast, they constitute just 14% of the population of Singapore, where the overwhelming majority are
© 2010. John Benjamins Publishing Company
All rights reserved
126 Salbrina Sharbawi and David Deterding
Chinese (Deterding 2007: 1) However, using data from females of the same ethnic
group facilitates comparison between the results for the two sets of speakers.
6. Data
he recordings for the Brunei speakers were made in a reasonably quiet oice at
UBD. he Singapore recordings were carried out in the Phonetics Laboratory of
NIE. In all cases, a high-quality microphone was placed a few inches from the
speakers’ mouth and the speech was digitized directly onto a computer at a sampling rate of 22 050 Hz. All the speakers were asked to read the Wolf passage as
below (but with no line numbers):
1
2
3
4
5
6
7
8
9
10
11
12
13
14
here was once a poor shepherd boy who watched his locks in the ields
next to a dark forest near the foot of a mountain. One hot aternoon, he
thought up a good plan to get some company for himself and also have a
little fun. Raising his ist in the air, he ran down to the village shouting
“Wolf, Wolf.” As soon as they heard him, the villagers all rushed from
their homes, full of concern for his safety, and two of them stayed with him
for a while. his gave the boy so much pleasure that a few days later he tried
exactly the same trick again, and once more he was successful. However,
not long ater, a wolf was looking for some change in its usual diet of
chicken and duck, so it actually did come out from the forest and began to
threaten the sheep. Racing down to the village, the boy of course cried out
even louder than before, but as the villagers were convinced that he was
trying to fool them a third time, nobody bothered to come and help him,
and so the wolf had a feast.
his passage is similar to the text proposed in Deterding (2006).1 he passage is
well suited for the description of English because it contains clear tokens of all
the vowels and consonants of English. A comprehensive analysis of the vowels
and consonants of these speakers, including their monophthongs, the FACE and
GOAT vowels, realisation of the TH-sounds, and L-vocalisation, is presented in
Salbrina (2009).
Here, analysis will just focus on tokens of non-prevocalic /r/ in the coda of
stressed syllables, as the presence or absence of R-colouring in unstressed syllables
can be hard to judge. he following tokens were selected for analysis (with the line
numbers shown in brackets):
1. he recordings were made before the text that is published in Deterding (2006) underwent
some minor changes on the suggestions of a reviewer.
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 127
–
–
word-inal: more (8), before (12)
before a consonant: dark (2), heard (5), concern (6), course (11), third (13)
hese tokens represent a range of phonological environments for non-prevocalic
/r/: word-inal in the middle of a phrase (more), word-inal before a pause (before),
and non-inal before /k/ (dark), /d/ (heard, third), /n/ (concern), and /s/ (course).
hree tokens of potential non-prevocalic /r/ in stressed syllables were excluded from the analysis: poor (line 1), near (line 2), and air (line 4). All these may
involve a centring diphthong in non-rhotic accents, /ʊə/, /ɪə/ and /eə/ respectively,
and the distinction between a centring diphthong and a monophthong followed
by /r/ is sometimes hard to determine. All the tokens investigated in this study
therefore involve a stressed syllable containing a monophthong vowel with or
without R-colouring.
One of the tokens, more (line 8) is followed by he, and if this pronoun is produced in its weak form without an initial /h/ (Roach 2009: 91), there is the possibility of a linking /r/ even in non-rhotic accents. We must therefore take care to
observe if the results include cases where more is the only token with a inal /r/.
If there are any speakers who have /r/ in more but not in any other tokens, this is
probably a linking /r/ and it should be disregarded as evidence of the rhoticity of
the speaker’s accent.
7.
Methodology
he presence or absence of non-prevocalic /r/ in the data was investigated in two
ways: perceptually and acoustically. Praat sotware (Boersma and Weenink 2009)
was used for both parts of the analysis.
For the perceptual investigation, both investigators listened to all the selected
tokens and judged whether each of the vowels is R-coloured or not. his was done
by identifying the location of the target word in the waveform and then playing the
sound repeatedly using the playback function of Praat. Overall, there was agreement on all but eleven tokens, which represents an agreement rate of nearly 95%.
he tokens where there was disagreement will be discussed in the Results section
below.
he acoustic analysis depended on measurement of the third formant. For
this, the default settings of Praat were adopted: Burg method linear prediction;
5 formants up to a maximum of 5 500 Hz; analysis window of 25 msec duration;
dynamic range of 30 dB; pre-emphasis of 50 Hz.
As mentioned above, R-colouring may be relected in dipping F3 or in overall lowering of F3. Attempts were made to measure changes in the F3 track by
© 2010. John Benjamins Publishing Company
All rights reserved
128 Salbrina Sharbawi and David Deterding
taking two measurements, one near the start of the vowel and one towards its
end. However, there are problems with accurate formant tracking of the higher
formants such as F3, even with sophisticated sotware such as Praat, and furthermore, as we have seen, some vowels may be characterised by an overall lowered
F3 rather than any dipping in its value. he results here will therefore just discuss
the absolute value of F3, even though it is acknowledged that there are limitations
to this approach.
In addition to F3, the second formant F2 was also measured. In investigating
the vowels of Brunei English, Salbrina (2006) showed that a pre-vocalic /r/ can affect the F2 in words such as traveller and wrapped, and this suggests that the value
of F2 may be an important indicator of R-colouring. Furthermore, measurement
of both F2 and F3 allows us to show the results on two-dimensional plots.
8. Results
Table 1 shows the perceptual results for the irst author for the Brunei data. he
italicised tokens are the ones where there was perceptual disagreement between
the two authors: four tokens for Brun9, two for Brun10, and one each for Brun8
and Brun13. All but one of these (before for Brun9) involve the irst author hearing
R-colouring while the second author perceived no R-colouring. Despite these differences, the overall results for the two listeners are quite similar: the irst author
judged nearly 47% of the tokens to be R-coloured, while the second author found
about 42%.
hree Brunei speakers (Brun2, Brun10 and Brun12) have /r/ in all seven tokens, and six speakers (Brun1, Brun7, Brun14, Brun15, Brun16, and Brun18) have
/r/ in none of the tokens. If we judge a speaker to be rhotic when at least four of
the seven tokens have non-prevocalic /r/, then half of the Bruneians (9 out of 18,
or 50%) can be classiied as rhotic, a result which both authors agree with.
Table 2 shows the results for the irst author for the Singapore data, with the
three tokens where there was disagreement between the authors shown in italics (one each for Sg1, Sg2 and Sg7). All these involve the irst author hearing Rcolouring while the second did not.
Using the results of the irst author, just over 8% of the Singapore English
tokens are R-coloured, and only one of the speakers might be classiied as rhotic.
he second author reported fewer instances of R-colouring and no speakers were
classiied as rhotic, though three out of seven of the tokens of Sg2 were judged to
have /r/. Clearly the overall instance of R-colouring is much lower for the Singapore data.
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 129
Table 1. R-colouring in the Brunei data
dark
heard
concern
more
course
before
third
Total
Brun1
0
0
0
0
0
0
0
0
Brun2
1
1
1
1
1
1
1
7
Brun3
0
1
0
1
0
1
1
4
Brun4
1
1
0
1
1
1
1
6
Brun5
1
1
1
1
0
1
1
6
Brun6
1
0
0
0
0
0
0
1
Brun7
0
0
0
0
0
0
0
0
Brun8
1
1
0
1
1
1
1
6
Brun9
1
0
1
1
0
0
0
3
Brun10
1
1
1
1
1
1
1
7
Brun11
1
1
0
1
0
1
1
5
Brun12
1
1
1
1
1
1
1
7
Brun13
1
1
1
0
1
1
1
6
Brun14
0
0
0
0
0
0
0
0
Brun15
0
0
0
0
0
0
0
0
Brun16
0
0
0
0
0
0
0
0
Brun17
0
0
0
0
0
1
0
1
Brun18
0
0
0
0
0
0
0
0
Total
59 (46.8%)
Table 3 shows the average F3 values and the standard deviation for the Brunei
speakers, who are classiied as rhotic or non-rhotic depending on whether four or
more of their tokens are R-coloured.
A t-test conirms that the average F3 for the rhotic Brunei speakers is signiicantly lower than that of the non-rhotic ones (t = 5.71, df = 16, independent samples, two-tailed, p < 0.001). his conirms a correlation between F3 and rhoticity,
though we should note that one rhotic speaker (Brun3) has a value that is closer
to that of the non-rhotic speakers, presumably because only four out of seven of
her tokens are actually R-coloured. Furthermore the value for Brun17 is actually
lower than that of Brun3.
Table 4 shows the acoustic measurements for the Singaporeans, with the rhotic / non-rhotic classiication based on the perceptions of the irst author.
We should note that the value for the one rhotic Singaporean (Sg2) is almost
identical to the average for the non-rhotic Singaporeans, so in this case the measurement of F3 has not separated out the two categories, presumably because, just
© 2010. John Benjamins Publishing Company
All rights reserved
130 Salbrina Sharbawi and David Deterding
Table 2. Rhoticity in the Singapore data
dark
heard
concern
more
course
before
third
Total
Sg1
0
0
0
0
0
0
1
1
Sg2
0
1
1
1
0
1
0
4
Sg3
0
0
0
0
0
0
0
0
Sg4
0
0
0
0
0
0
0
0
Sg5
0
0
0
0
0
0
0
0
Sg6
0
0
0
0
0
0
0
0
Sg7
1
1
0
0
0
0
0
2
Sg8
0
0
0
0
0
0
0
0
Sg9
0
0
0
0
0
0
0
0
Sg10
0
0
0
0
0
0
0
0
Sg11
0
0
0
0
0
0
0
0
Sg12
0
0
0
0
0
0
0
Total
0
7 (8.3%)
Table 3. Average F3 and standard deviations for Brunei speakers
Non-rhotic
Rhotic
F3 (Hz)
SD
F3 (Hz)
SD
Brun1
3081
x(214)
Brun2
2308
x(189)
Brun6
2867
x(329)
Brun3
2811
x(356)
Brun7
3029
x(288)
Brun4
2711
x(467)
Brun9
2891
x(418)
Brun5
2696
x(283)
Brun14
3292
x(216)
Brun8
2451
x(359)
Brun15
3099
x(180)
Brun10
2344
x(233)
Brun16
2984
x(259)
Brun11
2445
x(319)
Brun17
2718
x(444)
Brun12
2397
x(198)
Brun18
3010
x(148)
Brun13
2629
x(249)
Average
2997
x(277)
Average
2532
x(295)
as with Brun3, only four out of seven of the tokens for Sg2 were judged to be Rcoloured. In fact, as mentioned above, the second author only heard R-colouring
for three of these tokens, in which case Sg2 would not be classiied as rhotic.
In order to investigate whether individual R-coloured tokens tend to have
lower F3 than non-R-coloured ones, the formant measurements of the individual
tokens can be shown on two-dimensional scatter plots of F2 versus F3. For these
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English
Table 4. Average F3 and standard deviations for Singapore speakers
Non-rhotic
Rhotic
F3 (Hz)
SD
Sg1
3083
x(719)
Sg3
3089
x(169)
Sg4
2871
x(130)
Sg5
3007
x(315)
Sg6
2916
x(215)
Sg7
2586
x(303)
Sg8
2800
x(245)
Sg9
2669
x(202)
Sg10
2888
x(47)
Sg11
3003
x(215)
Sg12
2846
x(299)
Average
2887
x(260)
F3 (Hz)
SD
Sg2
2870
x(474)
Average
2870
x(474)
vowel plots, the Hertz measurements are irst converted to an auditory Bark scale
using the formula suggested by Zwicker and Terhardt (1980) to allow a visual representation of the quality of the vowels on a scale that is similar to the way that the
human ear perceives them (Hayward 2000: 142). In all the plots, following normal
practice, the F2 axis is inverted so that vowels with a fronted value are shown on
the let (Ladefoged 2006: 188). However, the F3 axis is not inverted, so tokens with
a lowered F3 are shown as lower on the plot.
Figure 1 shows the scatter plot for all 54 tokens of the Brunei data with the
NURSE vowel (heard, concern, third), with the white data points being the ones
where the irst author heard R-colouring. It can be seen that F3 separates out the
vowels quite efectively, with just one R-coloured token having a high F3 so intruding into the space occupied by the tokens with no R-colouring. his token is
concern for Brun5, which both authors heard as R-coloured. Further investigation
shows that Brun5 produced this word with the stress on the irst syllable, and measurement of F3 to indicate R-colouring for an unstressed second syllable may not
be comparable to the other tokens of NURSE.
Figure 2 shows the scatter plot for all 54 tokens of the Brunei data with the
FORCE vowel (course, more, before). Once again it can be seen that F3 separates
out the two categories quite efectively, with just one R-coloured token towards the
let of the plot having a high F3. his is before for Brun4, which both authors heard
as R-coloured. In fact the F3 is hard to track in this token, which illustrates that
formant-tracking sotware cannot always provide deinitive answers, especially for
© 2010. John Benjamins Publishing Company
All rights reserved
131
132 Salbrina Sharbawi and David Deterding
17
15
F3 (Bark)
16
Not R-coloured
R-coloured
14
13
15
14
13
12
11
10
9
8
7
F2 (Bark)
Figure 1. Scatter plot of all Brunei tokens of the NURSE vowel
17
Not R-coloured
R-coloured
15
F3 (Bark)
16
14
13
15
14
13
12
11
10
9
8
7
F2 (Bark)
Figure 2. Scatter plot of all Brunei tokens of the FORCE vowel
the higher formants. One other R-coloured token that intrudes a little into the
space for the tokens with no R-colouring is course for Brun8, even though once
again both authors heard it as R-coloured.
Figure 3 shows the scatter plot of the 18 tokens of the Brunei data of dark.
here are two R-coloured tokens with a high F3 so they intrude into the space of
the non-R-coloured tokens. hese are the tokens for Brun4 and Brun13, tokens
which both authors heard as having R-colouring. Alternatively, one might say that
there are two non-R-coloured tokens with a relatively low F3. hese are for Brun1
and Brun17, both of which were perceived by both authors to have no R-colouring.
In conclusion, the formant plots work quite well in conirming the auditory judgments, and the measurements of F3 can separate out the R-colouring of
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 133
17
15
F3 (Bark)
16
Not R-coloured
R-coloured
14
13
15
14
13
12
11
10
9
8
7
F2 (Bark)
Figure 3. Scatter plot of all Brunei tokens of dark
tokens reasonably successfully, though there are a few anomalous cases where the
formant measurements are not quite so reliable.
9. Discussion
Brunei English is clearly much more rhotic than Singapore English, with about
50% of speakers being classiied as rhotic. his seems rather surprising as it does
not tally with the expectation that English is non-rhotic in countries in which the
language arrived from Britain ater the start of the 19th century. Could the rhoticity of Brunei English be the result of inluence from American media? One issue
is that there is no evidence that Bruneians watch more American TV or listen to
more American music than Singaporeans.
In order to see whether there is a general tendency to follow American English
pronunciation in Brunei, three other features were considered: the vowel in ater,
as American English has /æ/ rather than /ɑː/ (Wells 1982: 133–5); the vowel in hot
for which American English usually has /ɑː/ rather than /ɒ/ (Wells 1982: 130–1);
and the sound in the middle of later which tends to be a tap in American English
(Wells 1982: 248). It was found that only one Bruneian (Brun13) has [æ] in ater;
none has [ɑː] in hot; and only two (Brun4 and Brun13) have a tap in later. Overall,
therefore, there seems little evidence of widespread adoption of an American accent in Brunei, though it is true that one or two speakers may have it.
here seems to be inconsistency in the usage of the American English features and rhoticity in the speech of the Bruneians. For example, with the exception of Brun4 and Brun13, those who have taps in little and later do not have
© 2010. John Benjamins Publishing Company
All rights reserved
134 Salbrina Sharbawi and David Deterding
non-prevocalic /r/ in their speech and those who are apparently rhotic do not
produce hot with American English [ɑː]. Despite these indings, we cannot rule
out the American media as one of the inluences on Brunei English. In fact, we
might consider how a range of inluences might combine to afect features of an
emergent variety of English, and so we should think about the pronunciation of
the indigenous languages of Brunei and Singapore.
All the Bruneian speakers claim that Brunei Malay is their mother tongue,
while the Malays in Singapore speak Standard Malay, and a notable diference
between Brunei Malay and the Standard Malay spoken in Singapore is in the realization of /r/ in non-prevocalic positions. Brunei Malay is a rhotic variety of
Malay, with the /r/ sound oten pronounced as a trill (Poedjosoedarmo 1996; Clynes 2001). In contrast, the Standard Malay of Singapore and most of Malaysia
generally has no non-prevocalic /r/, and the deletion of /r/ in word-inal position
is compensated for by the lengthening of the preceding vowel in words such as
tukar ‘change’ (Teoh 1996: 47).
It is likely that the rhoticity of Brunei Malay has a bearing on the occurrence of
rhoticity in Brunei English. his is not to say that it is the only contributing factor,
as inluence from the American media may also play a part in this phenomenon.
In other words, rhoticity in Brunei English may be a consequence of two sources:
Brunei Malay and American English, and these two inluences conspire to result
in the feature being adopted into Brunei English.
In fact, it may be naïve to look for a single source for a phonetic or grammatical feature in an emergent variety of English. his view is shared by Poedjosoedarmo (2000b) who claims that a combination of sources gives rise to many
features in the syntax of the written English of Singaporeans, so for example the
occurrence of null-subject structures may have arisen from the inluence of both
Malay and Chinese.
Finally, we might consider the status of Brunei English in terms of establishing
its own norms of pronunciation. We earlier suggested that Brunei English may be
in phase 3 of Schneider’s (2007) 5-phase model of the development of postcolonial
Englishes. If this is correct, it would mean that Brunei English is at an earlier stage
than Singapore English, which is already in phase 4. As a result, Brunei English
may be less mature as an independent variety and more susceptible to external
inluences than Singapore English.
It is interesting to note that Hong Kong English also seems to exhibit substantial inluences from American English, with Deterding, Wong and Kirkpatrick
(2008) reporting that six out of 15 of the speakers in their study had clear inluences from American English, including R-colouring of their vowels. We might
further note that, just like Brunei but unlike Singapore, English is not the lingua
franca of Hong Kong. Schneider (2007: 135) places Hong Kong in phase 3 of the
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English
cycle, and in general we expect varieties that are in phase 3 to be more susceptible
to outside inluences than those that are undergoing endonormative stabilization
in phase 4.
10. Conclusion
his paper has investigated the occurrence of rhoticity in Brunei English and Singapore English. It was found that the Bruneians displayed a higher tendency to
realize /r/ in non-prevocalic positions and this inding was supported both auditorily and acoustically. For the Singaporeans, however, only one speaker was perceptually judged to have a rhotic English accent and even then, only about half of
her tokens are R-coloured.
he inding does not support the expectation that Brunei English is non-rhotic
just like its counterparts in Singapore and Malaysia. At irst sight, this is surprising
as factors that determine whether an English variety is rhotic seem to be similar
in the three nations. Upon closer inspection, however, it was concluded that the
widespread nature of rhoticity in Brunei English is partly because Brunei Malay
is also rhotic, unlike the Malay spoken in Singapore and most of Malaysia, and
this combines with the inluence from American media in the country to result in
widespread rhoticity in Brunei English.
References
Baskaran, Loga. 2004. “Malaysian English: Phonology”. In Edgar W. Schneider, Kate Burridge,
Bernd Kortmann, Rajend Mesthrie and Clive Upton, eds. A Handbook of Varieties of English. Vol. 1: Phonology. Berlin: Mouton de Gruyter, 1 034–46.
Boersma, Paul and David Weenink. 2009. “Praat: Doing phonetics by computer”. http://www.
fon.hum.uva.nl/praat/ (accessed 18 Nov. 2009).
Boyce, Suzanne and Carol Y. Espy-Wilson. 1994. “Acoustic diferences between ‘bunched’ and
‘retrolex’ variants of American /r/”. Journal of the Acoustical Society of America 95: 2 823.
Clynes, Adrian. 2001. “Brunei Malay: An overview”. Occasional Papers in Language Studies, Department of English Language and Applied Linguistics, UBO 7: 11–44.
Crystal, David. 2003. A Dictionary of Linguistics and Phonetics. 5th ed. Malden, Oxford:
Blackwell.
Deterding, David. 2006. “he North Wind versus a Wolf: Short texts for the description and
measurement of English pronunciation”. Journal of the International Phonetic Association
36: 187–96.
———. 2007. Singapore English. Edinburgh: Edinburgh University Press.
———, Jennie Wong and Andy Kirkpatrick. 2008. “he pronunciation of Hong Kong English”.
English World-Wide 29: 148–75.
© 2010. John Benjamins Publishing Company
All rights reserved
135
136 Salbrina Sharbawi and David Deterding
Eu Lay Kim, Veronica. 2004. “he potential inluence of American English on some pronunciation features of young educated Singaporeans: An acoustic and attitudinal exploration”.
Unpublished English language honours academic exercise, NIE, Singapore.
Hagiwara, Robert. 1995. “Acoustic realization of American /r/ as produced by men and women”.
UCLA Working Papers in Phonetics 90: 2–187.
Hay, Jennifer, Margaret Maclagan and Elizabeth Gordon. 2008. New Zealand English. Edinburgh: Edinburgh University Press.
Hayward, Katrina. 2000. Experimental Phonetics. Harlow: Longman.
Hickey, Raymond. 2004. “South-East Asian Englishes”. In Raymond Hickey, ed. Legacies of Colonial English: Studies in Transported Dialects. Cambridge: Cambridge University Press,
536–58.
Hussainmiya, Bachamiya Abdul. 2006. Brunei Revival of 1906: A Popular History. Bandar Seri
Begawan: Brunei Press.
Jenkins, Jennifer. 2009. World Englishes: A Resource Book for Students. 2nd ed. London, New
York: Routledge.
Jones, Daniel, Peter Roach, James Hartman and Jane Setter. 2003. Cambridge English Pronouncing Dictionary. 16th ed. Cambridge: Cambridge University Press.
Jones, Gary M. 2007. “20 years of bilingual education: hen and now”. In David Prescott, ed.
English in Southeast Asia: Varieties, Literacies and Literatures. Newcastle, UK: Cambridge
Scholars Publishing, 246–58.
Labov, William. 1966. he Sociolinguistic Stratiication of English in New York City. Washington,
D.C.: Center for Applied Linguistics.
Ladefoged, Peter. 2006. A Course in Phonetics. 5th ed. Boston, MA: homson Wadsworth.
Low, Ee Ling and Adam Brown. 2005. English in Singapore: An Introduction. Singapore:
McGraw-Hill.
Martin, Peter W. 1996. “Brunei Malay and Bahasa Melayu: A sociolinguistic perspective”. In
Martin, Ożóg and Poedjosoedarmo, eds. 1996: 27–36.
———, Conrad Ożóg and Gloria Poedjosoedarmo, eds. 1996. Language Use and Language
Change in Brunei Darussalam. Athens, Ohio: Ohio University Center for International
Studies.
Ministry of Education. 2009. he National Education System for the 21st Century. Bandar Seri
Begawan: Ministry of Education.
Mossop, Jonathan. 1996. “Some phonological features of Brunei English”. In Martin, Ożóg and
Poedjosoedarmo, eds. 1996: 189–208.
Mugglestone, Lynda. 2003. Talking Proper: he Rise of Accent as Social Symbol. Oxford: Oxford
University Press.
Ożóg, Conrad. 1996. “he unplanned use of English: he case of Brunei Darussalam”. In Martin,
Ożóg and Poedjosoedarmo, eds. 1996: 156–72.
Poedjosoedarmo, Gloria. 1996. “Variation and change in the sound systems of Brunei dialects of
Malay”. In Martin, Ożóg and Poedjosoedarmo, eds. 1996: 37–42.
———. 2000a. “he media as a model and source of innovation in the development of Singapore
Standard English”. In Adam Brown, David Deterding and Low Ee Ling, eds. he English
Language in Singapore: Research on Pronunciation. Singapore: Singapore Association for
Applied Linguistics, 112–20.
———. 2000b. “Inluences of Malay on the written English of university students in Singapore”.
In Adam Brown, ed. English in Southeast Asia: Proceedings of the Fourth ‘English in Southeast Asia’ Conference. Singapore: National Institute of Education, 210–9.
© 2010. John Benjamins Publishing Company
All rights reserved
Rhoticity in Brunei English 137
Rajadurai, Joanne. 2006. “Pronunciation issues in non-native contexts: A Malaysian case study”.
Malaysian Journal of ELT Research 2: 42–59. http://www.melta.org.my/modules/tinycontent/Dos/RJoanne.pdf (accessed 8 Mar. 2007).
Ramasamy, Sheila Adelina. 2005. “Analysis of the usage of post vocalic /r/ in Malaysian English”.
Unpublished M.A. thesis, Universiti Malaya, Malaysia.
Roach, Peter. 2009. English Phonetics and Phonology. 4th ed. Cambridge: Cambridge University
Press.
Salbrina, H. Sharbawi. 2006. “he vowels of Brunei English: An acoustic investigation”. English
World-Wide 27: 247–64.
———. 2009. “he vowels and consonants of Brunei English”. Ph.D. dissertation, National Institute of Education, Singapore.
Schneider, Edgar W. 2007. Postcolonial English. Cambridge: Cambridge University Press.
Tan, Chor Hiang and Anthea Fraser Gupta. 1992. “Post-vocalic /r/ in Singapore English”. York
Papers in Linguistics 16: 139–52.
Teoh, Boon Seng. 1996. he Sound System of Malay Revisited. Kuala Lumpur: Dewan Bahasa
dan Pustaka.
Trudgill, Peter and Jean Hannah. 2008. International English: A Guide to the Varieties of Standard English. 5th ed. London: Hodder Education.
Wells, John C. 1982. Accents of English. 3 vols. Cambridge: Cambridge University Press.
———. 2008. Longman Pronunciation Dictionary. 3rd ed. Harlow: Pearson Education.
Zwicker, Eberhard and Ernst Terhardt. 1980. “Analytical expression for critical-band rate and
critical bandwidth as a function of frequency”. Journal of the Acoustical Society of America
68: 1 523–5.
Authors’ addresses
David Deterding
Universiti Brunei Darussalam
FASS
Jalan Tungku Link
Gadong, 1410
Brunei Darussalam
Salbrina Sharbawi
Universiti Brunei Darussalam
FASS
Jalan Tungku Link
Gadong, 1410
Brunei Darussalam
dhdeter@gmail.com
salbrina.sharbawi@ubd.edu.bn
© 2010. John Benjamins Publishing Company
All rights reserved