Beruflich Dokumente
Kultur Dokumente
Toni Lam
To cite this article: Toni Lam (2017) Intonational variation in Hong Kong English: a pilot study,
Asian Englishes, 19:1, 22-43, DOI: 10.1080/13488678.2016.1277411
Download by: [Hong Kong Polytechnic University] Date: 25 May 2017, At: 21:50
Asian Englishes, 2017
VOL. 19, NO. 1, 2243
http://dx.doi.org/10.1080/13488678.2016.1277411
Introduction
The problem with Hong Kong English from varieties-based perspectives
Ever since Luke and Richards (1982) established in their seminal article English in Hong
Kong: functions and status that the variety of English used in Hong Kong had developed
its own characteristics, the phonology of HKE and its distinct features of accent have been
identified and documented (e.g. Bolton & Kwok, 1990; Deterding, Wong, & Kirkpatrick,
2008; Hung, 2002). However, a weakness of this variety-based approach is that it tends to
focus on typical features at the expense of variation (Sewell, 2012). For instance, based on
Lukes (2000) seminal study, Gussenhoven (2012) derived the typical phonological features
of sentence intonation of HKE, without considering the issue of variation. While a certain
degree of generalisation is important for theorising and compiling features that are rep-
resentative of a particular variety, features enumerated in a hypothetical system may vary
according to other social factors including the age, proficiency level and speech style of the
speaker. Therefore, a variationist model is a necessary complement to the varieties-based
approach to accent studies (Sewell, 2012, p. 3).
global language in Hong Kong, Asias World City, thus provides an ideal context for stud-
ying the local variety along with its variation. Sewell (2016) has successfully employed the
global English approach to explore HKE accent and pronunciation to show how global and
local influences interact in this era of globalisation. In order to better capture the global and
local dimensions and avoid the misconceptions associated with the old labels, HKed and
HKbr should be renamed as HKE-gl (HKE with global orientations) and HKE-lo (HKE
with local orientations) correspondingly. The concepts pertaining to the sociolinguistics
of globalisation include scales, indexicality, polycentricity, ideology and commodification.
Scalar theory. It theorises spatial, social and temporal elements of communication. Space
can be represented both horizontally and vertically. The vertical dimension concerns how
different languages are perceived hierarchically as good, bad, better and worse, focusing
on issues of power and inequality. In the case of Hong Kong, for instance, HKE-gl is often
viewed as better than HKE-lo. Horizontally, acts of communication often have shifting
and multiple references to local and translocal/global dimensions as well as elements of
time. Language users may employ different norms, including accents, at the same time or
on different occasions, or mix features of different languages to express their identities or
stances. As a result of globalisation, many Hong Kong individuals, though born and brought
up locally, may have spent time overseas or at least had exposure to English-speaking cul-
tures through the media and/or the Internet, which would significantly diversify ones life
experiences and (socio)linguistic knowledge.
Indexicality. It points out from the text to the physical world and the social world (Collins
et al., 2009, as cited in Sewell, 2016, p. 21), and thus is pertinent to identity. A linguistic
aspect germane to identity is accent. Sung (2014) showed that some HK bilingual speakers
favoured English with a local accent, attributing their preference to their wish to signal their
Hong Kong identity, which was confirmed by Hansen Edwards (2015) study, concluding
that HKE has been increasingly regarded as a valued variety, having a symbolic value in
the construction and marking of its ethnolinguistic identity, and is used as an in-group
identity marker. The relationship between accent and identity, however, is a complex one.
Polycentricity. The concept can be used to account for this complex relationship. It can
take the form of multiple perceived centres, ranging from actual centres such as individuals
(parents, teachers and other cultural role models) and collectives (friends, schools, work-
places) to imagined ones such as ideals (equality, freedom). Speakers may orient towards
different centres of authority. For instance, in the same study, Sung (2014) also found that
other participants adopted a native or near-native accent to construct what they deemed
to be a desirable identity in English (p. 554) in order to present a positive self-image as
a competent L2 speaker and/or learner of English (p. 553), attempting to create a unique
local identity, comprising the ability to speak English with a standard prestige accent from
the West (Luk & Lin, 2006; quoted in Sung, 2014). Speakers of English in HK can then be
described, borrowing Dong and Blommaerts (2009, cited in Sewell, 2016) wording, to pro-
ject multilayered identities (p. 59), which simultaneously reveals their polyglot repertoire
(p. 58). This kind of multiple indexing, according to Sewell (p. 22), is inherently ideological,
relying on sociohistorically constructed meanings and group-based power relations.
Ideology. There is a strong link between language, indexicality and ideology. Sewell (2016,
p. 26) interprets ideology as an intrinsic aspect of the structure of language, its perceived
meanings, and its role in the construction of communities and identities. Indexicality can
thus be seen as using ideological structures to form a set of interactional norms for particular
ASIAN ENGLISHES 25
social groups to create meanings specific to their unique identities. Some of these ideological
structures can be found in accent, including intonation. Just as standard native accents are
often associated with formal, authoritative, and educated, e.g. HKE-gl, non-native accents
have the tendency to be informal, colloquial, friendly and less educated, e.g. HKE-lo.
Accent accommodation. Accents that are globally recognised as good or valued are
often commercially treated as products or commodities, e.g. in accent reduction schools,
hence the phenomenon of accent commodification. Speaking English with the right accent,
Sewell (2016, p. 29) explained, can become a personal attribute that increases perceived
competitiveness in certain marketplaces, taking advantage of the insecurity and aspiration
of the upwardly mobile.
Under the influence of globalisation, traditional boundaries such as language and vari-
ety have been blurred. Hybrid identities with multiple referencing can be captured by the
global English approach. A model aimed to depict the globalised sociolinguistic landscape
of Singapore is the Cultural Orientation Model (COM) by Alsagoff (2007, 2010).
Cultural Orientation Model. COM aims to elucidate intra-speaker variation in terms
of style-shifting through negotiating within the multidimensional space between the two
macro-cultural orientations of globalism and localism. The variation in the use of English
in Singapore stems from a cultural tension between being/doing global and being/doing
local. The former arises from a need for international intelligibility, motivated by economic
capital with a global perspective to use English as an instrument and means of global par-
ticipation in not only financial and economic markets, but also in science and technology,
while the latter involves the desire for national identity, a sense of a Singaporean identity
that is inextricably bound with Singlish, representative of their culture and identity as
Singaporeans. Table 1 depicts the details of COM.
Despite having a rather different sociolinguistic profile from Singapore, intra-speaker
variation in HKE speakers has been confirmed empirically by Sewell (2012). There is, how-
ever, one major flaw with COM: it excludes code-switching. HKE often includes frequent
instances of borrowing and code-switching. While exclusively concerned with variation
strictly within English, COM does not reflect the full picture of HKE.
Eckerts Indexicality (Eckert, 2008; Silverstein, 2003). It incorporates the issue of
code-switching, and involves various orders of indexicality, with the first order being
pragmatic, involving indexical presupposition of macro-social order (Silverstein, 2003, p.
227), the second metapragmatic and social, concerning appropriateness-to-context and
effectiveness-to-context (p. 196), and a higher order being conventionally recognised
(Leimgruber, 2012, p. 8) indexicality, involving a continual interpretation of forms in con-
text, and in-the-moment assigning of indexical values to linguistic forms (Eckert, 2008,
p. 463).
No visitors can leave Hong Kong without first going for some dim sum lor
In this example, for instance, a suggestion is made at the first indexical level by the use
of the sentence-final particle (SFP) lo (or lor) in order to make it obvious. A certain level of
informality is indexed by its presence at the second level. At a higher-order indexical level,
the SFP is used to index a combination of light-heartedness, localism, etc. SFPs are consid-
ered borrowed from Cantonese and often treated as code-switching, and thus should be
included for investigating the intonational variation in HKE. However, many HKE speakers
are L2 users. Despite some having high levels of proficiency, their pronunciation skills do
vary.
L2 Intonation Aptitude1. Learning the intonation of a different language is no easy task
as it involves a specific pronunciation aptitude, including awareness, analysis skills and
productive skills. Intonation awareness is defined by Ramirez Verdugo (2006, p. 142) as
the knowledge about the form, function and meaning of intonation in spoken discourse
that language users may have. However, L1 intonation becomes intrinsic and automatic as
L1 competence increases. The knowledge and use of intonation often become implicit and
unconscious, known as the unconscious character of intonation (p. 141). Over time this
will result in a discrepancy in abilities, with excellent knowledge of grammar and vocabulary
yet less comparable pronunciation aptitude and proficiency, referred to by Reiterer et al.
(2011) as the Joseph Conrad Phenomenon. Speakers would need to be aware of such L2
prosodic features and their differences from their L1 counterparts so as to have a chance of
grasping L2 intonation. However, awareness alone is insufficient. There are specific phonetic
skills involved in hearing pitch movements, classifying them as a particular tone type, and
mapping the pitch patterns correctly to the spoken texts (Dankovicova, House, Crooks, &
Jones, 2007). Such analysis skills are so unique that even if individuals have an aptitude
for segmental phonetics, that does not guarantee competence in intonation analysis. In
fact, only the gifted few are endowed with such a specific type of pronunciation aptitude.
Intonation aptitude is concerned with not just receptive analysis skills, but also produc-
tive. Anufryk (2009) confirmed a strong correlation between the ability to vary prosodic
patterns and pronunciation aptitude, such that a greater degree of variability in prosody
would correlate with a higher degree of pronunciation aptitude, establishing intonational
variation as a signature building block of general pronunciation aptitude. Others include
lexical units and grammatical patterns. L2 speakers who lack intonation awareness, analysis
and variation skills will tend to impose their L1 intonation on L2, even those with high
levels of L2 proficiency.
A unified approach, drawing upon multiple sources Blommaerts (2010) global English
approach, L2 intonation aptitude, Eckerts (2008) indexicality, Alsagoff s (2007, 2010) COM,
as shown in Table 2 seems apt for analysing the results of the present study.
Literature review
Intonation of native and non-native Englishes
Varieties of English differ prosodically. Dialectal differences in intonation in the British Isles
are well-attested (e.g. Cruttenden, 2001; Warren & Britain, 2000; Wells, 1982). Intonational
variation in English (IViE) in the British Isles has been researched extensively (e.g. Grabe
2004). Cross-linguistic differences in prosodic structure exist also in non-native varieties
ASIAN ENGLISHES 27
Table 2.The dynamics of Hong Kong English accent in the unified approach.
Hong Kong English with global orientations Hong Kong English with local orientations
Standard Colloquial Standard Colloquial
Polycentricity, including global and local identities
Global experiences Global experiences Local experiences Local experiences
High L2 intonation High L2 intonation aptitude Low L2 intonation aptitude High/low intonation aptitude
aptitude
Economic capital Sociocultural capital Economic capital Sociocultural capital
Authority Camaraderie Authority Camaraderie
Formality Informality Formality Informality
Distance Closeness Distance Closeness
Educational attainment Community membership Educational attainment Community membership
of English, including Asian Englishes (Low & Hashim, 2012) and other New Englishes
(Schneider & Kortmann, 2004).
HKE intonation
The segmental characteristics of HKE have traditionally received more academic attention
than its suprasegmental counterparts. I include some exceptions here.
Luke (2000) treated the sentence phonology of the variety of English spoken in Hong
Kong as being made up of the tone patterns of the individual words, which is based on
the Cantonese tone system. There are altogether six tones in Cantonese, including High,
Mid, Low, High Rise, Low Rise and Low Fall (Matthews & Yip, 2011). Only a subset of
these, however, are used in HKE, namely, High, Mid and Low. Luke described HKEs word
prosody in terms of the assignment of pitch accents to the syllables with primary stress
and those secondary-stressed before the primary stress. H tones are assigned to stressed
syllables while L and M tones to unstressed (Luke, 2000, 2008), which will in turn be stored
as phonological representations of the members of the new speech community. HKE (or
HKE-lo) intonation is then formed by assigning a sequence of Cantonese lexical tones to
represent the most common pitch contours of the English syllables for word stress to yield
a basic template of MHL!, and by applying a computation for compound/linking words,
phrases and sentences to give a choppy, as opposed to smooth, intonation contour for HKE
(Luke, 2008); for instance, the utterance I saw the manager this morning would have the
pattern of MHMHHHHHL!, as in Table 3.
An important point to note here is that a distinct characteristic of HKE intonation is the
fact that every single syllable in an utterance receives a pitch accent, borrowed from three
of the six Cantonese tones, i.e. H, M and L. To contrast that with a native pattern of the
exact same utterance, as shown in Table 4 below, only one syllable receives a pitch accent
in this case.
Lukes (2000, 2008) tonal analysis of HKE intonation has subsequently been followed up
by other scholars, with some additions and modifications. Adopting an ecological approach,
28 T. LAM
Lim (2009) also analysed HKE through the ecology of Hong Kong, where Cantonese is
clearly more dominant than English and Mandarin in both internal and external ecologies,
with major influences from Cantonese tones and SFPs (James, 2001). Since the SFPs contain
Cantonese tones, they introduce very specific pitch levels or contours into the prosodic
system of HKE. Gussenhoven (2012, 2014) modified Lukes (2000, 2008) tone assignment
algorithm to yield the intonational rules of HKE:
The intonational rules of HKE by Gussenhoven (2014) based on Luke (2000, 2008)
1.Pitch accent: H
2.Insert L after last H in the IP
3.Place M on empty word-initial syllable
4.Copy tones right
His analysis yielded three HKE intonations, namely declarative, emphatic and interrogative.
While the first is characterised by having no boundary tone, the last two are expressed
through the boundary tones L% and H% respectively.
All the studies mentioned previously, however, adopted the varieties-based approach,
which ignore intonational variation in HKE, effectively neglecting the sociolinguistic func-
tion of intonation to affirm local identity through the HKE-lo accent.
expectations and signalling noteworthiness (Luke, 1990), as well as indexing identity. Failing
to include SFPs will leave out important information about HKE intonation.
As discussed throughout the previous sections, there is an obvious gap in the research
area of HKE intonation. First, all the previous studies on HKE intonation adopted the
varieties-based perspective, failing to examine intonational variation in HKE. Some recent
exceptions include Sewell and Chan (2010) and Sewell (2012), which, however, focused only
on consonants. To date, and to the knowledge of the writer, there have been no studies inves-
tigating intonational variation in HKE. Second, none of the previous HKE prosody studies
considered the sociolinguistic function of intonation of HKE in the era of globalisation,
where the speaker, enabled by his/her advanced L2 intonation aptitude or relevant exposure,
projects multiple identities through socially and ideologically constructed and commodified
sociolinguistic styles and accents through intonational features, including SFPs. The next
section will outline the aims and scope of the present study in order to fill this gap.
The experiment
The subjects for the study consisted of six students of the Hong Kong University of Science
and Technology (HKUST), belonging to two ethnic groups: three local students (J, T and
K) and three Indians. All three HKE speakers were born and bred locally, and attended
local secondary schools. J obtained a 5*, K 5** (the highest level) and T 5 in the English
Language subject at the Hong Kong Diploma of Secondary Education (HKDSE), the public
examination at the end of the local secondary education. According to the official document
Level Descriptors published by the Hong Kong Examinations and Assessment Authority
(2014), speakers at this level typically express a wide range of ideas fluently in clear, accurate,
well-pronounced English (p. 1). The non-local subjects were there as non-HKE speakers
to interact with the local counterparts. The general information about the key informants
is as in Table 5.
There were altogether eight tasks (A-H), involving different patterns of interaction for
various purposes, as illustrated in Table 6.
The tasks were designed to incorporate various approaches to stylistic variation, namely,
Attention to Speech, Audience Design and Speaker Design, as categorised by Schilling-
Estes (2003). The Attention to Speech approach focuses on the amount of attention paid to
speech, so that the more the attention paid, the more careful the speech. Factors affecting the
amount of attention paid to speech include topics, task types and, in the case of HKE, the
mandatory inclusion of SFPs. Topics related to university studies (H) would be considered
more formal and thus evoke more careful speech whereas topics related to local Hong Kong
culture (D) should be deemed less formal and hence should encourage the use of more casual
speech. Similarly, reading tasks (A, B, E and F) should invite more attention than spontane-
ous tasks, resulting in more careful speech in the former and more casual speech in the latter
(C, D, G, and H). The presence of SFPs in A and B, being one of the hallmarks of natural
conversation both in Cantonese (Luke, 1990, p. 11) and HKE (James, 2001), is believed
to promote the use of casual speech. The Audience Design approach emphasises the role
of the addressee, maintaining that a HKE speaker might adjust his/her speech depending
on his/her interlocutor, whose nationality, variety of English and accent, might be at work,
hence the arrangement of HKE-HKE, HKE-nonHKE, nonHKE-nonHKE groups in tasks
C and G. The non-HKE group in this case was Indian. The third approach is considered to
involve active creation and re-creation of speaker identity. Speaker-internal factors such as
purpose and identity are at work, stressing speaker agency. The unified model adopted for
the present study attempts to be eclectic and reap the benefits from all three approaches.
The procedure of the experiment can be summarised as ABABCDDEFEFGHH, as shown
in Figure 1A, with the HKE-lo tasks taking place first in order of decreasing control and
ASIAN ENGLISHES 31
Findings
The findings are presented in Tables 7-9 and summarised in Table 10. Of all three HKE sub-
jects, only J and K style-shifted between HKE-gl and HKE-lo intonation, whereas T seemed
to display only features of HKE-lo intonation. Neither HKE-gl nor HKE-lo intonational
features were found in the non-HKE subjects, whose results are therefore not included here.
Js use of intonation for tasks A and B consisted solely of HKE-lo. Tasks C and D, how-
ever, managed to elicit only 2% and 1.5% of HKE-lo intonational features respectively. On
the other hand, though designed to be HKE-gl tasks, some of the prompts in tasks E and F
elicited a small amount of HKE-lo prosodic features, due to the use of some HKE vocabulary
item; for instance, cha chaan ting and dim sum. Those exceptions aside, the rest of tasks
E and F did conform to yielding the anticipated HKE-gl features. Tasks G and H yielded
only HKE-gl intonational features and none of HKE-lo.
Ks performance was largely similar to that of Js, demonstrating 100% of HKE-lo prosodic
features in tasks A and B, while showing only 0.6% and 3.8% of the local accent in tasks C
and D. Similar to J, K adopted only HKE-gl prosodic features in both tasks G and H. Unlike
J, however, he employed only HKE-gl features in E and F.
Unlike J and K, T used HKE-lo prosody consistently throughout all tasks, which means
he demonstrated an absolute 100% use of HKE-lo intonation.
ASIAN ENGLISHES 33
Discussion
Since all three participants were locally born, brought up and educated, their L2 experiences
can crudely be regarded as similar. Any disparity in L2 intonation development should be
primarily attributed to the difference in L2 intonation aptitude.
Contrary to popular belief, proficiency levels are not a good indicator of intonation
aptitude. Having above-average English proficiency, all three speakers English should be
well-pronounced; their L2 intonation aptitude, however, seemed to differ. One reason could
be that the criteria for being well-pronounced might have a stronger focus on segmentals
than suprasegmentals, as is traditionally often the case. Judging from the results, the fact
that both J and K employed both HKE-lo intonation (mainly in tasks A and B) and HKE-gl
intonation (in tasks C, D, E, F, G and H) was proof that they had access to a wider range of
intonation. T, on the other hand, demonstrated only the local features, as shown in Table
10. The fact that T did not (or failed to) use HKE-gl intonation even for formal and reading
aloud tasks such as tasks E and F, or any of the other tasks for that matter, suggested that his
intonation aptitude was relatively low, i.e. not capable of switching to HKE-gl intonation due
to its absence. He, however, did show some awareness of formality, e.g. through his word
choice and grammar, showing that his less than advanced ability was only in the area of
intonation, which seems to fit the description of the Joseph Conrad Phenomenon (Reiterer
et al., 2011), as introduced previously, to describe such a discrepancy in abilities. Given the
similarity in experience and proficiency, the unconscious character of intonation (Ramirez
Verdugo, 2006, p. 141) could only be overcome by advanced levels of L2 intonation aptitude,
which are generally considered rare and available only to a limited few. The Joseph Conrad
Phenomenon can be said to apply to T only, and not J and K. Just as Sewell (2016, p. 52)
described, people may wish to adopt certain identity positions, or roles, but be unable to
because they cannot make effective use of the accent features required. T is thus an example
of such people. This goes to show that educational attainment or language proficiency is
not a reliable indicator of ones pronunciation or intonation aptitude. It is hence interpreted
that J and K have better L2 intonation aptitude than T. Since intonational variation in HKE
was found only in J and K, sociolinguistic variation of HKE intonation is believed to be
possible only in speakers with higher intonation aptitude.
Table 11.JE3A.
You can drink tea or coffee in there
H H H L-L%
Table 12.JA3A.
You can drink tea or coffee in there ga
M M H H M HL M H L-L%
34 T. LAM
Table 13.JE3B.
You can drink tea or coffee in there
H H H H-H%
Table 14.JA3B.
You can drink tea or coffee in there ga meh
M M H H M HL M H M H-H%
Table 15.JE3C.
You can drink tea or coffee in there
H H H H L-L%
Table 16.JA3C.
You can drink tea or coffee in there ga
H M H H M HL M H M-M%
Table 17.JA4B.
Tsui Wah is a famous local cha chaan ting meh
H H M M HH HL L H H H-H%
Table 18.JA4A.
Tsui Wah is a famous local cha chaan ting ah
H H M M HH HH L H H M-M%
only. The first eight examples (Tables 1118) to be presented in the subsequent pages were
taken from J.
An example of HKE-gl intonation can be seen in Table 11, produced by J for prompt 3A
of the reading task (E). This example came from sentence group 3, uttered as a statement.
Note that Figure 1 was read with only three high pitch accents, on the words drink, tea
and the first syllable of coffee, ending with a low phrase accent and low boundary tone,
which is rather close to the intonation pattern of a native variety of English in the sense
that not every single syllable was assigned a pitch accent, as discussed in Table 4. The rest of
the syllables, e.g. or and in there, were phonologically toneless and only realised phonet-
ically between the assigned pitch accents. This is to be contrasted with HKE-lo intonation
presented in the next example.
Taken from 3A of Task A, similar to Task E except for the SFP, Table 12 presents the
prompt produced by the same speaker but with features of HKE-lo intonation. This ren-
dition of the exact same statement plus the SFP by the same speaker brought about some
interesting changes, demonstrating a series of lexical tone patterns, largely following the
intonational rules enumerated by Gussenhoven (2012, 2014) based on Luke (2000, 2008).
ASIAN ENGLISHES 35
Such an abundance of pitch accents in so many words in a sentence would be most unnat-
ural in native varieties of English (Nolan, 2006), and yet this is the case in virtually every
single syllable in the sentence. One prominent characteristic of HKE-lo intonation is thus
the assignment of pitch accents to all syllables in a given utterance based on the Cantonese
tone system, as discussed in the literature and represented in Table 3.
The use of a pitch accent in every syllable is not confined to statements; the same seems
to be true for questions and emphatic declaratives. Like Table 11, Table 13 shows the same
utterance read by J for Task E as a question. The only difference is this time it ends with
high edge tones. This is contrasted with the parallel question in task A uttered by the same
speaker, as shown in Table 14, where again every single syllable was given a pitch accent. The
same phenomenon can also be seen in the examples of emphatic statements in Tables 15 and
16, produced by the same speaker for tasks E and A respectively. The former was assigned
four high pitch accents in drink, tea, coffee and in, whereas the latter once again carried
a pitch accent in all syllables. The rest of the data from Task A and Task E did follow these
patterns. It seems reasonable at this point to state that HKE-gl intonation is distinctly dif-
ferent from that of HKE-lo, which largely follows the rules specified by Gussenhoven (2012,
2014). The same is true for speaker K; though not T, due to his lower intonation aptitude.
However, Gussenhovens rules do not seem to explain everything. It appears that the
low tone is not just inserted after the last high tone in the intonational phase (IP), as sug-
gested in Rule 2. Note that all the rules but one have been followed in Tables 12, 14 and 16,
namely the L tone in coffee. One explanation could be that the sentence is treated as two
IPs instead of one, rendering you can drink coffee and in there ga, hence placing the L
tone at the end of the first IP, i.e. boundary tone. However, one can argue that there does
not seem to be a significantly long enough pause for it to warrant the end of the IP. They
are, perhaps, closer to a lower ranking phonological constituent, e.g. intermediate phrases
(ip, Beckman & Pierrehumbert, 1986) or phonological phrases (PhP, Inkelas & Zec, 1990;
Nespor & Vogel, 1986; Selkirk, 1984, 1986) instead. Cheung (2009) suggested that this low
tonal element is a boundary tone of a phonological phrase. An alternative approach is that
at least one of the rules suggested above might be incomplete. For instance, Rule 2 might be
modified as insert L after last H in the IP or insert L in word-ending syllable(s). Another
example that seems to support the latter theory can be seen in Table 17, where the word
local, with HL, also appears to violate Rule 2. However, Table 18 gives an example of the
same word but following the original rule. Also notice that in both cases the word famous
is pronounced with two high tones, HH. Both tables are taken from Task A produced by
J, with the former being a question, the latter a statement. The fact that the violation can
appear in all three sentence types, as in Tables 12, 14 and 16, suggests that this has nothing
to do with sentence types. From inspecting Tables 17 and 18, it is observed that the HH
pattern, as represented in the word famous, can appear in both cases while the word local
can assume either HL (Table 17) and HH (Table 18). The phenomenon where the same
word can be pronounced in two different ways seems to warrant the revision of Rule 2
suggested above, which might be a case of variation.
Another violation of Gussenhovens rules is the use of the Cantonese mid-rising tone
(MH) in a single syllable in some of the recordings (see 5AC, 6AC, task B in Appendix
- J), appearing in words such as the first word of dim sum and the second word of har
gau and siu mai, as well as both words of cheung fun. This is believed to be the result of
indexicality, allowing the use of code-switching, rendering the elicitation and analysis of the
36 T. LAM
more casual yet stigmatised HKE-lo accent possible. As a result, the rest of the Cantonese
tones might theoretically be allowed into HKE, in addition to the H, M and L tones in the
original repertoire, especially when pronouncing typical HKE vocabulary items such as
the examples stated above. However, the use of Cantonese tones is not confined to lexical
items, but also the SFPs.
The presence of SFPs expands the repertoire of boundary tones. Not only does a Cantonese
SFP carry a lexical tone in itself, it can affect and reveal other edge tones that might otherwise
not show up, e.g. M-M% in Table 16, which is different from what Gussenhoven suggested
for the emphatic declarative boundary tone: L%. It would then be reasonable to expect that
the use of other SFPs might be able to elicit other tones. Table 14, for instance, demonstrates
the use of not one but two SFPs in succession, where the phrase accent and boundary tone
seem to follow the second SFP meh, hence H followed by H-H%. It is thus hypothesised
that since SFPs are such an essential part of HKE, and seem to play numerous important
roles in intonation, excluding them would fail to reveal the full range of HKE intonation
features, especially the HKE-lo ones.
Despite the violation/variation of Rule 2, the presence of the mid rising tone and the
extra boundary tones brought about by SFPs, it seems clear that there are apparent dif-
ferences between the intonational features of HKE-gl and HKE-lo. The fact that the same
speaker produced both HKE-gl and HKE-lo accents when pronouncing (almost) the same
utterance suggests that at least some HK speakers of English have both accents at their
disposal, being able to perform style-shifting when necessary, alluding to sociolinguistic
variation of HKE intonation. The next section hence explores the patterns of style-shifting
in the data.
Speakers J and K seemed to favour or be more comfortable using the HKE-gl prosody. This
could also explain why HKE-lo prosody could only be elicited in larger quantities when
coaxed and coerced with the use of SFPs in tasks A and B.
In addition to the stigmatised status of HKE-lo, the discrepancy in the use of the global
and local intonational features can also be attributed to the experimental tasks, shedding
light on the strengths and weaknesses of the different approaches to stylistic variation.
Though all of the first four tasks (ABCD) were meant to encourage the use of HKE-lo,
only the first two were successful in eliciting more HKE-lo prosodic features. On the other
hand, of the tasks intended for eliciting HKE-gl intonation (EFGH), E and F also yielded
some limited features of HKE-lo prosody. This implies that the topic (of HK culture) alone
was insufficient to invite the use of HKE-lo prosody, as shown in tasks C and D, potentially
discrediting the Attention to Speech approach to stylistic variation. The fact that only tasks
A and B were successful in eliciting more prosodic features of the taboo HKE-lo might be
pertinent to the mandatory inclusion of SFPs in the prompts, which did not exist in tasks
C and D, due to their spontaneous nature. It seems to suggest that the use of an SFP can act
as a primer to signal to the speaker to use the normally stigmatised, local accent. Whether
such a theory holds up remains to be seen in a larger study. Both topic and the compul-
sory use of SFPs are considered to belong to the Attention to Speech approach to stylistic
variation, yet only the latter is considered effective at eliciting the use of HKE-lo prosody.
The Audience Design approach seemed to have less significant effects on style variation.
In the two dyad discussion tasks between subjects C and G, the nationality and accent used
by one interactant did not seem to have any significant bearing on the accent used by the
other. In the first discussion (task C) between J and T, J employed mainly HKE-gl intona-
tion (98%), rendering a mere 2% use of the local counterpart. Similarly, in the same task
between K and S (non-HKE), Ks use of HKE-lo was almost non-existent (0.6%). In other
words, whether or not one was interacting with a HKE or non-HKE speaker did not seem
to matter in terms of ones use of HKE-lo or HKE-gl prosodic features, which can be used
to discredit the Audience Design approach to style shifting, at least to some extent. The
HKE speakers seemed to be more aware of and sensitive to something bigger and more
powerful that was at work, e.g. the stigmatised status of HKE-lo, than the nationality and
accent used by another speaker.
Espousing the active creation and re-creation of speaker identity, the Speaker Design
approach seems useful in explaining the use of HKE intonation by the three speakers. With
the exception of tasks A, B, G and H, all the other tasks yielded utterances that were impure,
or mixing both HKE-gl and HKE-lo prosodic features in the same utterances, even in tasks
intended solely for eliciting HKE-gl, e.g. prompts 4, 5 and 6 of task E and task F with some
typical HKE vocabulary items such as cha chaan ting in prompt 4, which both speakers J
and K chose to read with HKE-lo. As for dim sum in prompt 5, J chose to read with HKE-lo
while K opted for the HKE-gl version. Despite the rather low usage, there were HKE-lo
features being used in tasks C and D, namely 2% and 1.5% for speaker J; 0.6% and 3.8%
for K, alongside the majority of HKE-gl features being used. The co-occurrence of HKE-gl
and HKE-lo is a sign of polycentricity, showing both global and local orientations and
identities on the same occasions. Note that it is assumed that given the chance (appropriate
level of intonation aptitude), it is believed that speaker T would also have liked to perform
intonational variation in HKE. The reason he did not, as previously discussed, was due to
the constraints posed by his limited L2 intonation aptitude.
38 T. LAM
Despite those having a repertoire of both HKE-gl and HKE-lo accents, there was an
apparent disparity in the distribution of the two, due to reasons such as the stigmatised status
of the local accent, to the extent where even the use of informal task types, topics related to
local culture, and local interactants was unsuccessful in eliciting the taboo prosodic features.
However, the presence of SFPs seems to encourage the use of HKE-lo intonation, implying
the close relationship between the two. Although the participants did exclusively use either
HKE-gl or HKE-lo intonation in an utterance, there were instances of mixing the two within
the same utterance, alluding to their polycentricity. Due to the relatively low usage rate of
HKE-lo prosody, when it did occur, it would be expected to perform special functions.
Table 20.KF5.
No visitors can leave Hong Kong without first going for some dimsum
H HL H H H L-L%
ASIAN ENGLISHES 39
from a formal register (as indexed by the formal syntax) to an informal one. The rest of
the sentence, consisting of a series of HKE lexical tone patterns, also displays intonational
features of HKE-lo, strengthening the indexical effects. The particular instance of HKE-lo
intonation for the higher-order indexical process is believed to be more likely to happen
between HK speakers, as in this case, than inter-ethnic exchanges since it involves both
linguistic and cultural knowledge of Hong Kong. However, the use of HKE-lo intonation
in intercultural communication is still possible, provided that the speaker has reason to
believe the interlocutor would possess the linguistic and cultural knowledge of Hong Kong,
as can be shown in the next example. It is also worth noting that the instances of HKE-lo
features were mixed with HKE-gl features within the same utterances, lending support to
polycentricity.
There were but a handful of occurances of indexing in Extract 2, an exchange between
Speaker K and Speaker S, the latter being Indian, produced for Task C. There were occasions
when restaurants and local delicacies are mentioned (e.g. line 8) but no HKE-lo prosody
was employed. A strong indexical device, however, has been used in this exchange, namely
the SFP. The first one occurred in line 1 and the second in line 14. The use of the particles
by K might index a certain level of informality, e.g. the topic of food, and involve a high-
er-order indexicality, community membership, which, however, was clearly not echoed by
Speaker S with the absence of particles in his responses, leading to fewer instances of HKE-lo
intonation, compared to the previous example. The fact that there were still a number of
HKE-lo prosodic features used between a mixed-ethnic discussion seems to suggest that
style-shifting involves a combination of factors, e.g. informality, rather than just being based
on the interactant/audience.
ASIAN ENGLISHES 41
Conclusion
Owing to its mixed, Eurasian background, HKE has been influenced by both its substrate
tone language (i.e. Cantonese) and donor native varieties of English (e.g. British English
and/or American English), resulting in its rather unique and distinct phonological features,
both segmental and suprasegmetal. Adopting the unified approach in the pilot study, it was
demonstrated that there were different intonational features involved in the two accents of
HKE, namely HKE-gl and HKE-lo. A closer look at the HKE-lo intonational features with
the inclusion of SFPs in the HKE data has contributed to the revision of the phonological
rules of HKE intonation. Not only does HKE intonation, including the use of SFPs, func-
tion at the sentence level, but it also functions at the discourse level, performing pragmatic
functions. The fact that there were two distinct styles of HKE implies that it might involve
choosing (consciously or unconsciously) one style over the other. Such style-shifting would
arguably entail negotiating within the multidimensional space between two cultural ori-
entations, namely local and global, including numerous sociolinguistic dimensions such
as polycentricity, global/local identities, global/local experiences, L2 intonation aptitude,
formality/informality and so on. It is therefore worth researching further into the sociolin-
guistics of HKE intonation; for instance, the use of HKE-lo intonation for sociolinguistic,
indexical purposes, involving a larger group of participants.
In addition to sociolinguistic factors, it was found that the presence of an SFP could be
used to elicit the more tabooed and stigmatised HKE-lo accent, suggesting a linguistic factor
involved in the style-shifting process. The exact interactions between SFPs and intonation
of HKE will need to be studied in greater detail, yet it is safe to say that they play a rather
important role in HKE intonation.
Given the rather distinct features of intonation associated with the two accents of HKE
due to its mixed linguistic background and complex sociolinguistic milieu in the globalised
city, it is hypothesised that the intonational variation of the two HKE accents should be
governed by rather different phonological rules, with the HKE-lo accent inheriting more
features from its substrate tone language than the HKE-gl counterpart. A thorough inves-
tigation should be carried out to look into the intonational phonologies of both HKE-gl
and HKE-lo accents.
Apart from the phonological aspect, the phonetics of HKE intonation should also be
explored, as Ladd (2008, p. 14) explained: a complete phonological description does not
consist of abstract formulas alone, but must also specify how the abstract formulas are
realised; that is, it must describe the mapping from the categorical phonological elements
to the continuous acoustic parameters. Implementation and phonetic realisation rules of
HKE intonation such as downstep should also be investigated.
Due to the rather limited size of the subject pool in the pilot study, the results, though
important, were preliminary at best. A larger-scale study involving a more sizeable group of
informants is proposed, to confirm these preliminary results as well as extending the scope
of study to include a more thorough investigation into the intonation of HKE.
Note
1.
I use the term aptitude loosely here as an umbrella term to include aptitude, talent, skills,
ability and proficiency.
42 T. LAM
Disclosure statement
No potential conflict of interest was reported by the author.
References
Alsagoff, L. (2007). Singlish: Negotiating culture, capital and identity. In V. Vaish, S. Gopinathan, &
Y. B. Liu (Eds.), Language, capital, culture: Critical studies of language and education in Singapore
(pp. 2546). Rotterdam: Sense.
Alsagoff, L. (2010). English in Singapore: Culture, capital and identity in linguistic variation. World
Englishes, 29, 336348.
Anufryk, V. (2009). Prosodic factors in language proficiency: Intonational variation as a signature
of pronunciation talent. In G. Dogil & S. M. Reiterer (Eds.), Language talent and brain activity.
Berlin; New York: Mouton de Gruyter.
Beckman, M., & Pierrehumbert, J. (1986). Intonational structure in Japanese and English. Phonology
Yearbook, III, 1570.
Blommaert, J. (2010). The sociolinguistics of globalization. Cambridge: Cambridge University Press.
Bolton, K., & Kwok, H. (1990). The dynamics of the Hong Kong accent: Social identity and
sociolinguistic description. Journal of Asian Pacific Communication, 1, 147172.
Cheung, W. H. Y. (2009). Span of high tones in Hong Kong English. Proceedings of the 35th Annual
Meeting of Berkeley Linguistics Society (BLS 35) (pp. 7282). Berkeley CA.
Couper-Kuhlen, E. (1986). An introduction to english prosody. London and Tuebingen: Edward Arnold
and Niemeyer.
Cruttenden, A. (2001). Mancunian intonation and intonational representation. Phonetica, 58, 5380.
Dankovicova, J., House, J., Crooks, A., & Jones, K. (2007). The relationship between musical skills,
music training, and intonation analysis skills. Language and Speech, 50, 177225.
Deterding, D., Wong, J., & Kirkpatrick, A. (2008). The pronunciation of Hong Kong English. English
World-Wide, 29, 148175.
Eckert, P. (2008). Variation and the indexical field. Journal of Sociolinguistics, 12, 453476.
Grabe, E. (2004). Intonational variation in urban dialects of English spoken in the British Isles. In P.
Gilles & J. Peters (Eds.), Regional variation in intonation (pp. 931). Tubingen: Niemeyer.
Gussenhoven, C. (2012). Tone and intonation in Cantonese English. The Third International Symposium
on Tonal Aspects of Languages, Nanjing, May 26-29.
Gussenhoven, C. (2014). On the intonation of tonal varieties of English. In M. Filupula, J. Klemola
& D. Sharma (Eds.), The oxford handbook of world Englishes online. Retrieved March 1, 2016,
from http://gep.ruhosting.nl/carlos/Gussenhoven_Handbook_VarEng.pdf
Hansen Edwards, J. G. (2015). Hong Kong English: Attitudes, identity, and use. Asian Englishes, 17,
184208.
Hong Kong Examinations and Assessment Authority. (2014). Hong Kong Diploma of Secondary
Education Examination English Language Level Descriptors Subject Descriptors. Retrieved
November 29, 2016, from http://www.hkeaa.edu.hk/DocLibrary/HKDSE/Subject_Information/
eng_lang/LevelDescriptors-ENG-Subject.pdf
Hung, T. T. N. (2002). Towards a phonology of Hong Kong English. In K. Bolton (Ed.), Hong Kong
English autonomy and creativity (pp. 119140). Hong Kong: Hong Kong University Press.
Inkelas, S., & Zec, D. (Eds.). (1990). The phonology-syntax connection. Chicago, IL: Chicago University
Press.
James, G. (2001). Cantonese particles in Hong Kong students English emails. English Today, 17, 916.
Labov, W. (1972). Sociolinguistic patterns. Philadelphia, PA: University of Pennsylvania Press.
Ladd, D. R. (2008). Intonational phonology. New York, NY: Cambridge University Press.
Leimgruber, J. (2012). Singapore English: An indexical approach. World Englishes, 31(1), 114.
Lim, L. (2009). Revisiting English prosody: (Some) New Englishes as tone language? In L. Lim & N.
Gisborne (Eds.), The Typology of asian englishes Special Issue, English World-Wide 30(2). 218239.
Low, E.-L., & Hashim, A. (Eds.). (2012). English in Southeast Asia Features, policy and language in
use. Amsterdam/Philadelphia: John Benjamins Publishing Company.
ASIAN ENGLISHES 43
Luk, J. C. M., & Lin, A. M. Y. (2006). Uncovering the sociopolitical situatedness of accents in the
world Englishes paradigm. In R. Hughes (Ed.), Spoken English, TESOL and applied linguistics (pp.
322). New York, NY: Palgrave Macmillan.
Luke, K. K. (1990). Utterance particles in cantonese conversation. Amsterdam: John Benjamins.
Luke, K. K. (2000). Phonological Re-interpretation: The assignment of cantonese tones to english words.
The 9th International Conference on Chinese Linguistics, Singapore, June.
Luke, K. K. (2008). Stress and intonation in Hong Kong English. Paper presented at the 14th Conference
of the International Association for World Englishes (IAWE), Hong Kong.
Luke, K. K., & Richards, J. (1982). English in Hong Kong: Functions and status. English World-Wide,
3, 4764.
Matthews, S., & Yip, V. (2011). Cantonese A Comprehensive Grammar. London and New York:
Routledge.
Nespor, M., & Vogel, I. B. (1986). Prosodic phonology. Dordrecht: Foris.
Nolan, F. (2006). Intonation. In B. Aarts & A. McMahon (Eds.), The handbook of English linguistics
(pp. 433457). Malden, MA, USA: Blackwell Publishing.
Ramrez Verdugo, D. (2006). A study on intonation awareness and learning by non-native speakers
of english. Language Awareness Journal, 15, 141159.
Reiterer, S. M., Hu, X., Erb, M., Rota, G., Nardo, D., Grodd, W., Winkler, S., & Ackermann, H. (2011).
Individual differences in audio-vocal speech imitation aptitude in late bilinguals: Functional neuro-
imaging and brain morphology. Frontiers in Psychology, 2, 112.
Schilling-Estes, N. (2003). Investigating Stylistic Variation. In J. K. Chambers, P. Trudgill & N. Schilling-
Estes. (Eds.), The handbook of language variation and change. Blackwell Publishing. Retrieved from
http://www.blackwellreference.com/subscriber/book?id=g9781405116923_9781405116923
Schneider, E. W., & Kortmann, B. (Eds.). 2004. A handbook of varieties of English. Vol. 1: Phonology.
Berlin: Mouton de Gruyter.
Selkirk, E. (1984). Phonology and syntax: The relation between sound and structure. Cambridge, MA:
MIT Press.
Selkirk, E. (1986). On derived domains in sentence phonology. Phonology Yearbook, 3, 371405.
Sewell, A. (2012). The Hong Kong English accent: Variation and acceptability. Hong Kong Journal of
Applied Linguistics, 13, 121.
Sewell, A. (2016). English pronunciation models in a globalized world Accent, acceptability and Hong
Kong English. London and New York: Routledge.
Sewell, A., & Chan, J. (2010). Patterns of variation in the consonantal phonology of Hong Kong
English. English World-Wide, 31, 138161.
Silverstein, M. (2003). Indexical order and the dialectics of sociolinguistics life. Language and
Communication, 23, 193229.
Sung, C. C. M. (2014). Accent and identity: Exploring the perceptions among bilingual speakers
of English as a lingua franca in Hong Kong. International Journal of Bilingual Education and
Bilingualism, 17, 544557.
Trofimovich, P., Kennedy, S., & Ann Foote, J. (2015). Variables affecting L2 pronunciation development.
In M. Reed & J. M. Levis (Eds.), The Handbook of English pronunciation. Hoboken, NJ: John Wiley
& Sons.
Warren, P., & Britain, D. (2000). Prosody and intonation of New Zealand English. In A. Bell & K.
Kuiper (Eds.), New Zealand English (pp. 146172). Wellington: Victoria University Press.
Wells, J. C. (1982). Accents of English, Vol. 2. Cambridge: Cambridge University Press.
Zhang, L. (2014). Segmentless sentence-final particles in cantonese: An experimental study. Studies
in Chinese Linguistics, 35, 4760.
Zhang, Q. (2014). Investigating Hong Kong English. Oxford: Peter Lang.