Lim lisa singapore english a grammatical description

This volume is particular because its description of structural features of Singapore English is based on a large corpus of naturally-occurring spontane-ous speech of young, native speak

General Series

Volume G33

Singapore English: A grammatical description

Edited by Lisa Lim

Library of Congress Cataloging-in-Publication Data

Singapore English : a grammatical description / edited by Lisa Lim.

p cm (Varieties of English Around the World, issn 0172–7362 ; v g33)

Includes bibliographical references and indexes.

1 English language Singapore Grammar 2 English

language Variation Singapore 3 English language Dialects-language Variation Singapore 4

Singapore Languages I Lim, Lisa II Series.

Cover picture: The sculpture ‘The River Merchants’ (sculptor: Aw Tee Hong) is located at

the mouth of the Singapore River, and portrays Alexander Laurie Johnston, Scotsman and prominent merchant, interacting with a Chinese trader and Malay chief It is in encounters

of this type where different languages meet that innovative linguistic features – such as those found in Singapore English – arise

Acknowledgements XI

Chapter 

. Preamble 

.2 A brief history 2

.2. English arrives in Singapore 2

.2.2 English in education 3

.2.3 And English in the home and streets 5

.3 A brief history of Singapore English scholarship 7

.4 The Grammar of Spoken Singapore English Corpus (GSSEC) 0

.4. Collecting the data 

.4.2 Who are the Singapore English speakers? 

.4.3 The resulting corpus 2

2.2.2 Monophthong realisation and distribution 20

2.2.3 Diphthongs, and triphthongs? 23

2.3.6. Syllabic consonants 32 2.3.6.2 Final consonant deletion or replacement 32 2.3.6.3 Consonant cluster reduction 33

2.4..4 Exclamatives 4

2.4.2 Characteristic CSE forms 42

2.4.2. Sustained level steps 42 2.4.2.2 Phrase-final lengthening 43 2.4.2.3 Particles 45

2.4.2.4 Pitch patterns and particles from the substrates? 47 2.4.3 Focus and prominence 49

2.4.3. Focus placement 49 2.4.3.2 Prominence cues 52 2.4.3.3 Substrate prosody? 52 2.5 Conclusions 53

3.3 Number and agreement 63

4.2. Optionality of verbal inflections 77

4.2.2 Verb forms and their uses 79

4.3 The copula be and other predicate phrases 82

4.4 Be as an auxiliary 85

4.4. The progressive auxiliary be 85

4.4.2 The passive auxiliary be 86

4.4.3 Be as part of be supposed to 87

4.5 Factors determining the (co-)occurrence of auxiliaries 87

4.5. Auxiliary clusters 87

4.5.2 Subject NPs, negation, and the auxiliary 88

4.5.3 Wh-interrogatives and subject-auxiliary inversion 9

4.5.4 ‘Do-support’ 9

4.6 Verb reduplication and aspectual classes of events 92

4.6. Verb reduplication 92

4.6.2 Tests for aspectual classes in CSE 93

4.6.3 Constraints on verb reduplication 95

4.7 Passive constructions 97

4.7. Four types of passive constructions 97

4.7.2 The kena passive 98

4.8 Conclusion 99

5.3 The possibility of substratal influence 3

5.3. Other cases of substratal influence 3

5.3.2 Substratal influence in the case of CSE reduplication 4 5.4 Discourse particles 7

6.3.2.4 ‘Passives’ 38 6.3.2.5 Nouns and polyfunctionality 39 6.3.3 Final remarks 40

6.4 Origins of English in Singapore 4

6.5 Conclusions 45

My first expression of appreciation is for Umberto Ansaldo, who one day got

me to see that I could put this volume together, and who all days got me to keep at it and to believe it would be worth it I also owe my thanks to Joe Foley for having invited me to be Co-P.I on the project whose data formed the basis

of this book; and the other members of the research project team and authors

of chapters in this volume for being integral and interested parties in both deavours, and for enduring the numerous delays both have encountered, always with good cheer: I hope it’s been worth it I am especially beholden to Lionel Wee and Umberto Ansaldo for their extra little contribution, stepping in when things seemed at a stand-still

en-I am grateful to Salikoko Mufwene for his enthusiasm over the volume and his thoughts and feedback on earlier drafts of some chapters; and to Edgar Schneider for his critical eye and constructive comments on later drafts of the manuscript: I am glad that he felt this a volume that would make a valuable contribution to the Varieties of English Around the World series, and trust that

it does not disappoint I am also thankful for the incredible patience and good humour that Kees Vaes at John Benjamins must certainly possess, and his for-tunately-not-misplaced faith that the manuscript would ultimately come in.The descriptions in this volume would not be without the data I acknowl-edge the support provided by the National University of Singapore’s Academic

Research Grant R-103-000-003-112 which funded the research project Towards

a Reference Grammar of Singapore English in which the Grammar of Spoken

Singapore English Corpus (GSSEC) database was collected Our students at the Department of English Language and Literature at the National University

of Singapore, too numerous to mention by name, also played no small part

in this For toiling over the orthographic and phonetic transcriptions, we are obliged to the Honours and postgraduate students I am particularly indebted

to Diane Chang who, amongst many other tasks, digitised the data and checked and standardised the transcripts, and whose technological know-how with recordings and webpages has been invaluable Finally, where would we be if not for our native Singapore English speakers and their friends and family who put up with having their otherwise private lives recorded, and who provided

conversations which not only comprise a precious resource but which are,

above all, animated, amusing, and utterly Singaporean Thanks, hor.

LL

Amsterdam & Singapore; June 2004

Table 2.1 Standard Lexical Sets for SSBE, GA, SSE and CSE 2

Table 2.5 Monophthongs of (Singapore) Mandarin 26

Table 2.8 CSE (merged) phonemes matched with vowels

Table 2.12 Consonants of (Singapore) Mandarin 36

Table 2.15 A summary of the INTSINT symbols, adapted for SE 38

Table 5.4 Similarities in reduplication functions in CSE, Malay

Table 6.1 Reduplication patterns and functions in SE and

English in Singapore and Singapore English

Background and methodology

Lisa Lim† and Joseph A Foley‡

†University of Amsterdam

‡Southeast Asian Ministers of Education Organisation

(Regional Language Centre), Singapore

There are not many these days who are not acquainted with Singapore English (SE) – often fondly referred to as Singlish:1 much scholarly research has been published on it; it figures in current surveys of varieties of English as well as in textbooks; it has a listing in dictionaries, and even has dedicated websites What

is the need then for yet another volume on Singapore English?

This volume is particular because its description of structural features of Singapore English is based on a large corpus of naturally-occurring spontane-ous speech of young, native speakers of Singapore English, and hence what the reader will find in this volume is a description of a vibrant, current, contempo-rary, colloquial Singapore English, spoken by the Singapore English speaker of today, that is, those young Singaporeans who grew up during a period where, for all their school-going years, they have had to be what has been referred to

as “English-knowing bilinguals” (Pakir 1992), that is, they have had to be cient in English and their ‘mother tongue’ (these notions to be expounded upon

profi-in Section 1.2.2) This volume will describe not just the features which distprofi-in-guish SE from other varieties of English and, indeed, which characterise it as SE, but will also highlight the many aspects that it does share with other varieties.Singapore English has, since first written about three decades ago (Tongue 1974), morphed into quite a different animal, and it still continues to change This is in no small way because of the speed at which language policies are implemented, and the speed at which the country and its community has devel-oped The next section will provide a historical and sociolinguistic background

distin-to place the variety in perspective

.2 A brief history

.2. English arrives in Singapore

In a small island of 4 million people made up of 76.8% Chinese, 13.9% Malay, 7.9% Indian and 1.4% persons of other races (Leow 2001), it may seem strange that English, the language of the colonial rulers, should have the dominance and scope that it commands in the day-to-day life of Singaporeans The situation is partly the legacy of colonial history and partly the effect of post-independence policies in which English has been recognised as a resource to increase the country’s rate of economic and social development

1819 tends to be recognised as marking the first significant event,2 when Sir Stamford Raffles acquired the island for the British East India Company Singapore consequently became part of the Straits Settlements, which at that time consisted of Penang and Malacca Singapore’s location at the crossroads

of the trade routes in Southeast Asia and China made it an effective counter to Dutch trading operations in Indonesia, and in less than a year, more than five thousand people had settled on the island, including, in addition to those that came with the colonial power, migrants from the neighbouring countries The most important groups were the Malays from Malaya, the Javanese, Balinese, Bataks and Buginese from Indonesia, the Tamils, Malayalees, Punjabis, Sikhs and other ethnic groups from India, and the Hokkiens, Cantonese, Teochews, Hakkas, Hainanese and Chinese from other dialect groups from China; other minorities included Arabs and Filipinos

It should be noted that, even prior to this influx of settlers, the British administration already found on arrival a heterogeneous population of local Malays and Indonesians, Southern Chinese and Indian traders, and some mixed ethnic groups such as the Peranakans and possibly Eurasians Also al-

ready in place was a capitan (‘captain’) system, which divided the society into three groups: the Malays, Chinese, and Indians, plus a capitan-less group of

‘others’, and each ethnic community had in effect its own legal system under

the jurisdiction of its own capitan (Bloom 1986: 352) The British found it to

their advantage to preserve this ethnically-based division, and to this day, it constitutes the cultural logic of Singapore’s “multiracialism” (Benjamin 1976) that underpins its current language policy (see Section 1.2.2)

While as a whole the ethnic Malays were the largest single group in ninsular Malaysia, they did not form a majority in the main urban areas of Singapore, Penang, Malacca, Kuala Lumpur and Kuching The Chinese settled

pe-the urban areas, and pe-the Malays’ small majority of 52% in pe-the Straits Settlements

in 1871 was gradually eroded by the influx of Chinese, as reported in the Straits Settlements Censuses (Gupta 1994: 34) Two decades later in 1891, the Chi-nese had grown from 34% to 44% of the population, creeping past the Malays’ 42% By 1931, the Chinese had attained a strong majority of 60% compared to

the Malays, comprising now only 26% In spite of this, Bazaar Malay (Bahasa

Pasar), a pidginised form of Malay, was always the main lingua franca Many

of these migrants came with the intention of making their fortunes and then returning home, but some stayed on and thus contributed to the cultural mix

of racial and linguistic groups in early Singapore

in the three Rs [reading, ’riting and ’rithmetic] in English” (Bloom 1986: 351)

It appears though that the British were nonetheless keen to cultivate a group

of English-educated elites, and by the 1830s there was some English-medium education in all three settlements of Penang, Malacca and Singapore; these were either government schools or schools run by church missions (‘mission schools’) However the actual meaning of the term ‘English-medium’ changed over time: in the early years, in fact, some of these ‘English-medium’ schools used or taught other languages, principally Malay (Gupta 1994) Indeed, in

1870, the British reported having produced young men “competent to earn a livelihood in Government and mercantile offices, but it is much to be regret-ted that the majority of these clerks know only how to read, write and speak English imperfectly” (cited in Bloom 1986: 358) What is crucial is that English had been established as the language for socio-economic mobility, and by 1900, this group of elites had come to enjoy a much greater degree of English language proficiency and to also cover a much wider occupational range

In the mid-nineteenth century, the number enrolled in the English-medium schools was only 922 In 1872 there were 1,722, and this enrolment remained constant for the next twenty years, rising to 2,883 in 1891 The enrolment then

began to go up much more rapidly such that by the end of the century, the students enrolled in the English-medium schools totalled 7,264 Whatever may have been the attitude of the various ethnic groups in the Straits Settle-ments and the other Malaysian States to British Colonial policy or to European culture, it was obvious that English-medium education had many material advantages By 1937 the total enrolment in the Federated Malay States was 17,161 (Loh 1974), and in Singapore, the enrolment just before the Japanese occupation was 27,000 (Doraisamy 1969) Non-English-medium schools, that

is, those which were Malay- or Chinese-medium, did not teach English at all in the early years, but by the 1920s and 1930s, many of these schools were at least teaching English as a subject

After the hiatus of the second world war (1942–1945), English became even

more the lingua franca of the more elite sections of society, partly due to the

growing importance of the United States as a global power in the region Indeed,

in these pre-independence days in the 1940s to 1950s, the British still sought

to promote English-medium schooling among the elite, while subsidising the Malay language streams (Tickoo 1996: 434); this lead, however, to discontent amongst the majority Chinese who continued sending their children to Chi-nese-medium schools In 1947, about 32% of students were enrolled in English-medium schools In 1950s’ Singapore, education became effectively universal, and English-medium education became more and more the norm By 1952, 43% of the school enrolment was English-medium, with the numbers register-ing for English-medium education overtaking those registering for Chinese-medium education by the end of that decade (Doraisamy 1969) The language education policy in 1959 strove for equality among the four language streams, and recommended Malay as the national language (Tickoo 1996: 435)

When the People’s Action Party (PAP) government came into power – first

in 1959 with Singapore self-governed, followed by a brief period of unification with Malaysia, and finally independence in 1965 – and, additionally, with the withdrawal of British military in 1967, economic and sociopolitical insecurity necessitated an overhaul of the system (Tickoo 1996: 436) Measures were im-plemented to create national unity and forge national identity and conscious-ness that transcended ethnic boundaries (Chiew 1983: 45ff.) One of these measures was the institutionalisation of English as a compulsory language in schools The economic consideration of this move was that a usable compe-tence in English, the language of science and technology, and of international trade and commerce, was seen as a basic need; the political consideration was that, being a neutral non-native language, not associated with any of the Asian

cultures, and not the mother tongue of any of the ethnic groups, it gave none of the ethnic groups an advantage (Kuo 1980: 59ff.) English education thus be-came the lynchpin of a bilingual policy, and standards in its use began to receive attention at the highest levels of decision-making (Tickoo 1996: 437).

In other language-medium schools then, English was learnt as a second language; in English-medium schools, alongside English as a ‘First Language’, the ‘Second Language’ learnt was a child’s ‘mother tongue’ (Bokhorst-Heng 1998) It should be pointed out that the term ‘mother tongue’3 as used in Sin-gapore policy making does not reflect linguistic reality in many cases, since it

is the practice to assign an official language, either Mandarin, Malay or Tamil,

to a person, depending on the ethnic group they are categorised as belonging

to, respectively Chinese, Malay and Indian (though note too that ‘ethnicity’ is

a vague concept and an arbitrarily assigned designation in Singapore); this is regardless of their actual mother tongue which could be any one of the Chinese

or Indian languages or Malay, or even English The label ‘ethnic mother tongue’ (Pakir 1993: 23–24) or ‘ethnic mother tongue Second Language’ (Foley 1998: 130–131) is thus more appropriate In the Singapore situation then, the ethnic mother tongue for some is often not their mother tongue in terms of origin but

is so in terms of ascription and external identification.4

By the 1980s, enrolment in the other language-medium schools became low, less than 20% The then Permanent Secretary of the Ministry of Educa-tion recommended a national school system with English as the medium of instruction in all schools, with the ‘mother tongue’ taught as a subject, and by

1987, all schools were converted to become English-medium As a result, dren are schooled to become what has been referred to as “English-knowing bilinguals” (Pakir 1992), that is, they have to be proficient in English and their

chil-‘mother tongue’

.2.3 And English in the home and streets

As a consequence of these language policies, by the 1970s English was

recog-nised as having become the primary working language in Singapore, the de

facto national language (Llamzon 1977), used in the government, in commerce

and business, in legislation and the law courts, and in science and technology,

as well as being the main medium of instruction in schools and tertiary

insti-tutions and serving as the lingua franca for international communication and diplomacy It has more recently also become the lingua franca for inter-ethnic

communication, especially among the younger and more educated, particularly

in more formal settings, effectively replacing the ubiquitous Bazaar Malay and Hokkien which performed such service in previous generations (although Mandarin, promoted by the government, is certainly a language of choice for many younger Chinese Singaporeans’ intra-ethnic communication).

A rise in the literacy rate for English has been observed in the population: from 46.7% in 1970 to 55% in 1980 (Tay 1983: 63) to 63% in 1990, and to 71%

in 2000 (Leow 2001) Concurrent with the rise in the level of English literacy, the usage of English has also been reported to have become more prevalent at home While the principal languages most commonly reported to be spoken at home are still, for the respective ethnic groups, Mandarin or a Chinese dialect, Malay and Tamil, there has, in fact, been a shift in all groups towards English

as the predominant home language between 1980, 1990 and 2000: the Chinese from 10% to 19% to 24%, the Malays from 2% to 6% to 8%, and the Indians from 24% to 32% to 36% (Leow 2001; Ng 1995) Overall, while English was the language most frequently spoken at home for 12% of resident households in

1980 (Foley 1998: 130), this rose to 19% in 1990 and continued rising to 23%

in 2000 (Leow 2001)

It is clear too that English is emerging as the language of the young In the

1990 Census of Population, 26% of six-year-old school children claimed lish as the most frequently spoken home language, higher than the national average of 20% (Pakir 1994) In 2000, 36% of Chinese children aged 5–14 years spoke in English, compared to 22% of those aged 15–24 years and 25% aged 25–54 years (Leow 2001)

Eng-Tay (1979), three decades ago, describes six characteristic uses of English

in Singapore, namely, as an official language, a language of education, a ing language, a language of inter- and intra-ethnic communication, a language for the expression of national identity, and an international language To this,

work-a decwork-ade lwork-ater, Bloom (1986: 388) work-adds its uses work-as work-a lwork-anguwork-age of religion, work-and work-as

a home language; in other words, English is not used only in public domains, but also in the more private domains of family and friendship (Platt & Weber 1980) Indeed, by the late 1980s and 1990s, it has been noted that “Singapore is,

in fact, well on the way towards becoming a largely English-speaking country” (Newbrook 1987: 12), certainly one that is English-dominant (Schneider 1999: 193); for many young people, English is the only language spoken confidently There is thus a growing body of English users for whom English has gone be-

yond the lingua franca stage, who are native speakers of the language, following

the simple definition that a native speaker is a fluent speaker of the language, typically after having learned the language as a child (Pakir 1994)

What is to be noted as well is that the English that these Singaporeans are native speakers of is one that has gone through what has been referred to as a process of structural nativisation (Schneider 2003: 265) While the more stand-ard variety of English is the variety taught in schools, alongside it has developed

a variety which has a distinctive phonology, syntax and lexicon, which shows

a high degree of influence from other local languages such as Hokkien, tonese, Malay and Tamil (Platt & Weber 1980: 18) Some scholars believe this colloquial variety to have developed at the English-medium schools, though more in the playgrounds than in the classrooms (Platt & Weber 1980), a contact variety pioneered by children (Gupta 1994: 47) But Mufwene (2001) strongly asserts that children do not make language; adults do; Ansaldo (this volume) will expound on this in the final chapter of this volume

Singapore English has been described in various ways, and it would seem that prescriptive views of SE were the starting point of attempts at describing the language Bloom (1986) identifies three phases

The first concentrated on what was wrong with Singapore English, and search mainly consisted of finding errors of various kinds and trying to account for them Lim (1986) identifies a meeting that Lee Kuan Yew, the then Prime Minister of Singapore, called on 27 February 1979 as a landmark in the pre-scriptive approach to the English language and its use in Singapore The meeting was to discuss the falling standards of written English in the Civil Service The discussions that followed in various ministries as a consequence of this meeting had far-reaching effects, especially within the education system Emphasis was placed on ‘error analysis’, and researchers produced work which was mainly a tabulation of errors However error analysis does not seem to have contributed

re-a grere-at dere-al to the description of Singre-apore English Perhre-aps the fre-ailure lre-ay in the fact that it was an unconscious attempt at getting rid of anything that was

‘Singaporean’, meaning non-standard, in the language It was and possibly still

is part of the uncertainty about status and solidarity that Lim (1986) describes,

in spite of the assumption of Singapore having moved into the phase of normative stabilisation by the 1960s and 1970s (Schneider 2003: 264)

endo-Though its beginnings chronologically pre-date the first stage, the second stage was the debate launched by the publication of Tongue’s (1974) classic, regarded as trailblazing and revolutionary, which proposed that there were

forms of English spoken in Singapore and Malaysia which might be considered standard in their own right What Tongue called ‘standard’ was that variety used

by educated Singaporeans and Malaysians in contexts where a standard form would be considered appropriate It would be the kind of English used by those people born or brought up in the once colonial Malaya who had been educated

in the English medium, where the teachers would have been ‘native’ English speakers, i.e more often than not, Irish priests or nuns in the Catholic schools and American missionaries in the American missionary schools Tongue also identified what he refers to as a ‘substandard’ variety which would be used by Singaporeans and Malaysians, whether they had been educated in the English, Chinese, Malay or Tamil medium, used in informal situations with their peers

or as a lingua franca within the community Tongue seemed to suggest that

a Standard Singapore English exists and the only thing missing to make it a legitimate standard is some form of codification Along similar lines, Crewe (1977a, b), categorising Singapore English as a ‘non-native dialect’, also identi-fied two varieties of English spoken in Singapore, one formal and the other informal The formal one he described as something like a standard but which not everyone could speak proficiently, while the informal variety, which a great number of people knew, was definitely non-standard Crewe saw the school system as playing the major role in any future codification of the language.But probably the most significant studies undertaken at the time were those

of Platt and colleagues (Platt 1975, 1977; Platt & Weber 1980; Platt et al 1983, 1984) It was in particular Platt and Weber’s (1980) description that became

the de facto authority on a number of areas relating to English in Singapore

and Malaysia, leading to such epithets as Lowenberg’s (1984: 114): “[t]he most detailed account of this variation, and therefore the best point of departure for the analysis of this study, is that of Platt and Weber (1980)”

Platt (1977) and Platt and Weber (1980) described the development of English in Singapore and Malaysia in terms of a ‘lectal continuum’ within a

‘post-creole continuum’ The range was seen as extending from the basilect at one end, which it was claimed showed many features of creoles, to the acrolect

at the other, which approximated to Standard (superstrate) English The olect mediated transitionally between the former and the latter The linguistic differences in the usage of socially and educationally differentiated groups of people were represented by the place this usage took on the lectal range How-ever, those people whose usage generally figured at the acrolectal end could also move to the mesolectal or even to basilectal usage While most linguists have recognised that the basilectal is rule-governed, there was a strong feeling

mes-in government bodies that it was unacceptable and even wrong This feelmes-ing was reinforced by the association of the basilect with minimal educated users, meaning that these users only controlled the basilect, so that they were unable

to move along the continuum to mesolectal or acrolectal usage

Later researchers, in particular Tay (1993), Gupta (1994) and Foley et al (1998), identified a number of weaknesses in using the lectal continuum model based on Bickerton’s (1981) theory of the development of Creole The English

in Singapore and Malaysia was claimed then to never have been a creole, as its development did not follow the established patterns identified in the Creole model The alternative was to call it a creoloid (Platt 1975) This may have been even more misleading in that the variety of English found in Singapore and Malaysia was spread through the formal school system Also Platt and Weber did not seem to distinguish between uneducated speech and the informal vari-ety of an educated speaker Thus their different lects were difficult to identify in actual discourse because they lacked sociolinguistic and psychological reality Further, such a classification may be considered obsolete, creole studies having been readdressed and old labels having acquired new readings in recent decades (see Ansaldo this volume)

The third stage was an attempt at describing Singapore English as a variety

in its own right; here we can more clearly see Schneider’s phase 4 in place, viz

the gradual adoption and acceptance of an indigenous linguistic norm, ported by a new, locally rooted linguistic self-confidence (2003: 249) The first step in this direction was the work of Tay and Gupta (1983) Their research set out to define Singapore English mainly in terms of who speaks it, and secondly

sup-in terms of its context of use Tay and Gupta rejected ussup-ing an exonormative standard for their description, opting for what they termed an endonormative one The main reasons given were the functions English serves in Singapore, the multiplicity of foreign standards so that whichever one was chosen would not be backed up by the linguistic environment, and finally the fact that the English-educated Singaporean wanted to sound like a Singaporean What they were attempting to do was summarised in a final paragraph:

Research on Standard Singapore English has at last begun A start has been made to collect examples of non-standardisms but a data bank on deviations from general Standard English needs to be built up By engaging a team of competent researchers, the data can be built up and by testing these deviations with several subjects for their acceptability, a fuller description than what has been presented in this paper will emerge (Tay & Gupta 1983: 188)

Their publication pointed in the direction of establishing a database and a sible codification of what could be Standard Singapore English The terms used

pos-at thpos-at time were often still couched in devipos-ationist language much as one found

in the earlier error analysis type of research But there was emerging another picture, that of a variety of English having an existence in its own right, with subvarieties and its own registers to be used according to the context of culture and situation that is appropriate

Since Bloom’s (1986) original description of the three stages of research into English in Singapore and Malaysia, we have what may be seen as developments within the third stage, involving a move away from the description of the local variety as a post-creole continuum An approach which caught on with and was expanded upon by many researchers in the 1990s was that of considering the situation in Singapore as diglossic, where a single language or variety is seen

to have two subvarieties, a High and a Low, each serving different functions and domains in society (Ferguson 1959) Many proposed that what is seen in the Singapore English speech community is a form of diglossia with Standard Singapore English (SSE) constituting the High variety and Colloquial Singapore English (CSE) the Low variety (Foley et al 1998; Gupta 1994; Richards 1977) If this may be seen as too simplistic and inadequate an account (see Ansaldo this volume), a more complex tridimensional model (Tay 1993) was also proposed, which built on the dimensions of user, usage and use, involving respectively the speaker’s language background, the microlinguistic codes used for multilingual communication, and the factors which make up the context in which English

is used

As has become evident, various characterisations of Singapore English have been made in the past three decades, and it is important to note that over this time, socioeconomic circumstances have changed, as have language policies, and, consequently, inevitably, so has the language itself It is considered timely

to produce a description of the English used by these young Singaporeans who – as outlined in Section 1.2 – are from a generation where being a native speak-

er of (Singapore) English is the norm, where a speaker of English is not from

an elite stratum of society, and is one who uses Singapore English in different situations, ranging from the most formal to the most informal The corpus upon which the description in this volume is based is also a spoken one (cf Biber et

al 1999; Carter & McCarthy 1997); variation from standard usage (if we accept talking about variation from a standard) is after all more marked in spoken English than written English, and it is the former which we are interested in describing here The creation of the Grammar of Spoken Singapore English Corpus (GSSEC) involved an extensive collection, in the period 1998–1999, of naturally-occurring conversational data from native speakers of SE.

.4. Collecting the data

Thirteen undergraduates at the National University of Singapore were equipped with portable audio cassette recorders which they carried on their person and which, whenever they were with friends, or, on a few occasions, with family, they switched on, and the ensuing conversations were recorded While this may sound almost too simplistic, the method actually worked very well These data collectors returned with recordings of naturally-occurring conversation, both dialogues and group interactions, both single-sex and mixed-sex interactions The recordings were made amongst friends or family, and occurred at home,

on campus, in shopping centres, in food courts, even in cars Extraneous noise

in such environments naturally made the transcription of the recordings less than easy, but the high return was in the naturalness of speech

A common concern with such data collection is that of informant modation to the standardising influences of the interviewers or researchers, thus obscuring their vernacular However, these recordings were made by the young Singaporeans themselves, within the social networks of their family and/or their peers, so accommodation is not an issue here And although the interlocutors were aware that they were being recorded, one only has to listen

accom-to the recordings accom-to see that speakers were not caught in an observer’s paradox (Labov 1972a), with speakers sounding completely at ease and uninhibited, and often extremely involved and animated

.4.2 Who are the Singapore English speakers?

The data collectors were deliberately drawn from Chinese, Malay and Indian ethnic groups, and although the majority of data collected by them, as expected, were intra-ethnic interactions, a number of recordings included inter-ethnic interactions as well The corpus is thus satisfying in that it is representative of SE

as spoken by the different ethnic groups in Singapore, and not solely of Chinese

Singapore English, as many studies on SE tend to be Additionally, many of the intra-ethnic interactions included instances of code mixing, which, while not

the focus of this volume, are a rich and invaluable inclusion in any corpus of a variety found in a bi- or multilingual community.

The following provides a general description of the speakers comprising the corpus, with the first three forming the criteria for the speakers to be considered native speakers of Singapore English:

i They are Singaporean, having been born in Singapore and having lived most, if not all, of their lives in Singapore

ii They use English as their main language at home, with friends, at school

or at work; at the same time, most also speak other languages at home, at work, and with friends

iii They have been educated in English as a first language

iv They have educational qualifications ranging from Cambridge GCE ‘A’ level (General Certificate of Education Advanced level) to a bachelor degree at a local university

v They are between 20 and 40 years old

vi They are of Chinese, Malay or Indian origin

.4.3 The resulting corpus

The Grammar of Spoken Singapore English Corpus (GSSEC) can be seen to

be distinctive from other existing corpora of Singapore English on at least two significant counts

First, as mentioned earlier, the speakers in our corpus are young reans from a generation where being a native speaker of (Singapore) English – a variety of English which has undergone endonormative stabilisation – is the norm, where a speaker of English is not from an elite stratum of society, but is one who has had education up to post-secondary or tertiary level, in English, and is one who uses Singapore English in different situations, ranging from the most formal to the most informal This contrasts, in particular, with

Singapo-a published corpus Singapo-as thSingapo-at of PlSingapo-att et Singapo-al (1983) There, the Singapo-authors regSingapo-ard their samples as representing “typical English (SgE [Singapore English]) [as] spoken

by the English-medium educated sector of the community, particularly those who use it in everyday verbal interactions” (1983: 12), apart from two of their spoken texts which they highlight as examples of ‘borderline cases’, the speak-ers having had non-English-medium education However one in fact finds that the speakers had at most English-medium education up to their fourth year in secondary school, with the majority having only three to six years of English-medium education in primary school; further, all speakers do not use English

as a home language at all Although considered the “English-medium educated sector of the community” of two decades ago, their background is markedly different from that of the speakers in our GSSEC corpus Given this then, the Singapore English described in their volume is quite different from that which has undergone endonormative stabilisation and that which is spoken by native speakers If we instead consider Singaporeans with a comparable background from one or two generations back, we find that they would speak a variety that

is more acrolectal (see Ansaldo this volume, note 2) The point is that the lish-medium educated sector of the community” of two decades ago – whether

“Eng-we use Platt et al.’s definition, or whether “Eng-we match speakers by their social indices – is substantially different from now

Second, the corpus recordings are made of spontaneous speech in the speakers’ natural environments, where interlocutors are all Singaporeans and have close network ties with each other, being family or friends In other words,

we have access to the vernacular, which is a most precious asset This again sets our corpus apart from other corpora which may have used contemporary native Singapore English speakers but are composed of less natural data elicited under laboratory conditions (additionally, with a non-Singaporean interlocutor), for example, as in the NIE Corpus of Spoken Singapore English (Deterding & Low 2001) While this certainly affords recordings of greater clarity which, for ex-ample, lend themselves much better to instrumental acoustic phonetic analyses, the speech is certainly not the same as what one would find in a more natural, Singaporean environment

Both these considerations make our GSSEC a valuable and rich source, complementing other existing corpora such as those mentioned above; other qualitative aspects of the data in our corpus are described in Ansaldo (this volume) Comprising 32 recorded extracts, coming to approximately 8 hours of conversation, totalling over 60,000 words, the audio recordings and their orthographic transcriptions making up GSSEC are available for use by other scholars

.4.4 Transcription

Recordings were orthographically transcribed by a number of undergraduate and postgraduate students who were provided with written and verbal in-structions for transcription conventions and whose initial transcriptions were checked and verified by a member of the project team These transcriptions

were then double-checked by a number of English Language Honours and postgraduate students.

Phonetic transcriptions were made by the trained phonetician of the project team, as well as by three students (two Master’s students and one English Lan-guage Honours graduate) who had all previously received training in phonetic transcription and were deemed competent in it Students’ transcriptions were checked by the phonetician

This first chapter has acquainted the reader with various sociolinguistic and historical aspects which are vital to understanding, appreciating and placing the study in its context The history of scholarship in the area of Singapore English is also outlined, noting the kind of shifts in emphasis that SE studies have undergone through the decades, from an error analysis approach to re-garding SE as non-standard, to the more recent descriptions of SE as a variety

in its own right The background to the corpus of SE on which the structural descriptions in the rest of the volume is based is also presented, and its signifi-cance highlighted

The next four chapters take up the more significant structural aspects of CSE, describing patterns in detail The description in this volume is meant as a coherent one, in that the evidence on which its various structural descriptions are based is kept consistent, drawing upon the corpus for its data In addition

to the corpus data, the authors also at times supplement their observations with examples or findings from other sources; when this occurs, it is explicitly stated where these derive

In Chapter 2, Lisa Lim describes the sound system of CSE and the logical processes that occur, addressing both segmental and suprasegmental aspects The latter, often only given passing mention in most descriptions, is described in detail: this is significant as it has often been said that it is the stress, rhythm and intonation which are extremely characteristic of SE In Chapter 3, Lionel Wee and Umberto Ansaldo provide an account of the noun phrase in CSE, highlighting, in particular, characteristic CSE features involving number

phono-and subject-verb agreement, phono-and the use of one as the head of a nominal in

various syntactic environments In Chapter 4, Vivienne Fong describes the grammatical patterns in the verb and auxiliary cluster in CSE, focusing on main clauses, and highlighting structures characteristic of CSE, including zero

copula, verb reduplication and passive constructions Lionel Wee in Chapter 5 looks at the two more salient semantic and pragmatic features of CSE, namely, reduplication and discourse particles, which have often been held up as symbols

of – what is ‘bad’ about – CSE

In the final chapter, Chapter 6, Umberto Ansaldo looks at Singapore lish from a diachronic aspect, tracing the origins of this variety and identifying its core ingredients by looking at the original ethnic groups that were responsi-ble for its very emergence The structural observations presented in the preced-ing chapters are thus contextualised here, being related to the socio-historical and cultural background in which SE evolved, as well as to creolisation studies

Eng-in general

As will become evident in the reading of this volume, all the chapters which provide a structural description do so in a consistent manner: they are all highly descriptive and informative, are systematic in coverage of structures, and are rich in convincing examples It is nonetheless inevitable, since the different chapters are contributions by different individuals, that the authors, coming from different linguistic backgrounds, should naturally also draw upon their own training and schools of thought

Such a combination of various methodological approaches in a somewhat eclectic manner is, we feel, making the best of all worlds Fong, for example,

in her chapter on the verb cluster, supplements her observations based on the corpus data with native speaker judgements This is in fact also seen as a con-sequence of the area of study: with the verb phrase where, it is believed (see Fong this volume, but also cf Wee & Ansaldo this volume), there is often code switching between dialects of the more Standard Singapore English and col-loquial Singapore English, Fong feels that she has to develop other diagnostics

to help establish if the features she highlights are indeed those characteristic of the colloquial variety On the other hand, a chapter like Wee’s on reduplication and particles addresses what one might refer to as ‘core features’ of colloquial Singapore English, and there is thus no need to double-check the ‘validity’ of the features as CSE

In engaging with the data, the different authors at times propose slightly different analyses for structures, for example, Wee vs Ansaldo on the source of reduplication patterns Again, this is not felt to make the volume any less coher-ent; on the contrary, the reader’s understanding of the complex issues and even more complex possibilities of interpretation can only be enriched by a range of well-argued and solidly-supported views

As mentioned in Section 1.4 above, the data on which the descriptions are based are those of naturally-occurring spontaneous speech of native speakers

of Singapore English in informal situations To use a term commonly found

in recent scholarship, we will be largely referring to the variety as colloquial Singapore English (CSE) or occasionally simply as Singapore English (SE), and

by this we mean the colloquial variety of SE used by speaking, educated Singaporeans.6 Some scholars (in this volume, Ansaldo) prefer to refer

English-to this variety of English found in Singapore as mesolectal Singapore English (or mesolectal Singlish); the acrolectal/mesolectal/basilectal classification is so preferred as it takes into account a greater range of varieties and more obviously recognises a cline than the SSE/CSE approach does For our purposes in this volume, these two labels, CSE and mesolectal SE, may be seen as interchange-able and encompass the same variety

What is termed basilectal SE is also common in Singaporeans and is used

by less highly-(English-)educated groups, and, in fact, quite possibly also by our young English-educated group, as another variety of SE in which they operate that is lower along the lectal continuum (see Ansaldo this volume for elaboration) It is a relevant point to be made, because variation within the

corpus is certainly noted (see Lim; Wee & Ansaldo this volume), and should be

noted – linguistic homogeneisation is after all not a feature of SE (Schneider 2003: 266)

While there have been a number of recent publications of language use in Singapore (e.g Foley et al 1998; Ooi 2001; Ooi et al 1997), no single volume provides a comprehensive grammatical description of a corpus of Singapore English in combination with an insightful diachronic perspective This descrip-tion goes beyond older accounts, such as Platt and Weber’s (1980) description,

to name but one, which, in categorising SE into acrolect, mesolect and basilect, according to speakers’ socioeconomic status, is an account which while valid two decades ago does not have quite the same contemporary reality Taken together, such descriptions clearly illustrate significant shifts in language use from a longitudinal perspective This description also departs in the extent and depth of structural analysis from previous accounts that mostly focus on issues

of sociolinguistic nature

The value of this grammar for linguists is obvious, from theoretical linguists interested in the linguistic patterns highlighted, to variationists, typologists and creolists interested in the features characterising this particular English variety and the possible sources of such features, to people interested in English as a global language: all will find this description useful and relevant.

The implications of such a grammatical description for pedagogy (cf Greenbaum & Quirk 1990) are also obvious: a more effective language peda-gogy is possible only if attention is given to the use of language in spontaneous speech, and, consequently, to the need to develop descriptive tools for analysing spoken language

This volume is also particularly significant in the context of Singapore’s language policies, the most recent being the Speak Good English Movement (SGEM) whose goal is to raise the level of consciousness of the public to speak-ing a more standard English instead of a ‘bad’, ‘ungrammatical’ Singlish (see endnote 1; for discussions on Singlish, see Fong et al 2002; Rubdy 2001) This volume in fact illustrates that Singapore English does have its rules and gram-mar, and this can be an insightful and instructive example for other ‘non-stand-ard’ varieties of English It also suggests that the evolution of such a variety is a perfectly natural linguistic phenomenon which best reflects the multiethnic and multilingual society that Singapore is and has been for the past two centuries.Indeed, although this volume does involve analyses at a fairly technical and sophisticated level, the grammatical descriptions herein should still be acces-sible to the layperson and to teachers Indeed it is hoped that individuals in the government and education will read this and, in so doing, realise that colloquial Singapore English is not a deviant, substandard variety of English, but rather that it does have its own rules and regulations, and is a viable, indeed crucial, variety of English for Singaporeans

Notes

 We have deliberately avoided the use of the term ‘Singlish’ in this volume, even

though it has more common, more popular currency, and is surely more pithy than the more cumbersome ‘Singapore English’ This is because the term is used in

at least two different ways, viz either (i) to colloquial Singapore English (CSE), the

interpretation found in current academic discussions; or (ii) to the cal’, ‘non-standard’ variety of Singapore English – this more common, derogatory interpretation is found in most public and official (government) discussions For

‘ungrammati-further discussion of this, see Fong et al (2002) It should be pointed out that the term is nonetheless preferred by some scholars, e.g Ansaldo (in prep a); Gil (1994, 2003).

2 Though it is observed that some variety of English was already existent prior

to the arrival of the British in 1819; see Ansaldo et al (in prep)

3 The idea of ‘mother tongue’ is essentially a construction or an artefact that has

grown up in monolingual societies, where linguistic focusing is much stronger than in multilingual communities, where it is more diffused The identification and operation with one language is therefore not so clear in a multilingual com-munity as it is in a monolingual country

4 Skutnabb-Kangas and Phillipson (1989: 452–453) define ‘mother tongue’

ac-cording to the criteria of origin, competence, function, or internal or external identification, and state that (a) a person can have several mother tongues, (b) the same person can have different mother tongues, depending on which definition

is used, and (c) a person’s mother tongue can change during one’s lifetime, even several times, leaving out the criterion of origin

5 The Grammar of Spoken Singapore English Corpus (GSSEC) was collected

for the project Towards a Reference Grammar of Singapore English (Lim 2001),

which was funded by a National University of Singapore Academic Research Grant R-103-000-003-112 (Joseph A Foley and Lisa Lim, Principal Investigators; Vivi-enne Fong, Ni Yi-Bin and Lionel Wee, project members) Researchers are welcome

to use the corpus, as long as they acknowledge the source

6 The classification of Standard Singapore English (SSE) and Colloquial Singapore

English (CSE), as pointed out in Section 1.3, really only takes into account usage in proficient English speakers in Singapore

It should be noted at the outset that there is a range of variation observed

in GSSEC of two kinds First, some speakers show realisations which are closer

to a more acrolectal or SSE pronunciation than others Second, features ticular to certain ethnic groups are also observable Neither of the above types

par-of variation is problematic, as it is in fact quite common to find features typical

of different levels of a lectal continuum co-existing in one and the same variety (see also Wee & Ansaldo this volume) as well as ethnic features in multiethnic communities While instances of both will be mentioned where relevant, the more stable and characteristic pan-Singaporean features which all Singaporeans

do share is what is central to the description here

The chapter will in addition suggest possible reasons for the features observed in CSE The scope of this chapter does not afford great depth and detail in addressing this issue, but a quick survey will be conducted of some

of the phonological systems of the more common local languages spoken by

the majority of Singaporeans, viz Mandarin, Hokkien, Cantonese, Malay and

Tamil, to be able to suggest possible substrate influences The rationale for the choice of five of the many Asian languages spoken by the Singapore popula-tion to account for substratal influence is as follows Mandarin, because of its

official institutionalisation through language policies (see Lim & Foley this volume), has now become the language most frequently spoken at home for the Chinese (Leow 2001: 98) However, as the vast majority of the ethnic Chinese Singaporeans are descendants of immigrants from Southern China, mainly the provinces of Fujian and Guangdong, Hokkien and Cantonese are still dominant home languages for them (third and fourth, after Mandarin and English) In fact, Hokkien was until recently the most frequently understood and spoken Chinese language, followed by Cantonese and Mandarin, particularly in the 1970s (Lock 1982: 302) when Singapore English was developing Malay is

by far the most frequently spoken language at home for the ethnic Malays; and for Indians, it is Tamil, although English is a very close second for them (Leow 2001: 98)

2.2. Inventory

Perhaps the clearest way to examine the vowel phonemes of CSE is to present them in Standard Lexical Sets (see Wells 1982), as in Table 2.1 Standard South-ern British English (SSBE) and General American (GA) are included as the two more widely-documented English varieties, and are useful as models of com-parison We also include Standard Singapore English (SSE)4 for comparison, though our discussion will primarily focus on CSE

It can be clearly seen that many of the vowel oppositions found in SSBE and

in GA, as well as in SSE, are absent in CSE.5

In Sections 2.2.2 and 2.2.3, we go on to look in greater detail at the tions and distributions of CSE monophthongs and diphthongs, and in 2.2.5

realisa-we briefly examine the contact influence which accounts for the system and realisation of vowels

2.2.2 Monophthong realisation and distribution

Table 2.2 places the CSE vowels in a traditional vowel quadrilateral to provide

an indication of their actual quality

The contrast between tense-lax vowel pairs, as seen in varieties such as SSBE and GA, tends to be neutralised in CSE, with the distinction both in quality and quantity being reduced, such that the following pairs of words are

produced as virtual homophones: beat and bit as [bit], cart and cut as [kat],

Table 2.1 Standard Lexical Sets for SSBE, GA, SSE and CSE.

I6 7

"˜

f˜

f˜

~6 i 6 6

i 7 7 f a u a f 6 i e a f o u ai fi au i6 7 a f f u6 i 6 6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

KIT DRESS TRAP LOT STRUT FOOT BATH CLOTH NURSE FLEECE FACE PALM THOUGHT GOAT GOOSE PRICE CHOICE MOUTH NEAR SQUARE START NORTH FORCE POOR/CURE 6

HAPPY LETTER COMMA

Table 2.2 Monophthongs of CSE.

a

u o f

caught and cot as [kft], and pool and pull as [pul] In SSE which still appears

to have these tense-lax pairs as separate phonemes, the quality of the tense vowels is less in the extremities of vowel space, such that, for example, [i] is not as front and high in quality as that in SSBE, and so the quality of the two vowels [i] and [I] are further apart in SSBE, closer in SSE, and significantly indistinguishable in CSE (Lim in prep a)

In CSE, then, the kit and fleece sets have the same vowel, represented by /i/, as illustrated by these examples from GSSEC:

(2) titëw teach cf SSBE ti˜tëw

The vowel in the strut, palm and start sets is the same in CSE As indicated

in Table 2.2, this vowel in CSE, represented by the symbol /a/, is low and central: lower in quality than the strut vowel in SSBE, and certainly much less low and back than the palm and start vowel in SSBE

(3) laki lucky cf SSBE l%ki

(4) as ask cf SSBE "˜sk

The lot, cloth, north and thought sets all have the same vowel in CSE CSE /f/ is much less low and back than the lot/cloth vowel in SSBE, and while similar in vowel height, has a quality that is less back than the north/thought vowel in SSBE While it has rounding to a lesser degree than that found in the vowels in SSBE, this vowel is not unrounded like the lot vowel in GA

(5) nft not cf SSBE n#t, GA n"˜t

(6) sfs sauce cf SSBE sf˜s

The vowel in the nurse, letter and comma sets is the same in CSE, produced with minimal tenseness and duration, with no length distinction as that found between the vowels in nurse and letter in SSBE or GA This is particularly evident in example (7)

(9) sw7t sweat cf SSBE sw7t

(10) p7mfl6t pamphlet cf SSBE pæmfl6t

A small number of lexical items, such as extra and elegant, do have a particular

realisation where what would normally be a vowel of the dress set is nounced with a much more open realisation, as [ækstpa] and [æl6:6nt] in CSE These may be considered instances of hypercorrection

pro-There is however also a regular realisational difference between these two sets The trap vowel /æ/ certainly has a less open realisation as [7], compared

to SSBE, in all environments However, while the dress vowel /7/ is realised as [7] in most environments, it is realised as [e] when followed by a voiced plosive,

suggesting that the two phonemes have not conflated So while bat and bad are respectively [b7t] and [b7d] with same vowel quality, we find bet realised

as [b7t], while bed is realised as [bed] We find instances of this in GSSEC, in

examples (11) and (12):

(12) flpedi already cf SSBE 6lp7di

This pattern also results in the neutralisation of an otherwise consistent trast As the vowel in the face sets is regularly a monophthong, compared to the diphthong in SSBE and GA (see 2.2.4), in the context of preceding voiced plosives then, the contrast between /7/ of the dress set and /e/ of the face set

con-is suspended, leading to bed and bade being homophonous in CSE.

Finally, there is a tendency for full vowel quality to be maintained, which in

general appears to conform to orthography, particularly orthographic o, with

no reduction to schwa even in unstressed position:

(13) kfp7k correct cf SSBE k6Áp7kt

(14) poziw6n position cf SSBE p6ÁzIw6n

(15) kfmfft comfort cf SSBE Ák%mf6t

(16) l7]:wetëw language cf SSBE Álæ]:wIdëŠ

2.2.3 Diphthongs, and triphthongs?

As indicated in Table 2.1, the vowels in the face, goat and square sets, which are diphthongs /eI/, /o~/ and /76/ in SSBE, are regularly reduced to a more monophthongal quality of the sort closer to [e], [o] and [7] respectively in much

of GSSEC, illustrated in the following examples

(17) l7s mekap less makeup cf SSBE meIk%p

(18) don(t) don’t cf SSBE do~nt/d6~nt

(19) kfmp7d compared cf SSBE k6mp76d

In other work which examines data in different formality modes (Lee & Lim 2000), it is found that, while these three diphthongs undergo monophthongi-sation in more informal contexts such as in spontaneous picture description, diphthongal quality is still maintained in more formal or structured modes of discourse, such as in carrier frames and read passages

The price vowel in CSE, as seen in Table 2.1, is a diphthong not dissimilar

to that in SSBE or GA Certain lexical items however have [a] instead; this seems to be in syllables whose coda is [l]

(20) wal while cf SSBE waIl

Diphthongs found in other accents such as SSBE are falling diphthongs, that is, the first element of the diphthong is more prominent than the second (Crut-tenden 2001: 129) The centring diphthongs of the near and poor/cure sets

in CSE tend instead to be rising diphthongs, that is, the second element of the diphthong is the more prominent one, and the first element is realised as a glide,7 as in (21) and (22) In contexts where the poor/cure diphthong is pre-ceded by /j/, as in (23), the diphthong is completely monophthongised to [f]

CSE does not have any triphthongs – and whether other English varieties have them as well is arguable (cf Cruttenden 2001; Roach 2000) Syllables that are considered, at least by some, as triphthongs in SSBE, are in CSE produced as two syllables, as in (24) and (25) This resyllabification seems the preferred process in CSE, rather than, as in SSBE, having the triphthongs smoothed and reduced to a diphthongal glide by losing the middle element, or reduced even further to a monophthong

(24) aw6 hour cf SSBE a~6 > "˜6 > "˜

(25) taj6d tired cf SSBE t"I6d > t"˜6d > t"˜d

2.2.4 Rhotacisation

Rhotacised vowels are sometimes, though not often, observed in the GSSEC corpus, as in (26) They are, however, documented in SE in more formal situa-tions (Lim in prep a) and in public domains, such as in the speech of DJs on local radio stations (Schneider 1999) In other work involving elicited material and perceptual tests, rhotacisation is found in the speech of Singaporeans judged in identification tests to be Malay-sounding, but not in those judged as non-Ma-lay-sounding (Lau 2002) All the above suggest that this is a variable which is still very much in flux, possibly largely to the more recent and increasing status

of GA, which does retain such post-vocalic (both consonantal and pausal) usage of /r/, compared with non-rhotic SSBE (Cruttenden 2001: 208) (26) pft port

pre-2.2.5 Vowel substrate influence?

Tables 2.3 to 2.7 sketch the vowel inventories of the five local languages that the majority of speakers of the three main ethnic groups in Singapore speak, the rationale for this having been presented in the introduction of this chapter

As there is no good description of the system of the Singaporean varieties of these languages, these are based on Bodman 1955; Chiang 1940; Deterding & Poedjosoedarmo 1998; Killingley 1993; Lim 1988; Lock 1982

There are, of course, obvious differences between the vowel systems of the languages: nasalisation is phonemic in Hokkien, while in Cantonese and Mandarin, vowel rounding is phonemic, and Tamil has phonemic vowel length However, we should also note two things which are relevant: (i) as far

as vowel height and backness are concerned, all the languages depicted have

no more than a 6-way distinction between vowels; and (ii) none of them have the distinctions in terms of quality that are found in varieties such as SSBE and

GA where vowels are differentiated in terms of the extent of vowel height and

front-backness, as with /I/ and /i˜/, /~/ and /u˜/, and /#/ and /f˜/, for example Although no one-to-one correspondence between each of these languages and CSE can be established, it is obvious that the vowel inventory of CSE, as seen

in Table 2.2, is much closer to those of the local languages than it is to the vowel inventory of SSBE

20-Where diphthongs are concerned, we note that, while both Malay and Tamil do not have diphthongs, the diphthong inventories of both Cantonese (Killingley 1993: 4–5) and Mandarin do include /ei/ and /ou/, and one of

Table 2.3 Monophthongs of Hokkien.

Table 2.4 Monophthongs of Cantonese.

Table 2.5 Monophthongs of (Singapore) Mandarin.

Table 2.6 Monophthongs of Malay.

Table 2.7 Monophthongs of Tamil.