вход по аккаунту


[Varieties of English Around the World General Series] David Blair Peter Collins - English in Australia (2001 John Benjamins Publishing Co)

код для вставкиСкачать
Varieties of English Around the World
General Editor:
Edgar W. Schneider
Department of English & American Studies
University of Regensburg
Universitätsstraße 31
Editorial Assistants:
Alexander Kautzsch, Andreas Hiltscher, Magnus Huber (Regensburg)
Editorial Board:
Michael Aceto (Puerto Rico); Laurie Bauer (Wellington)
J.K. Chambers (Toronto); Jenny Cheshire (London)
Manfred Görlach (Cologne); Barbara Horvath (Sydney)
Jeffrey Kallen (Dublin); Thiru Kandiah (Colombo)
Vivian de Klerk (Grahamstown, South Africa)
William A. Kretzschmar, Jr. (Athens, GA)
Caroline Macafee (Aberdeen); Michael Montgomery (Columbia, SC)
Peter Mühlhäusler (Adelaide); Peter L. Patrick (Colchester)
Volume 26
David Blair and Peter Collins (eds.)
English in Australia
Edited by
Macquarie University
University of NSW
The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for
Printed Library Materials, ANSI Z39.48-1984.
Cover illustration
Cover photograph by Mirjam Medema.
Library of Congress Cataloging-in-Publication Data
English in Australia / edited by David Blair and Peter Collins.
cm. -- (Varieties of English around the world. General series, ISSN 0172-7362 ; v.
Includes bibliographical references and index.
1. English language--Australia. 2. English language--Variation--Australia. 3.
Australianisms. I. Blair, David, 1942-. II. Collins, Peter, 1950-. III. Varieties of English
around the world. General series ; v. 26.
ISBN 90 272 4884 2 (Eur.) / 1 55619 729 2 (US) (alk. paper)
© 2001 – John Benjamins B.V.
No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other
means, without written permission from the publisher.
John Benjamins Publishing Co. • P.O.Box 36224 • 1020 ME Amsterdam • The Netherlands
John Benjamins North America • P.O.Box 27519 • Philadelphia, PA 19118-0519 • USA
Language and identity in Australia
Peter Collins and David Blair
Section A: English in Australia: structure
Vowel change: synchronic and diachronic evidence
Felicity Cox and Sallyanne Palethorpe
Variation and change in Australian consonants: reduction of /t/
Laura Tollfree
The vocalisation of dark l in Australian English
Toni Borowsky
Hypocoristics of place-names
Jane Simpson
Syntactic features and norms in Australian English
Mark Newbrook
Australian English and indigenous voices
Bruce Moore
Australian English — an identity crisis
Susan Butler
Corpus evidence on Australian style and usage
Pam Peters
Section B: English in Australia: variation
Torres Strait and Aboriginal
Torres Strait English
Anna Shnukal
Aboriginal English: adopted code of a surviving culture
Ian Malcolm
Ethnic varieties
Ethnic varieties of Australian English
Michael Clyne, Edina Eisikovits and Laura Tollfree
Australian English and recent migrant groups
Scott Kiesling
Diachronic and generational variation
The acquisition of colloquialisms by non-native speakers
Jane Curtain
Changing attitudes to Australian English
David Bradley and Maya Bradley
A.G. Mitchell and the development of Australian pronunciation
Colin Yallop
Lexicography and national identity: the Australian experience
Arthur Delbridge
Australian English in interaction with other Englishes
Brian Taylor
Regional variation
Short A in Australian English: a geolinguistic study
Barbara M. Horvath and Ronald J. Horvath
Language and identity in Australia
Peter Collins and David Blair
Australians seem to have a perennial fascination with the question of national
identity. In the minds of most citizens (at least, as represented by those social
surveyors and media commentators who report the national mindset), Australia
is still undergoing a process of establishing a clear self-image and a national
sense of purpose. In contrast, there is considerable evidence that Australia’s
linguistic identity was established early in the history of the colony.
According to the late A. G. Mitchell, the founder of Australian English
(henceforth, “AusE”) studies, it is likely that the colony of New South Wales
saw the formation of a variety approximating what we now know as Broad
Australian as early as the 1830s (see Yallop, this volume). It is generally
accepted that the origins of AusE lay in the dialects spoken in the south-east of
England in the late 18th century. Language historians may argue over whether
those dialects were mixed in the London area and then transported to Australia
as a new amalgam, or whether the dialect contact processes that created AusE
took place on Australian soil (Blair 1975; Trudgill 1986); but Mitchell was
almost certainly correct in his view that the process of dialect levelling
produced an identifiable variety within the first 50 years.
Early printed sources give a strong indication that this was so. James
Dixon wrote in 1822 of the “amalgamation of such various dialects assembled
together” which had already produced “a better language, purer, more harmonious, than is generally the case in most parts of England” (Dixon 1822: 46).
George Bennett visited NSW twice, in 1829 and in 1832, before settling in
Sydney in 1836. He noted that “the English spoken is very pure” and that, as in
the United States of America, it was not “corrupted by so many different
provincial dialects”; the native-born could be clearly distinguished from those
who had emigrated from England, by this characteristic alone (Bennett 1834:
I, 331).
Louisa Meredith (1844: 50) had a very different assessment of the local
…a very large proportion of both male and female snuffle dreadfully; just the
same nasal twang as many Americans have. In some cases English parents
have come out here with English-born children; these all speak clearly and
well, and continue to do so, whilst those born after the parents arrive in the
colony have the detestable snuffle. This is an enigma which passes my
sagacity to solve.
Despite the different appreciation, she was at one with both Dixon and
Bennett in noting the clear distinction between the native-born and the English-born pronunciation.
Three commentators, Peter Cunningham (1827: II, 60), Samuel Mossman
(1852: 19) and Friedrich Gerstaecker (1853: II, 269), reported differently.
They identified the dialect of NSW as Cockney. It is likely that all three, as
Cunningham explicitly acknowledged, were responding to the accents of adult
emigrants: Gerstaecker noted the presence of “broad Irish brogue” as well.
But it should not surprise us that there was a strong London component in the
original version of AusE, even though Australians have been reluctant to
acknowledge the similarities of their pronunciation, in some respects, to that
of working-class London. If Mitchell was right, that the original Broad AusE
was formed in the first 50 years, it is inconceivable that it did not carry a set of
Cockney-related features.
Those features, of course, have been somewhat attenuated over the years,
and AusE progressively expanded its repertoire of variation to include RP-like
varieties. (The dialect is now usually described as having several sociolects, in
an accent continuum which leads from “Broad Australian” at one end to a
more RP-like “Cultivated Australian” at the other; the middle ground is held
by the variety named “General Australian”.) In addition, Australia’s multiethnic society has recently generated varieties which are associated with
community groups of various non-English migrant backgrounds (Clyne,
Eisikovits and Tollfree, this volume); and together with the English of Aboriginal communities, these form a distinct subset of Englishes in Australia.
As a result of this, it is now common for linguists to distinguish AusE (or
“Anglo-English”, the unmarked form of English in Australia) in its several
varieties, from Aboriginal English and the other Englishes of Australia.
Attitude and identity
Many sociolinguists have argued that language functions as a badge of social
identity. (See Fasold (1984) as a typical example.) It is certainly true that the
many identifiable social groups in Australia are marked by, and can to some
extent be identified by, variation in their language use. Language is part of the
social dynamic, and may undergo considerable modification by individual
speakers and by groups, from context to context. The slightest difference in
language may be detected by listeners and perceived to have social significance (Giles and St Clair 1979: 17).
This being so, we might expect changes in language and in social identity
to go hand-in-hand. For example, Blair (1993) notes the linguistic consequences of the shift in Australian national consciousness ushered in by the
election of the a Labor government led by E.G. Whitlam in 1972. Corresponding to the break with the British-oriented outlook of the previous long-serving
Menzies government and heightened awareness of Australian cultural icons,
claims Blair, there was a greater acceptance of the Broad Australian accent. In
fact a number of ministers in the new Whitlam government spoke with a
Broad accent. It was during this time that aggressively local “Ocker”1 TV
commercials became highly popular, and there was a resurgence of Australian-produced films portraying an overtly Australian culture and lifestyle.
According to Delbridge (this volume: 310) it was not until the 1940s that
a positive attitude towards AusE began to develop, the only earlier sign being
a temporary surge in national feeling towards the end of the nineteenth
century, “a small manifestation of interest, even pride, in some of the words
and phrases of Australian idiom…”
What is happening today? Are Australians over their ‘identity-crisis’ of
the 1970’s, when they began to lose confidence in the Mother Country, and
began to question their traditional emotional, political, cultural, and economic
Arguably, the most transparent reflection of speakers’ attitudes, values and
self-perception is to be found in the lexicon. The Australian vocabulary
embodies the ideals that Australians cherish, those of egalitarianism and anti-
authoritarianism, sympathy for the battler and desire for a fair go. The
fondness with which words like mateship have been preserved and the benign
regard for archaisms like cobber illustrate the close relationship between
words and the culture they reflect.
The most significant lexical development in recent decades has been the
influx of words and expressions from American English (“AmE”), reflecting
the increasing influence of American culture in Australia since World War II.
Rapid developments in communication such as the Internet are bringing AusE
closer than ever before to the “now” of AmE.
Taylor (this volume: 334–5) comments on the changing nature of the
exposure to AmE, as “the increasing speed of technological development
brings the inhabitants of the global village even closer together”, and as
younger Australians “communing day in, day out for hours on end with their
AmE-speaking computers are being interactively exposed to that variety in a
way earlier generations never were with passive TV watching”.
Opinions differ as to the extent and impact of American influence on
AusE. Members of the public and journalists constantly bemoan the debasement of AusE through Americanisation (see Taylor, this volume) suggesting
or at least implying that AusE is merely a passive receptacle for Americanisms. Butler (this volume) speaks of an “identity crisis” for AusE, a nervousness in self-image prompted by the feeling that AmE is “taking over”.
Sussex’s studies of radio, film and TV data led him to the conclusion that
Australians “still exhibit the classic signs of cultural insecurity” (1989: 167)
and “don’t seem to place a great deal of emphasis on their variety as part of the
national culture” (1995: 22).
Despite talking of Australians’ linguistic nervousness, Butler (this volume) observes the operation of a “filtering process” in AusE that blocks the
entry of items from AmE that have no interest or relevance to Australia (e.g.
advanced television, Anita-Hill, attack-fax). In an earlier article she noted
(Butler 1996) that some AmE borrowings, such as bushranger and phoney
have become so integrated into AusE that speakers are no longer aware that
they are AmE in origin. Peters (1993: 25) also argues that Australian borrowings from AmE are selective, and that show “no sign of going all the way with
the USA.” Corpus evidence adduced by Peters indicates a preference, for
example, for past tense burnt over the form favoured in AmE, burned.
Taylor’s (1989) research shows that AmE influence on AusE has by no
means been limited to the lexical level. Phonologically, Taylor notes, there
has been a tendency for the stress patterns in certain words to move from a
traditional British to an American pattern (e.g. finANCE to FINance,
reSEARCH to REsearch). Graphologically, simplification of digraphs such as
<ae> and <oe> as in medieval and fetal follows American practice. Syntactically, Taylor notes, amongst other things, the American-influenced elision of
the in structures of the type I play (the) piano.
Let us examine several further lexical trends in AusE and their possible
explanation in terms of Australians’ self-perception and identity. According to
Moore (this volume) there has been an influx of words from Aboriginal
culture since the 1960s which parallels the development of Aboriginal political and cultural activism and a growing interest in Aboriginal languages and
culture amongst white Australians. Examples are native title and Mabo, which
entered AusE following the High Court decision in 1992 to recognise the
claim by Koiki Mabo, a Mer islander from the Torres Strait, that his people’s
land had been illegally annexed by Queensland. A number of Aboriginal place
names have risen to prominence in recent years beside their European counterparts, the most well known being Uluru (for Ayers Rock).
Taylor (1989, and this volume) identifies a revival of British influence
from the 1970s, through television series such as Till Death Us Do Part,
Steptoe and Son, Heartbeat, A Touch of Frost, and The Bill, resulting in the
importation of terms such as telly (‘television’), loo (‘toilet’), knickers
(‘women’s underpants’), and cheers (‘goodbye’). The lexical effects of other
foreign influences in Australia have been more semantically restricted. Bearing testimony to the multicultural identity that Australia has gradually developed since the commencement of large-scale immigration at the end of World
War II are the many terms for food and drink introduced by migrants from a
large variety of language backgrounds. Terms such as cappuccino, goulash,
souvlaki and hummus have, as Clyne, Eisikovits and Tollfree (this volume)
note, “become a tangible indicator of multiculturalism”.
Finally, as Seal (1999: 235) has observed in his lively account of the
“Lingo”, the continuing vitality and creativity of the colloquial vernacular
wordstock is a constant reminder of how directly the lexicon enshrines the
deep-seated beliefs and ideals of its speakers. The most ardent efforts of
political correctness advocates have not succeeded in curtailing the intrinsically “incorrect” elements in the Australian lexicon. Alongside trends suggesting a new pluralism in Australia’s identity there persists a rich array of
terms suggesting intolerance of those who are somehow different: these
include racist expressions such as wog, chink, pom, yank and more recently
ethno. Many Australians today — particularly males — retain a penchant for
the coarse and irreverent, in some cases perpetuating “flash” expressions
originating in early nineteenth century prison contexts (e.g. stink (‘uproar’),
bludger (‘lazy person), sort out (‘fight’)). Australians continue to display
considerable colloquial creativity, building up large sets of expressions using
a single stem: scared shitless (‘very scared’), shit a brick! (expression of
surprise), up shit creek (‘in a difficult predicament’), built like a brick
shithouse (‘strongly built’), bullshit artist (‘one who tells lies’), shit-faced
Syntax and morphology
Typically the dialects of a language differ least at the syntactic level, so it is
significant — and perhaps suggestive of an attempt to find a separate identity
— that some syntactic phenomena have emerged in AusE that are different
from those of both BrE and AmE. According to Newbrook (this volume)
AusE allows both singular and plural concord with the names of sports teams
and other singular collective proper nouns (e.g. North Melbourne is/are
playing well), whereas plural concord is favoured in BrE, and singular concord in AmE. Newbrook also claims that some Australian teachers propagate
a reversed form of the traditional rule governing the use of commas around
relative clauses, resulting in the following types of pattern: Joanne and Jane
who had finished left the hall; Any students, who have finished, may leave the
hall. Peters’ corpus studies (this volume) indicate that contractions (it’s, don’t,
there’s, etc.) are more widely distributed and more frequent in non-fictional
genres in AusE than in BrE and AmE.
Some examples restricted to the syntax of nonstandard spoken AusE and
not attested in nonstandard BrE or AmE, are also adduced by Newbrook (this
volume), including the use of genitive ’s with relative that as in This is the girl
that’s book I borrowed.
Studies of AusE morphology have not been common. Apart from Dabke
(1976) and Taylor (1989) — the latter being restricted to morphological
transfers from other languages and dialects — such work has largely concentrated on nominal modification such as nicknames, diminutives and abbreviations. (See Dermody 1980; McAndrew 1992; Poynton 1984, 1989.) Simpson
(this volume) treats these alternative forms of words and names (“hypocoristics”), and shows that the AusE pattern entails fitting the full form to a one- or
two-syllable word template which is stressed on the first syllable. What is less
certain is whether there are meaning differences implied in the choice of
different hypocoristics; this is possibly so for common nouns, less apparent for
toponyms. Simpson shows that there are regional differences in type of usage
and in frequency; but, as with all regional variation within AusE, much more
research needs to be done.
What can we learn about Australian identity from recent developments in the
phonology? Horvath (1985), in her survey of English variation in Sydney,
found a pattern suggesting the occurrence of a change from the vernacular
Broad and prestigious Cultivated speech varieties, towards the in-between
General. Some support for her claims is available from Cox’s (1998) reanalysis of Bernard’s acoustic data. Cox’s analysis of apparent time differences in
Bernard’s subjects suggests a movement away from Broad with /u/ (towards
reduced ongliding) and with /oυ/ (towards a more retracted first target), and a
movement away from Cultivated with /ə/ (towards monophthongisation).
These trends suggest that Australians may be growing increasingly dissatisfied with their traditional British ties, these being suggested by the often
noted closeness of Cultivated Australian to the British Received Pronunciation. Furthermore the increasingly urbanised, better educated and wealthier
Australian population seems to be becoming disenchanted with the rural,
“ocker” association of Broad Australian.
Cox and Palethorpe (this volume) confirm the nature of these changes
(with the second target of /oυ/ now also undergoing a fronting shift), and they
also detect a lowered /æ/ and a lowered first target in /e/. The lowered /æ/
might be seen as counterevidence to the move away from RP-sounding
vowels, but Cox and Palethorpe hypothesise that this is in fact a sociolinguistic hypercorrection towards the prestigious /a/ form (with which /æ/ often
forms a “free variation” pair, as in dance and chance).
It has been suggested in the past that AusE, or at least some speakers of
the dialect, may be moving under the influence of New Zealand English. As
well as raising (rather than lowering) of /æ/, the data might have been
expected to reveal raising of /e/ and retraction of //; such shifts are not present
in the data, however.
Australian phoneticians have focused strongly on vowel analysis, since
the vowel space responds most readily to phonetic change. The consonants in
AusE, however, are not immune to change and variation. Mitchell and
Delbridge (1965), as well as Ingram (1989), noted some assimilation effects in
consonants, and Wells (1982) mentioned characteristics of the consonants /t/
and /l/ in AusE; but Tollfree’s analysis of /t/-reduction and Borowsky’s paper
on /l/-vocalisation (both in this volume) are an important step forward in the
description of AusE.
Tollfree’s research shows a high frequency of reduced forms of /t/ in
everyday speech, either tapped or glottalised. Fricated forms appear to be agerelated and are declining in currency. The study reported by Borowsky shows
that the vocalisation of dark l is highly context dependent: l tends to be
vocalised especially when it is syllabic, when it is followed by a consonant or
when it is preceded by a back vowel. In other words, both syllable position and
adjacent backness (of either vowel or consonant) are predisposing factors.
Social factors in interspeaker variation should not be ignored: Horvath
(1985) showed that the Greek and Italian migrants in her study were leading
the reversal to General Australian from the Broad and Cultivated varieties
(mentioned above), moving more rapidly than the Anglo-Australians. One
might speculate that Anglo-Australian and ethnic speakers are shifting for
different reasons, the former being motivated by a desire to distance themselves from the stigma associated with Broad and from the RP associations of
Cultivated, and the latter by a desire to distance themselves from the low
prestige ethnic broad speech of their parents. According to Kiesling (this
volume: 241) many younger second generation migrants learn AusE no differently from their Anglo-Australian peers. Nevertheless, gender differences are
detectable, with migrant women differentiating themselves more than men, a
fact which suggests that “there is more change to come, given the important
role that women generally play in sound change”.
Gender differences in phonological change are also evident in the data
analysed by Cox and Palethorpe (this volume: 41). Interestingly, all the female
changes relate to vowels that do not relate to social variables; such a phenomenon is usually an indication of a “new change below conscious awareness”.
Regional and social variation
Recent work suggests the presence of more regional variation in AusE than
was once thought to exist. Bradley (1989) comments on such phonological
differences between Sydney and Melbourne as the tendency for the front lax
vowels /i, e, a/ to be higher in Melbourne than in Sydney, for /ə/ as in here to
be monophthongised in Sydney but not Melbourne. Horvath and Horvath (this
volume) find /æ/ to be more strongly preferred over /a/ in words such as dance
and advance in Brisbane than Melbourne, more in Melbourne than Sydney,
more in Sydney than Hobart, and more in Hobart than Mt Gambier (in South
Australia), while this pattern is reversed for the vocalisation of /l/ in these
same cities.
Trudgill and Hannah (1982) argue that for the most part regional variation is lexical, a claim given substance by the work of Pauline Bryant, which
confirms that there is considerable variability in the Australian lexicon
(Bryant 1985, 1989a, 1989b, 1991). Consider some examples. A drinking
fountain in Perth is what is known as a bubbler in New South Wales. A
cocktail frankfurt is a cheerio in both the Northern Territory and Queensland,
but is a little boy in Melbourne. Devon in Sydney is German sausage or
Strassbourg sausage (Straz) in Melbourne, fritz in Adelaide, Windsor sausage
in Queensland and polony in Perth. Spouting in Melbourne is known as
guttering elsewhere. Waste goes into a rubbish bin in Perth, but in Brisbane
and Sydney it goes into a garbage bin, and in Melbourne and Adelaide into a
dust bin.
It is important to keep the facts about regional variation in Australia in
perspective, as Bernard (1989) has noted. The overall picture must remain one
of a continent across whose vast reaches there is comparatively little variation.
“Australians are for practical purposes almost completely unmarked by region
within their language community” (Bernard 1989: 255). Blair (1993) claims
that the regional uniformity of AusE, while largely explicable in terms of the
patterns of early settlement, has also been influenced by the recent strength of
Australians’ perception of themselves as a single nation. The development of
linguistic divergence, which recent social and political factors might have
been expected to promote, has apparently been retarded by the fact that an
Australian’s sense of national identity is stronger than any sense of regional
Australian linguists have, until recently, been consistently reluctant to
recognise the importance of social factors in AusE variability. Even the
Mitchell and Delbridge (1965) study, which in the absence of significant
regional variation presented almost exclusively socially-correlated phonetic
variants, refrained from labelling these as indicators of social dialects. In more
recent times, however, the effects of gender and social class on phonology
have been established and have been extensively reported (see, for example,
Cox 1998; Horvath 1985; Cox and Palethorpe, this volume; Kiesling, this
volume); and work on sex differences at the syntactic/morphological level has
been reported (for example, Eisikovits 1989).
Dialect and ethnicity
Increasingly Australia’s emerging identity is coming to embrace a diversity of
ethnic voices. Varieties of AusE that mark the ethnicity of second and later
generation Australians are referred to as “ethnolects” by Clyne, Eisikovits and
Tollfree (this volume: 226), and their function described as follows:
… an ethnolect, like a community language, offers a means of expressing
linguistic identity, of demonstrating solidarity with one’s ethnic group. Importantly, it provides a means for those who may no longer be fluent in their
ethnic language to express their identification with, and sense of belonging to,
their ethnic group.
The strength of ethnolinguistic identity varies from individual to individual,
and from community to community. Particularly influential factors in ethnolect maintenance are geographical concentration and membership of a distinctive religious group. Research reported in Clyne, Eisikovits and Tollfree’s
chapter is suggestive of the continuing vitality of ethnolects in Australia.
For many, Australia’s new pluralist identity also embodies an indigenous
“voice”, in the form of Aboriginal English. According to Malcolm (this
volume: 201) the range of indigenous varieties that together represent Aboriginal English, “while maintaining certain regional markers and exhibiting a
spectrum of stylistic variation, which may emphasise their relatedness to
pidgins/creoles on the one hand and AusE on the other, have developed
sufficient commonality both of form and of function to enable them to be
comprehended, collectively, as a dialect”. This commonality of features may
be accounted for in part by the common basis of varieties in NSW pidgin, in
part by a continuing process of koinéisation of the increasingly culturally
integrated, geographically mobile and nationally-oriented Aboriginal speech
Any survey of the status and nature of English in Australia will reveal how
closely the national language, in its strength and in its variability, reflects the
essential cohesiveness and diversity of its home culture. From its beginnings
at the end of the 18th century, English in Australia had to provide a medium of
communication that could cope with a new physical environment and a new
social order. As the language has developed over 200 years, its lexicon has
continued to expand to serve the needs of a developed southern hemisphere
nation, and its varieties have continued to reflect the multi-ethnic and multicultural nature of Australian society.
The role of language as a badge of social identity means that English in
Australia serves a double social function. Within Australia, the range of
varieties (or Englishes) provides a set of cultural and social indicators of
ethnicity, social class, gender and age. From an external viewpoint, and
primarily through its prestige dialect (AusE), the language provides a marker
of “Australian-ness” which is increasingly recognisable to speakers of other
Englishes around the world.
Ocker is defined as ‘the archetypal uncultivated Australian working man’, or adjectivally
as ‘distinctively Australian’ ( Macquarie Dictionary, 3rd edition, 1998).
Bennett, George. 1834. Wanderings in New South Wales, Batavia, Pedir Coast Singapore,
and China; Being the Journal of a Naturalist in those Countries during 1832, 1833, and
1834. In two volumes. London: Richard Bentley.
Bernard, John R. 1989. “Regional variation in Australian English: A survey”. In Collins
and Blair, eds. 1989: 255–259.
Blair, David. 1975. “On the origins of Australian pronunciation”. Working Papers of the
Speech and Language Research Centre, Macquarie University July: 17–27.
———. 1993. “Australian English and Australian identity”. In Gerhard Schulz, ed. The
Languages of Australia. Canberra: Australian Academy of the Humanities, 62–70.
Bradley, David. 1980. “Regional differences in Australian English phonology”. University
of Melbourne Working Papers in Linguistics 6: 73–93.
———. 1989. “Regional dialects in Australian English phonology”. In Collins and Blair,
eds. 1989: 260–270.
Bryant, Pauline. 1985. “Regional variation in the Australian English lexicon”. Australian
Journal of Lingustics 5: 55–66.
———. 1989a. “Regional variation in the Australian English lexicon”. In Collins and Blair,
eds. 1989: 301–314.
———. 1989b. “The south-east lexical region in Australian English”. Australian Journal
of Linguistics 9: 85–134.
———. 1991. “A survey of regional usage in the lexicon of Australian English”. In
Suzanne Romaine, ed. Language in Australia. Cambridge: Cambridge University Press,
Butler, Susan. 1996. “Hasta la vista, strine, the lingo’s gone feral”. Sydney Morning Herald
7 December, Spectrum 4.
Collins, Peter and David Blair, eds. 1989. Australian English: The Language of a New
Society. St Lucia: University of Queensland Press.
Cox, Felicity. 1998. “The Bernard data revisited”. Australian Journal of Linguistics 18: 29–
Cunningham, Peter M. 1827. Two Years in New South Wales. (vols I and II.) London:
Henry Colburn.
Dabke, Roswitha. 1976. Morphology of Australian English. (Ars Grammatica, Band 6.)
Munich: Wilhelm Fink Verlag.
Delbridge, Arthur et al. 1997. The Macquarie Dictionary. 3rd edition. Sydney: Macquarie
Dermody, Anthony C. 1980. “Word abbreviation and suffixing in Australian English”. BA
(Honours) thesis, La Trobe University.
Dixon, James. 1822. Narrative of a Voyage to New South Wales and Van Diemen’s Land in
the Ship ‘Skelton’ During the Year 1820. Edinburgh: John Anderson.
Eisikovits, Edina. 1989. “Girl-talk/boy-talk: Sex differences in adolescent speech”. In
Collins and Blair, eds. 1989: 35–54.
Fasold, Ralph. 1984. The Sociolinguistics of Society. Oxford: Basil Blackwell.
Gerstaecker, Friedrich. 1853. Narrative of a Journey Round the World, Comprising a
Winter Passage Across the Andes… With a Visit to the Gold Regions of California and
Australia… (vols I, II, and III.) London: Hurst and Blackett.
Giles, Howard and Robert N. St Clair. 1979. Language and Social Psychology. Oxford:
Basil Blackwell.
Horvath, Barbara. 1985. Variation in Australian English. Cambridge: Cambridge University Press.
Ingram, John C. L. 1989. “Connected speech processes in Australian English”. Australian
Journal of Linguistics 9: 21–49.
McAndrew, Alex. 1992. “Hosties and garbos: a look behind diminutives and pejoratives in
Australian English”. In Claudia Blank, ed. Language and Civilization: A Concerted
Profusion of Essays and Studies in Honour of Otto Hietsch. Frankfurt am Main: Peter
Lang Publishers, II: 111–184.
Meredith, Louisa Anne. 1844. Notes and Sketches of New South Wales, During a Residence
in that Colony from 1838 to 1844. London: John Murray.
Mitchell, Alexander George and Arthur Delbridge. 1965. The Speech of Australian Adolescents: A Survey. Sydney: Angus & Robertson.
Mossman, Samuel. 1852. The Gold Regions of Australia; a Descriptive Account of New
South Wales, Victoria and South Australia. (vols I, II, and III.) London: Hurst and
Peters, Pam. 1993. “American and British English in Australian usage”. In P.H. Peters, ed.
Style on the Move: Proceedings of Style Council 92. Macquarie University: Dictionary
Research Centre, 20–27.
Poynton, Cate. 1984. “Names as vocatives: forms and functions”. Nottingham Linguistic
Circular 13: 1–34.
———. 1989. “Terms of address in Australian English”. In Collins and Blair, eds.
1989: 55–69.
Seal, Graham. 1999. The Lingo: Listening to Australian English. Sydney: UNSW Press.
Sussex, Roland. 1989. “The Americanisation of Australian English: prestige models in the
media”. In Collins and Blair, eds. 1989: 158–168.
———. 1995. “Americanisms roll in to Australian English”. Australian Style 3(2): 2–3.
Taylor, Brian. 1989. “American, British and other foreign influences on Australian English
since World War II”. In Collins and Blair, eds. 1989: 225–254.
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Basil Blackwell.
——— and Jean Hannah. 1982. International English: A Guide to Varieties of Standard
English. London: Edward Arnold.
Wells, John. 1982. Accents of English. (3 vols.) Cambridge: Cambridge University Press.
Section A
English in Australia: structure
The changing face of
Australian English vowels
Felicity Cox and Sallyanne Palethorpe
Speech Hearing and Language Research Centre
Macquarie University, Sydney
During the 1960s there was a flurry of activity at Sydney University with
researchers carefully documenting aspects of Australian English (henceforth,
AusE) pronunciation. The most prominent amongst this group were Alex
Mitchell, Arthur Delbridge and John Bernard. Mitchell and Delbridge (1965)
conducted the first sociolinguistic analysis in this country and their large
survey of the speech of Australian adolescents firmly established their place in
Australian linguistic history. Their colleague, John Bernard, pioneered the
acoustic analysis of AusE vowels and the result of this work became the
standard reference for vowel acoustics (Bernard 1967; Bernard 1970). These
three scholars described the dialect in very detailed terms and provided the
foundation for continuing research into the pronunciation of AusE.
The Australian accent can be differentiated from other forms of English
primarily by the pronunciation of the vowel sounds although there are certain
distinguishing consonantal modifications and suprasegmental effects. It is for
this reason that vowels have been most carefully examined in AusE. Mitchell
and Delbridge (1965) identified three accent types for Australian speakers,
based primarily on how individuals pronounced the six vowel sounds that
occur in the words HAY, HE, HIGH, HOE, HOW and WHO. The three accent
types, referred to as “Broad”, “General” and “Cultivated”, represent divisions
across a range of continuous acoustic variation (Bernard 1970; Cox 1998;
Harrington, Cox and Evans 1997; Mitchell and Delbridge 1965). This
broadness continuum is believed to describe socio-stylistic variation (Ingram
1989) and although it has become convenient to refer to the three varieties as
separate types, they are not discrete entities as there is considerable phonetic
overlap between them. Regional accent differences have not been empirically
identified for AusE. Blair (1993: 68) comments that “there is probably no
other inhabited landmass of such a size that exhibits as little regional variation
as the Australian mainland”. Comparable variation in the broadness continuum
can be seen in all parts of the country (Bernard 1981), with minor regional
effects reportedly restricted to variation in the proportions of speakers from
each accent category who live in a particular area (Mitchell and Delbridge
1965). However, see Oasa (1989) and Bradley (1989) for an alternative
The early descriptions of AusE vowels relied on researchers’ auditory
impressions of the important descriptive characteristics of vowel height and
fronting. These parameters relate very generally to the reference point defined
as the highest point of the tongue during vowel production (Catford 1994).
Vowels that are high (or close), such as the vowels in HE and WHO, have a
high tongue position whereas low vowels, such as those that occur in HARD
and HUD, have a more open vocal tract and a low tongue position. Front
vowels, such as in HID and HEAD make use of a fronted tongue position,
whereas for back vowels, such as in HOARD and HOD, the tongue is further
back in the mouth. The horizontal and vertical tongue positions are used to
define the vowel space and provide a frame of reference for auditorily positioning the vowels. There are other important vowel characteristics such as
whether the lips are rounded as in WHO or unrounded as in HE and whether
the vowel is long as in HARD or short as in HUD. Two distinct vowel types
have also been described for AusE: monophthongs and diphthongs. During
monophthong production, the vowel can be produced for a prolonged period
without the articulators moving significantly from their target position. However, for diphthong production the articulators are required to move in order
for the characteristic diphthong glide to be achieved. The glide for a diphthong
moves from one articulatory position in the vowel space (referred to as
“Target 1”) to another (referred to as “Target 2”). The two targets define the
extent of the glide and each will often correspond to particular monophthong
positions. For instance, for most AusE speakers, Target 1 of the vowel in HAY
is quite similar to the monophthong in HAD and Target 2 is similar to the
monophthong in HID (Cox 1996; Harrington et al. 1997). For the HAY vowel
we can say that the tongue moves from a position near the HAD vowel to a
position near the HID vowel. The impressionistic phoneticians used their
perception of the characteristics of height, fronting, lip position and length to
describe the vowels of AusE and the variations that they encountered. John
Bernard (1967), however, made use of a new acoustic technique to describe
vowels in a more objective manner. Using sound spectrography, he was able
to ascertain the positions of the first three major peaks of energy in the
acoustic vowel spectra. The first two of these peaks, the first and second
formants, are very highly correlated with the parameters of vowel height and
fronting (Fant 1960) and could therefore be used to verify impressionistic
Once AusE vowels had been impressionistically and acoustically documented in the 1960s very little work in these areas continued. There was a
paucity of acoustic vowel research throughout the 70s and early 80s, with few
studies focused on speakers outside Sydney and no acoustic work on the
vowels of females. Dynamic social systems like language inevitably experience change, and impressionistic evidence suggested that the Australian accent was evolving. Cox (1996) recognised that such change should be
documented and in a re-examination of Bernard’s data (Cox 1998) found that
the vowels in the words HEARED, WHO’D and HODE displayed variation
that could be interpreted as representative of change. These vowels exhibited
an acoustic progression of variation across an age range. It has long been
recognised that linguistic variation is the key to understanding language
change: Bloomfield (1933) considered the process of change to be the result of
a group of speakers gradually favouring one phonetic variant over another and
Hockett (1958) suggested that change is the result of asymmetry of acoustic
distribution eventuating in a sound shift. An examination of vowel variance
therefore provides valuable information in the assessment of change. Over a
hundred years ago Sweet (1888) recognised that vowel change was a nonrandom phenomenon which proceeded in a patterned manner. After examining historic data, he proposed three major tendencies for vowel change: “…
short vowels tend to lowering … long to raising” and “the tendency is from the
back to the front” (Sweet 1888: 19–20). Such tendencies in vowel change
have been examined more recently by Labov, Yaeger and Steiner (1972) who
restated Sweet’s observations as three universal constraints on change. Labov
(1994) has provided detailed descriptions of the principles of vowel change
based on current sociolinguistic research and the essence of Sweet’s (1888)
observations remain. There is general agreement with the idea that sound
shifts usually obey natural universal laws and these can often be attributed to
the physical properties of the speech mechanism. However, Ohala
(1983: 190) cautions that “psychological and social factors shaping speech
may run counter to the influence of purely physical factors”. Ladefoged
(1983) believes that phonetics can help to explain why certain historical
changes and processes are more probable than others but argues that a biological basis cannot be attributed to many (maybe most) sound changes.
It has been found that changes to the sound system of a language typically
involve the realignment of phonetic elements (Labov 1994; Lindblom 1986).
As vowel systems are organised within physiological limits to ensure minimal
confusability (Liljencrants and Lindblom 1972), disruptions often lead to
chains of internal shifts which restore symmetry and re-establish maximal
contrast. The Great Vowel Shift is an example of a chain of historic changes
which primarily affected the long vowels of English some time around the 15th
century (Wolfe 1972). More recent English chain shifts have been discussed
extensively by Labov (1994, 1996), and also by Bauer (1979) and Matthews
(1981) in relation to the movement of “Australasian English” vowels. Vowel
variation within various dialects of English has been carefully documented
with the aim of describing ongoing evolution (for example see Dailey-O’Cain
1997; Eckert 1988; Labov 1990, 1994, 1996; Trudgill 1988; Zeller 1997).
However, the examination of phonetic change in AusE has been rather neglected as an area of study. A few researchers have made passing comment on
vowel change but there has been no systematic study of this phenomenon.
Horvath’s sociolinguistic study of the Sydney speech community shows an
increasing Anglo-Australian preference for the central General Australian
accent type over the marginal, Broad and Cultivated, varieties (Horvath 1985).
Bradley (1989) suggests short front vowel raising, monophthongisation of the
centring diphthongs, and fronting of /u/ and /oυ/ as changes. Cochrane (1989)
also comments on a fronting of the second element in /oυ/, and Holmes and
Bell (1990) suggest a New Zealand English influence with the raising of the
front vowels /ε/ and /æ/, and the centralisation of //. These comments,
however, do not have their basis in objective acoustic analysis and it is clear
that such an analysis of vowel change in AusE is required to assess the merit of
these suggestions. It has been firmly established that systemic readjustment of
phonetic elements often occurs as a consequence of phonological change but
there has been little discussion in the literature about the impact of chain shifts
on the relationship between monophthongs and diphthongs. The effects of
change on monophthong/diphthong interdependency have not been documented, nor has there been any discussion on how the two vowel types operate
within the changing vocalic system. AusE provides an interesting source of
data in this regard as the monophthongs and diphthongs are closely linked in
phonetic space (Cox 1996), yet constitute two distinct classes within the
vocalic type (Harrington and Cassidy 1994; Watson and Harrington 1999).
There are two major approaches that one can take when examining
change. The first is a diachronic real-time or trend analysis where two sets of
equivalent data are collected from socially similar people at each end of a time
span. Chambers and Trudgill (1980: 164) state that the results obtained from
analysing such two groups can be compared “with a high level of confidence
that any significant discrepancy between them is a result of a linguistic change
in progress”. The second approach is a synchronic apparent-time analysis
where data is collected from a population and examined for age effects. Age
related variation in speech production is very well documented in the literature
and physiology is the major factor responsible for this variability (Fant 1966).
During the course of a person’s life, the vocal apparatus increases in size and
changes in shape, mobility and structure, all of which affect vocal output (see
Mackenzie Beck 1997, for a review) but speech also varies as a consequence
of non-biological factors associated with ageing (Chambers 1995). Language
changes over time and therefore young people often sound different from
older people because their speech contains more new or innovative forms
(Eckert 1988). Trudgill (1983) believes that an individual’s accent is unlikely
to change substantially past puberty, therefore changes that come into the
language will generally be innovations in the speech of pre-adolescents and
adolescents (Chambers 1995; Wells 1982). This group displays a very high
degree of phonological innovation relative to other age groups and is therefore
at the forefront of change. Sociolinguists have devised the synchronic apparent-time technique for examining sound change in progress that relies on such
age-related variation in speech production. The researcher observes the
speech of people from different age groups while controlling for sex differences, as well as social, stylistic and regional characteristics. This technique
can indicate change if results show differences in a particular form across the
age range (Labov 1994). Where feasible, the results of apparent-time analysis
should be validated with real-time observations due to the possibility that age
related differences may be the result of age-grading rather than change (Bauer
1985; Labov 1994). Age-grading refers to a pattern of age related difference
that repeats itself in successive generations (Chambers 1995). One type of
age-grading is the result of the increasing use of standard forms in language
that may occur in the transition from adolescence to young adulthood as a
consequence of change in social circumstances (Chambers 1995).
In this chapter, we will present the results of diachronic and synchronic
analyses of AusE using independent sets of speech data. The diachronic, realtime analysis, compares acoustic vowel data collected by Bernard (1967) with
similar data collected by Cox (1996) and is based on male speakers. The
synchronic, apparent-time analysis, examines age related vowel variation in a
sample of male and female speakers selected from the Australian National
Database of Spoken Language (ANDOSL) (Millar, Vonwiller, Harrington
and Dermody 1994). Acoustic analysis of vowels provides an objective
method of examining the important descriptive vowel characteristics of height
and fronting using the parameters formant 1 (F1) and formant 2 (F2) which
have very well defined physiological correlates. The frequency of F1 is
inversely related to tongue height and the frequency of F2 is related to tongue
fronting (Fant 1960). These two acoustic parameters therefore allow objective
measurements to be made that relate to the relative articulatory positions of the
vowel sounds. Furthermore, numerical data of this type allows statistical
comparisons to be made between different groups of speakers.
There are four major aims to this analysis:
to test the hypothesis that Australian vowels have recently undergone
to compare real-time and apparent-time analysis procedures;
to examine whether changes to the AusE vowel system conform to
general principles of chain shifting;
to examine the relationship between monophthongs and diphthongs in
vowel change.
Diachronic analysis: A comparison of vowels from the 1960s and
2.1. Method
Twenty-seven General AusE speaking male school pupils from Sydney were
selected from Bernard’s (1967) database to represent the 1960s data. This
group possesses characteristics that best approximate those of the present day
(1990s) group. For the 1990s sample, Cox (1996) collected data from sixty
second generation General Australian speaking 15 year-old males from a
single representative local government area in Sydney. Both groups of speakers read eighteen AusE vowels in the fully stressed h – d context in citation
form from flash cards (see Table 1). The 1960s subjects read each word once
and the 1990s subjects read each word four times in random order. The
symbols that have been chosen to represent the vowels in this analysis are
based on the traditional Mitchell and Delbridge (1965) system that is employed in the Macquarie Dictionary. We recognise that there are shortcomings
to this system but it is the most widely used in Australian linguistics. Some
comments about transcription of AusE vowels can be found in Harrington et
al. (1997).
Table 1. Words and phonemic symbols used for data elicitation
The vowel data was acoustically processed and the single target of the
monophthongs, / i, , ε, æ, a, , ɒ, ɔ, υ, u, /, and the two targets for the
diphthongs, / a, e, ɔ, aυ, oυ/, were established. The vowels /ə/ and /eə/
have been excluded from the present analysis. For the 1960s data, Bernard
(1967) measured the centre frequencies of the first three formants at the vowel
targets by hand from hard copy wideband spectrograms. The present-day
speech data (Cox 1996) was hand labelled using Waves+ (Entropic Research
Laboratory Inc.) and Mu+ speech analysis tools. Vowel targets for the first
three formants were established from high-resolution grey-scale wideband
digital spectrograms and aligned waveforms. The vowel targets were derived
using procedures outlined by Di Benedetto (1989). The four productions of
each token for each speaker in the 1990s data were averaged to satisfy the
statistical requirement for independence of observation. A series of oneway
analyses of variance was performed to examine the relationships between the
independent variable (data set: 1960s and 1990s) and the dependent variable
(F1 or F2).
2.2. Results
Monophthong data is presented by means of plots showing the monophthong
vowel spaces superimposed onto one another in the standard F1/F2 plane with
axes oriented to replicate the traditional vowel map indicating height and
fronting. Each vowel label represents the mean value of the target position for
the vowel.
Table 2 summarises the statistical results with significance defined as p <
.01. This conservative significance level was chosen because the two sets of
data were derived using measurement techniques that differ from one another
due to technological sophistication.
Table 2. Anova results for the significant differences in monophthong and diphthong
realisations in the diachronic analysis between the Bernard (1960s) and Cox (1990s) data.
F value
p value
F value
p value
Figure 1 shows that there are many differences in the vowel target positions
between the two data sets. In the 1990s data /æ/ is more open and retracted, /ɒ,
u, υ/ and // are phonetically closer, and /u/ and // are fronted. The fronting of
/u/ and // represents a parallel shift. Some of these findings are in direct
contrast to claims made by Bradley (1989) in relation to change in Sydney
English. He suggests a retracted // and a raised /æ/ which are not evident in
the present data. However, Bradley’s comments regarding the fronting of /u/
are supported.
Figures 2 to 5 display the diphthong results using F1/F2 plots that depict
the schematic (straight-line) trajectories from Target 1 to Target 2 (the diphthong glide) superimposed onto the monophthong vowel space. The results
are summarised in Table 2. Differences for /e/ (Figure 2) are apparent. The
first target of /e/ in the 1960s data is midway between /æ/ and //. However,
in the 1990s data, the first target is more fronted making it closer to /æ/. The
1960s F2 values at Target 1 for /a/ (Figure 3) are significantly higher than the
1990s values indicating that they are more fronted. The major difference
between the two data sets for /a/ is that the first target is phonetically closer in
the 1990s data than in the 1960s data. There is also a Target 2 effect with the
1990s group displaying a more open position. The 1990s F1 values at Target 1
for /aυ/ (Figure 4) are significantly higher than the 1960s values, therefore the
target is lower in the vowel space. The 1990s Target 2 is higher than in the
1960s data. These differences seem to be a reflection of monophthong differences. As /æ/ is lowered, Target 1 of /aυ/ is also lowered and as /ɒ/ is raised,
Target 2 shows a concomitant raising. For /oυ/ (Figure 5), the 1960s Target 2
F2 values are significantly lower and therefore the 1990s data has a more
fronted second target which appears to be associated with the fronting of /u/.
The shifts for the diphthong trajectories appear to imply a chain of
movements. Figure 6 shows that the /oυ, a, aυ/ trajectory orientation has
rotated in a chain shift fashion so that the relationships between the three
vowels and their affiliated monophthongs have been preserved.
Summary of changes from the 1960s to 1990s:
Raised //
Lowered and retracted /æ/
Raised /ɒ/
Raised /υ/
Raised and fronted /u/
Fronted //
Fronted /e/ Target 1
Retracted /a/ Target 1 and lowered Target 2
Lowered /aυ/ Target 1 and raised Target 2
Fronted /oυ/ Target 2
* Cox
^ Bernard
F1 in Hertz
F2 in Hertz
Figure 1.Monophthong F1/F2 space at target for the Bernard (1960s) and Cox (1990s)
data sets.
* Cox
^ Bernard
F1 in Hertz
F2 in Hertz
Figure 2. Schematic diphthong trajectories for the Bernard (1960s) and Cox (1990s) data
sets from Target One to Target Two for the vowel /e/ superimposed on the F1/F2
monophthong vowel space.
* Cox
^ Bernard
F1 in Hertz
F2 in Hertz
Figure 3. Schematic diphthong trajectories for the Bernard (1960s) and Cox (1990s) data
sets from Target One to Target Two for the vowel /a/ superimposed on the F1/F2
monophthong vowel space.
* Cox
^ Bernard
F1 in Hertz
F2 in Hertz
Figure 4. Schematic diphthong trajectories for the Bernard (1960s) and Cox (1990s) data
sets from Target One to Target Two for the vowel /aυ/ superimposed on the F1/F2
monophthong vowel space.
* Cox
^ Bernard
F1 in Hertz
F2 in Hertz
Figure 5. Schematic diphthong trajectories for the Bernard (1960s) and Cox (1990s) data
sets from Target One to Target Two for the vowel /oυ/ superimposed on the F1/F2
monophthong vowel space.
Bernard: diphthong trajectories
F2 in Hertz
F1 in Hertz
F1 in Hertz
Cox males: diphthong trajectories
F2 in Hertz
Figure 6. The relative diphthong shifts showing the interrelationships between the
monophthongs and diphthongs.
2.3. Discussion
These results strongly suggest that vowel changes have occurred during the
past 30 years and that many of these changes follow patterned relationships.
For instance, raised /aυ/ Target 2 follows raised /ɒ/, and fronted /oυ/ Target 2
follows fronted /u/. Fronted /u/ and // represent a parallel shift and raised /ɒ/
and /υ/ provide an example of a chain shift. However, this analysis has only
examined the speech of males because data from female speakers has not been
previously available for acoustic analysis. Cox (1996) has shown that it is not
valid to generalise from male data to the female speech patterns as gender
specific effects are present in vowel realisation. As Eckert (1989: 247) points
out “there is no apparent reason to believe that there is a simple constant
relation between gender and variation”. We must therefore find a way to
include females in our examination of change. To address this issue, the
speech of both male and female speakers was studied using a synchronic
analysis that does not utilise older archive data. Synchronic analysis is based
on the assumption that new forms will be more predominant in the speech of
the innovative young.
Synchronic analysis: An examination of age related differences in
vowel production
3.1. Method
The subjects used for the analysis were selected from the Australian National
Database of Spoken language (ANDOSL) (Millar et al. 1994). Selected were:
those who spoke AusE as their first language
those who were born in Australia
those who could be included in one of the three major age ranges (18–30, 31–
45, 46+)
those who were unambiguously classified by three phoneticians as speakers
of Broad, General or Cultivated AusE (see Harrington et al. 1997).
There were 41 young (20 males and 21 females), 39 middle age (17 males and
22 females), and 38 older speakers (17 males, 21 females). The General
category is the largest in this analysis with 59 speakers, the Cultivated has 25
speakers and the Broad has 34 speakers. There was also a predictable distribu-
tion of accent type across gender with more Cultivated females than males and
more Broad males than females.
The data was selected from the ANDOSL isolated word materials and the
words chosen for the monophthongs were the same as those used in the
diachronic analysis. However, as ANDOSL does not have a full complement
of vowels in a single context, the diphthong words that were available for
selection varied from the h-d context. For this reason, HAY, HIDE, HOIST,
HOE, and HOW were selected. A single production of each isolated word was
elicited from the subjects who read the words from a computer screen in an
anechoic environment at the National Acoustic Laboratories. The speech data
was acoustically processed and labelled in the manner described for the 1990s
data. The values of the first three formant frequencies were measured at the
vowel target(s) (a single target for monophthongs and two targets for diphthongs) and the accuracy of the process was checked by a trained phonetician
and hand corrections were made if necessary. All data was then rechecked by
another phonetician to ensure consistency of labelling.
Two-way analysis of variance was used to assess the effect of the factors
AGE and ACCENT on the formant frequencies. The results are found in
Table 3, with significance defined as p < .05.
Accent was included as a factor because of the unequal distribution of
accent types across the age groups. Because Harrington et al. (1997) have
Table 3. Anova results for the significant differences in the monophthong and diphthong
realisations in the synchronic analysis.
F value
p value
/i/ F2
// F1
/a/ F1
/υ/ F1
/υ/ F2
/e/ F1T1
/ɔ/ F1T1
/ɔ/ F2T1
/ɔ/ F2T2
/oυ/ F2T1
/oυ/ F2T2
/a/ F1T2
F(2,61) 4.26
F(2,61) 3.38
F(2,59) 5.07
F(2,61) 3.20
F(2,61) 3.57
F(2,61) 7.60
F(2,61) 4.55
F(2,61) 3.18
F(2,61) 3.50
F(2,61) 5.26
F(2,61) 7.79
F(2,61) 4.15
F value
p value
// F1
/oυ/ F2T1
/oυ/ F2T2
/aυ/ F1T2
/aυ/ F2T2
F(2,51) 4.48
F(2,52) 8.62
F(2,52) 12.6
F(2,52) 4.63
F(2,52) 4.73
established accent effects for this group of speakers, an age effect might have
been predicted from accent as a result of this characteristic of the data.
However, no significant AGE X ACCENT interactions were found, age is
considered independent of accent in this analysis and we will therefore not
discuss the accent effects any further as they are described in detail in
Harrington et al. (1997). In reporting the results, a significance level of .05
was used. Trends in the data (p < .10) will also be reported as they may suggest
the possibility of new change.
3.2. Results
The results reported here again focus only on the first two formants but are
illustrated here by means of ellipse plots which present an elliptical range of
two standard deviations from the mean for each vowel. The ellipses are
presented in the standard F1/F2 plane and are based on values from the entire
database with the means for the age groups (“o” = old; “m” = middle; “y” =
young) superimposed onto the average ellipse for each vowel. In vowel
change, we expect to see a progression of vowel movement from the older
through to the younger groups.
In Figure 7, significant age progressions are evident for // of both
males and females where the progression of change is one of lowering. For the
females, there are also significant age effects for /i, υ/ and /a/: the young group
produces the most fronted /i/ and /υ/ and more open /a/ and /υ/. There are also
female trends for /ε/ lowering (p = .080) and // fronting (p = .072). The //
and /a/ movement for females is suggestive of chain shift.
The diphthongs are illustrated in Figures 8 to 11 by showing the actual
movement of the vowel from the first to the second target through the F1/F2
space. Target 1 is labelled as T1. Both the male and female plots (Figure 8)
show the lowered first target of /e/ for the young group which parallels the
lowering of the monophthong //. This effect is highly significant for the
females and a trend is apparent (p = .093) for the males.
The male and female plots (Figure 9) show a general lowering of the
trajectory of /a/ for the younger group. There is a Target 1 trend for the
females toward a lower position (p = .094) and a significant effect for the
second target in F1 confirming vowel lowering. For males the first target
effect is non-significant, however there is a trend toward a more fronted
second target (p = .062).
Wom =
F1 in Hertz
F1 in Hertz
i yom
F2 in Hertz
F2 in Hertz
Figure 7. Mean values of each age group for vowels that showed significant age effects
superimposed on the average monophthong F1/F2 space for females (a) and males (b).
F1 in Hertz
F2 in Hertz
F1 in Hertz
T1 a
F2 in Hertz
Figure 8. Trajectories of each age group for /e/ superimposed on the average monophthong F1/F2 space for females (a) and males (b).
F1 in Hertz
a T1
F1 in Hertz
F2 in Hertz
F2 in Hertz
Figure 9. Trajectories of each age group for /a/ superimposed on the average
monophthong F1/F2 space for females (a) and males (b).
The age effects for /oυ/ are quite striking. Figure 10 shows the significant F2 differences for both males and females indicating progressive changes
in F2 from the old through to the young group with retraction in Target 1 and
fronting in Target 2. These two target effects are the result of a trajectory that
has changed in direction from a nearly vertical initial movement to one that
has become much more fronted. It should be stressed that in this analysis the
vowel was produced in an open syllable that will impact on the realisation of
the final element. We therefore cannot compare the absolute positions of this
vowel with that in the previous analysis although the trend of trajectory tilt
remains the same.
For /ɔ/ (Figure 11), the young females have a rather lower and more
fronted trajectory throughout with F1 and F2 Target 1 and F2 Target 2
differences being significant. There do not appear to be any consistent patterns
for /ɔ/ for the males and no age effects were present for F1 or F2 for the first
or second targets.
There are no significant age effects for Target 1 for /aυ/ despite the
apparent progression in F1 shown in Figure 12. This first target movement
again appears to parallel the lowering of //. For the males only there is a
difference between the old group and the young and middle groups which is
significant at p = .014 for F1 and at p = .012 for F2.
The results from the synchronic analysis above show that there are
significant differences between the age groups for certain vowels. In summary, the direction of movement at the vowel target is:
For both males and females
// is lowering
/oυ/ Target 1 is retracting and Target 2 is fronting
/e/ Target 1 is lowering
For females
/a/ is lowering
/i/ is fronting
/υ/ is lowering and fronting
/a/ Target 2 is lowering
/ɔ/ Target 1 is lowering and fronting, Target 2 is fronting
For males
/aυ/ Target 2 is lowering
3.3. Discussion
The consistencies between the two analyses for males are the lowered // and
fronted Target 2 of /oυ/ which were found using two separate analysis
techniques and two independent data sets. These two effects were also found
in the synchronic analysis for females. In addition, the synchronic analysis for
females showed a lowered Target 1 of /e/. A systematic male trend was also
apparent and the diachronic analysis also showed this effect at p = .037.
Therefore, lowered //, fronted /oυ/ Target 2, and lowered /e/ Target 1 can be
considered to represent valid changes in progress in the Australian vowel
In the synchronic analysis there were many more significant age effects
for females than for males supporting the contention that females are more
progressive with regard to some types of change (Eckert 1989; Labov 1990).
Female-only effects occurred for lowered /a/ and /υ/, lowered and fronted /ɔ/
Target 1 and fronted Target 2, as well as fronted /i/ and /υ/. Labov (1990)
suggests that females are more progressive with regard to change that is below
conscious awareness and most of the vowels that show movement effects here
are indeed those that have not traditionally been shown to have any social
associations for AusE. The retraction of Target 1 of /oυ/ which is found for
both males and females in the synchronic analysis, but not at all in the
diachronic analysis, will be discussed below.
F1 in Hertz
F1 in Hertz
F2 in Hertz
F2 in Hertz
F1 in Hertz
Figure 10. Trajectories of each age group for /oυ/ superimposed on the average monophthong F1/F2 space for females (a) and males (b).
F2 in Hertz
Figure 11. Trajectories of each age group for /ɔ/ superimposed on the average monophthong F1/F2 space for females.
Figure 12. Trajectories of each age group for /aυ/ superimposed on the average monophthong F1/F2 space for males.
There are some inconsistencies between the two analyses and to discuss
them we need to look more closely at the different characteristics of the data sets.
3.3.1. Contextual differences
ANDOSL used open syllables for /e, aυ, oυ/ and HOIST for /ɔ/ whereas Cox
and Bernard used h—d frame contexts. This would have most impact on the
second target of the diphthongs and may therefore account for the /aυ, ɔ/
Target 2 differences between the two analyses.
3.3.2. Age differences
Bernard’s pupils and the Cox group were aged approximately 15 years and the
speakers in the ANDOSL young group were between 18 and 30. We expect
the younger Cox group to be more innovative and therefore some of the
diachronic differences may indicate very new change that will need to be
further assessed by examining a more extensive age range.
3.3.3. Regional differences
Perhaps the biggest discrepancy between the three data sets relates to region.
Bernard provides no detailed regional information; however, we do know that
his young male General speakers were from Sydney. The Cox subjects were
from Sydney’s Northern Beaches. Eighty percent of ANDOSL speakers came
predominantly from Sydney with the remainder from country N.S.W. and
other states. Regional differences may provide some of the answers to the
discrepancies we have seen.
Regional effects in the ANDOSL database
Regional effects have been largely overlooked in Australian phonetic studies
as no prior empirical research has provided any substantial evidence for such
effects on vowel pronunciation. Therefore, there is an assumption that the
uniformity theory for AusE holds true and researchers have not generally
sought to control for regional differences in their sample research populations.
We decided to make use of the small amount of regional information available
in ANDOSL to explore this issue with the aim of explaining some of the
discrepancies between the results of our two previous analyses. For about half
the ANDOSL speakers it was possible to determine area of upbringing during
the late primary and high school years: periods which are influential in the
development of accent (Eckert 1988). Only preliminary work on regional
effects is possible with this data given the small numbers of subjects available.
However, trends may become evident that help us to explain some of the
results of our previous work and give us future directions to explore in
research. A detailed explanation for all the differences between the two
analyses would require an examination of a larger and more carefully controlled sample of speakers. We have therefore chosen to restrict the discussion
of regional effects to two examples of the inconsistencies in the synchronic
and diachronic analyses: The fronting of /u/ and // which was found to be
present in the diachronic but not the synchronic analyses and the retraction of
the first target of /oυ/ which was present in the synchronic but not the
diachronic analysis.
The inconsistency of results regarding the fronting of /u/ and // can be
explained in relation to the age and regionally distributed nature of the
ANDOSL sample. A possible reason that age progressions were not seen for
these two vowels in the synchronic analysis is that the age and region data
conflict with one another. Cox and Palethorpe (1998) have shown that /u/ and
// are more retracted for Northern Suburbs speakers than for Western Suburbs speakers. If these vowels are in the process of fronting, as indicated in the
diachronic analysis, then young speakers should exhibit more fronting than
old speakers. The ANDOSL Western suburbs group contains a larger number
of older speakers and the Northern suburbs group contains a greater number of
younger speakers. There is a potential conflict between region and age and the
age effect in the synchronic analysis will be diluted. If region had been
controlled as a variable, the age progression would be seen in the analyses.
Figure 13 shows the difference between the old and young groups for the
vowels /u/ and // produced by speakers from a single region. The progression
of fronting is quite clear.
The inconsistencies regarding the retraction of the first target of /oυ/ can
be examined by looking at plots that show the diphthong trajectories for the
young and old Northern and Western Suburbs groups (Figure 14). The retraction of the first target can be clearly seen in the Northern Suburbs data for both
young and old female groups. However, Target 1 retraction is not present in the
Western Suburbs plots indicating a regionally restricted effect. For males, the
first target retraction is present for both the young Western and Northern groups
indicating an age effect. The male and female differences in this variable also
indicate a gender effect. The /oυ/ Target 1 retraction may be a relatively recent
and regionally distributed phenomenon and therefore not present in Bernard or
Cox male data sets. Figures 14a and 14b show the progression of retraction for
the young that is more prominent in the Northern Suburbs. The interesting fact
about /oυ/ is that the entire trajectory varies according to age and region for both
males and females. The trajectory tilt which can be traced back to the 1960s for
the vowel preceding /d/ (Cox 1998) is present in this open vowel context for both
the old and young female Northern Suburbs groups but only for the young
Northern Suburbs male group.
These findings indicate that regional effects are important in the interpretation of results, strongly suggesting region as a valid factor in the examination of variation in AusE. Change is not uniform across region and gender so
research that fails to control for these factors is open to misinterpretation.
Apart from explaining some of the discrepancies in the results of the two
independent analyses of change, the preliminary examination of regional
effects calls into question the validity of the uniformity theory for AusE which
states that vowel variation is consistent across the country and restricted to the
broadness continuum. Further detailed investigation of regional variation is
necessary if we are to understand the nature of change and the social significance of variation.
* +
F1 in Hertz
F2 in Hertz
Figure 13: Mean values of /u/ and // superimposed on the average monophthong F1/F2
space for young and old Western Suburbs females.
F2 in Hertz
no ny
wo ny
F1 in Hertz
F1 in Hertz
F2 in Hertz
Figure 14: Trajectories of /oυ/ superimposed on the average monophthong F1/F2 space
for young and old Western Suburbs and Northern Suburbs females (a) and males (b).
The results of both the synchronic and diachronic analysis strongly suggest
that the vowels of AusE have undergone a process of evolution. Particularly
strong is the evidence for lowered /æ/, lowered /e/ Target 1 and fronted /oυ/
Target 2. The theory that AusE is moving under the influence of New Zealand
English (Holmes and Bell 1990) is not supported by this analysis. New
Zealand English displays a raised /æ/ and /ε/ and centralised/retracted //
relative to AusE (Bauer 1986; Watson, Harrington and Evans 1998) and such
movements have not been observed here. Some of the changes discussed
above do share similarities with New Zealand English, such as the fronting of
/u/ and // and the fronting of the second target of /oυ/, but similar movements
have been recently documented for other English dialects (Bauer 1985;
Henton 1983; Labov 1990; Luthin 1987; Moonwomon 1985). Therefore,
these changes cannot be attributed to a New Zealand influence.
The observed lowering of /æ/ in this data may be the result of sociolinguistic hypercorrection (Labov 1972) occurring in response to a perceived
prestigious alternative. There are many words in AusE for which /a/ or /æ/ are
acceptable alternative choices, particularly in prenasal environments such as
in “dance” and “chance”. The editors of the Macquarie Dictionary (Delbridge
1981) have chosen to place /æ/ as primary to /a/ due to its more widespread
usage, but the /a/ form is considered prestigious. Many speakers, particularly
from the eastern states, chose not to use this prestigious form as they consider
it “indicative of affectation, pedantry, or snobbishness” (Wells 1982: 599).
They may, however, continue to use the /æ/ variant but lower its realisation to
a more open position while maintaining the distance from /a/ in F2.
The change in /oυ/ that is clearly documented in both the synchronic and
diachronic analyses is of particular interest. Mitchell and Delbridge (1965)
made specific comment about an unusual variant of /oυ/ found chiefly in
South Australia which exhibited a range of realisation including variants from
[ɔυ] to [ɒy]. Cochrane (1989) also contributes to a discussion of this variant,
noting his impression that it is becoming more common; and Oasa (1989)
claims a regional effect in the rising F2 pattern for /oυ/ used by Adelaide
speakers. The data presented here from Sydney speakers strongly suggest that
the upward glide is quite fronted, approaching [y] in some cases preceding
/d/, and that the first target has shifted towards [ɔ]. It is possible that the new
/oυ/ has arisen from the original Adelaide variation but the fronting tendency
may be attributable to a natural progression inherent in vowel systems as is
proposed by Labov (1994). Luthin (1987) has found that fronting of /oυ/ and /
u/ is increasing in California, and Henton (1983) and Bauer (1985) discuss the
fronting of /u/ in RP. The interesting aspect of the /oυ/ change is that not only
the first target but also the second target participate strongly in the change,
indicating a reorientation of the trajectory. Cox (1998) found evidence in
Bernard (1967) of a change in the second target of /oυ/ suggesting that this
change has been in progress for some time.
The interrelationships between monophthong and diphthong realisations
are clearly documented in the analyses. It appears that the second target of /oυ/
preceding /d/ is closely affiliated with the fronting of /u/ (see Figure 5).
Movements of one vowel affect others, and the effects are the result of vowel
systems’ propensity to obey phonological patterning laws and maintain the
principles of maximal contrast (Lindblom 1986). Parallel shifts are observed
between /u/ and //, and chain shifts are observed between /ɔ/ and /υ/ in the
diachronic analysis; a chain shift can also be seen for female /æ/ and /a/ in the
synchronic analysis. The /oυ, a, aυ/ shift in the diachronic analysis represents
a chain of movements of trajectory orientation rather than simply target
change. This implies that in any discussion of diphthong change more emphasis could be placed on dynamic characteristics of the vowel.
The universal tendencies discussed by Labov (1994), that long vowels
move up and forward and short vowels move down and back, appear to be
supported for the fronting of /u/, // and the second target of /oυ/. The
lowering of short nuclei is supported by the lowering of //. The more minor
changes often run counter to Labov’s principles: for example, the lowered /a/
(a long vowel) and fronted /υ/ (a short vowel) found in the female analysis.
The synchronic analysis provides evidence for gender differences in
vowel change as there are many more female effects than male effects. All
female changes relate to vowels that do not have social associations in AusE,
perhaps suggesting the possibility of new change below conscious awareness.
The specific gender effects occur for the lowering of /a/ and /υ/, the fronting of
/i/ and /υ/ and the /ɔ/ effects; but these will need to be further assessed in light
of our findings relating to the importance of regional control.
The limited examination of regional effects in the ANDOSL provides
strong evidence against the uniformity theory for AusE and suggests that
regional effects should be more carefully controlled in phonetic research.
Further research on change should ensure selection of speakers from different,
but controlled, age and regional groups to determine the vowels undergoing
change, the extent of the movement, the gender differentiation and the identity
of the initiating groups. Socioeconomic information may also provide useful
information in the assessment of linguistic evolution. The progress of change
across other urban and rural centres should also be examined to gain a full
picture of the nature and extent of vowel movement and the geographical
diffusion of identified changes in progress in Australian English.
Bauer, Laurie. 1979. “The second great vowel shift?”. Journal of the International Phonetic
Association 9: 57–66.
———. 1985. “Tracing phonetic change in Received Pronunciation of British English”.
Journal of Phonetics 13: 61–81.
———. 1986. “Notes on New Zealand English phonetics and phonology”. English WorldWide 7: 225–258.
Bernard, John R. L. 1967. “Some Measurements of Some Sounds of Australian English”.
Ph.D., Sydney University.
———. 1970. “Toward the acoustic specification of Australian English”. Zeitschrift für
Phonetik 2/3: 113–128.
———. 1981. “Australian Pronunciation”. In Delbridge et al., eds. 1981: 18–27.
Blair, David. 1989. “The development and current state of Australian English: A survey”.
In Collins and Blair, eds. 1989: 171–175.
———. 1993. “Australian English and Australian national identity”. In G. Schultz, ed. The
Languages of Australia. Canberra: Australian Academy of the Humanities, 62–70.
Bloomfield, Leonard. 1933. Language. New York: Holt and Company.
Bradley, David. 1989. “Regional dialects in Australian English phonology”. In Collins and
Blair, eds. 1989: 260–270.
Catford, John C. 1994. A Practical Introduction to Phonetics. Oxford: Oxford University
Chambers, John K. 1995. Sociolinguistic Theory: Linguistic Variation and its Social
Significance. Cambridge, Mass: Basil Blackwell.
——— and Peter Trudgill. 1980. Dialectology. Cambridge: Cambridge University Press.
Cochrane, George R. 1989. “Origins and development of the Australian accent”. In Collins
and Blair, eds. 1989: 176–186.
Collins, Peter and David Blair, eds. 1989. Australian English: The Language of a New
Society. St Lucia: University of Queensland Press,
Cox, Felicity. 1996. “An Acoustic Study of Vowel Variation in Australian English”. Ph.D.,
Macquarie University.
———. 1998. “The Bernard data revisited”. Australian Journal of Linguistics 18: 29–55.
——— and Sallyanne Palethorpe. 1998. “Regional variation in the vowels of female
adolescents from Sydney”. Proceedings of the Fifth International Conference on Spoken Language Processing. Sydney, Volume 6: 2359–2362.
Dailey-O’Cain, Jennifer. 1997. “Canadian raising in a midwestern U.S. city”. Language
Variation and Change 9: 107–120.
Delbridge, Arthur et al., eds. 1981. The Macquarie Dictionary. Sydney: Macquarie Library.
Di Benedetto, Maria-Gabrielle. 1989. “Vowel representation: Some observations on temporal and spectral properties of the first formant frequency”. Journal of the Acoustical
Society of America 86: 55–66.
Eckert, Penelope. 1988. “Adolescent social structure and the spread of linguistic change”.
Language in Society 17: 183–207.
———. 1989. “The whole women: Sex and gender in variation”. Language Variation and
Change 1: 245–267.
Fant, Gunnar. 1960. Acoustic Theory of Speech Production. The Hague: Mouton.
———. 1966. “A note on vocal tract size factors and non-uniform F-pattern scalings”.
Speech Transmission Laboratories, Quarterly Progress Status Report 4: 22–30.
Harrington, Jonathan and Steve Cassidy. 1994. “Dynamic and target theories of vowel
classification: Evidence from monophthongs and diphthongs in Australian English”.
Language and Speech 37: 357–373.
——— and Steve Cassidy. 1999. Techniques in Speech Acoustics. Dordrecht: Kluwer.
———, Felicity Cox and Zoe Evans. 1997. “An acoustic phonetic study of broad, general
and cultivated Australian English Vowels”. Australian Journal of Linguistics 17: 155–
Henton, Caroline G. 1983. “Changes in the vowels of Received Pronunciation”. Journal of
Phonetics 11: 353–371.
Hockett, Charles F. 1958. A Course in Modern Linguistics. New York: Macmillan.
Holmes, Janet and Allan Bell. 1990. “Attitudes, varieties, discourse: an introduction to the
sociolinguistics of New Zealand English”. In Allan Bell and Janet Holmes, eds. New
Zealand Ways of Speaking English. Clevedon: Multilingual Matters, 1–20.
Horvath, Barbara. 1985. Variation in Australian English: The Sociolects of Sydney. Cambridge: Cambridge University Press.
Ingram, John. 1989. “Connected speech processes in Australian English” In David Bradley,
Roland D. Sussex and Graham K. Scott, eds. Studies in Australian English. La Trobe
University for the Australian Linguistic Society, 21–50.
Labov, William. 1972. Sociolinguistic Patterns. Pennsylvania: University of Pennsylvania
———. 1990. “The intersection of sex and social class in the course of linguistic change”.
Language Variation and Change 2: 205–254.
———. 1994. Principles of Linguistic Change, Volume 1, Internal Factors. Oxford: Basil
———. 1996. “The organisation of dialect diversity in North America”. The fourth
International Conference on Spoken Language Processing, Philadelphia SuA1L2.2.
———, Malcah Yaeger and Richard Steiner. 1972. A Quantitative Study of Sound Change
in Progress, Vols I and II. Philadelphia: U.S. Regional Survey.
Ladefoged, Peter. 1983. “The limits of biological explanations in phonetics”. In Antonie
Cohen. and Marcel P. R. van den Broeke, eds. Abstracts of the Tenth International
Congress of Phonetic Sciences. Dordrecht: Foris, 31–37.
Liljencrants, Johan and Lindblom, Bjorn. 1972. “Numerical simulation of vowel quality
systems: The role of perceptual contrast”. Language 48: 839–862.
Lindblom, Bjorn. 1986. “Phonetic universals in vowel systems”. In John J. Ohala and Jeri J.
Jaeger, eds. Experimental Phonology. Orlando: Academic Press, 13–44.
Luthin, Herbert W. 1987. “The story of California (owe): The coming of age of English in
California”. In K.M. Denning et al., eds. Variation in Language: Proceedings of the
Fifteenth Annual Conference on New Ways of Analysing Variation. Department of
Linguistics, Stanford University, 312–324.
Mackenzie Beck, Janet. 1997. “Organic variation to the vocal apparatus”. In William
Hardcastle and John Laver, eds. The Handbook of Phonetic Sciences. Oxford:
Blackwell, 256–297.
Matthews, Richard. 1981. “The second great vowel shift?” Journal of the International
Phonetic Association 11: 22–26.
Millar, Bruce, Julia Vonwiller, Jonathan Harrington, and Philip Dermody. 1994. “The
Australian National Database of Spoken Language”. In Proceedings of the International Conference on Acoustics, Speech and Signal Processing, Volume 2, Perth: The
University of Western Australia, 67–100.
Mitchell, A.G. and Arthur Delbridge. 1965. The Speech of Australian Adolescents. Sydney:
Angus and Robertson.
Moonwomon, Birch. 1987. “Truly awesome: (O) in Californian English”. In Keith M.
Denning et al. eds. Variation in Language: Proceedings of the Fifteenth Annual
Conference on New Ways of Analysing Variation in Language, Department of Linguistics, Stanford University, 325–330.
Oasa, Hiroaki. 1989. “Phonology of current Adelaide English”. In Collins and Blair, eds.
1989: 271–287.
Ohala, John. J. 1983. “The origin of sound patterns in vocal tract constraints”. In Peter F.
MacNeilage, ed. The Production of Speech. New York: Springer-Verlag, 189–216.
Sweet, Henry. 1888. History of English Sounds. Oxford: Clarendon Press.
Trudgill, Peter. 1983. On Dialect. Oxford: Basil Blackwell.
———. 1988. “Norwich revisited: Recent linguistic changes in an English urban dialect”.
English World-Wide 9: 33–49.
Watson, Catherine and Jonathan Harrington. 1999. “Acoustic evidence for dynamic formant
trajectories in Australian English”. Journal of the Acoustical Society of America, 106:
———, Jonathan Harrington and Zoe Evans. 1998. “An acoustic comparison between New
Zealand and Australian English vowels”. Australian Journal of Linguistics 18: 185–
Wells, John. C. 1982. Accents of English 3: Beyond the British Isles. Cambridge: Cambridge University Press.
Wolfe, Patricia M. 1972. Linguistic Change and the Great Vowel Shift. Berkeley: University of California Press.
Zeller, Christine. 1997. “The investigation of sound change in progress: /æ/ to /e/ in
Midwestern American English”. Journal of English Linguistics 25: 142–155.
Variation and change
in Australian English Consonants
Reduction of /t/1
Laura Tollfree
Monash University
This chapter supplies a descriptive phonetic account of variation in modern
Australian English (AusE) /t/. The consonant /t/ has been chosen for its typical
reducing (leniting) behaviour in Australian and other varieties of English. The
primary data in the current survey derive in small part from a 1993–4 survey of
urban speakers,2 and in larger part from continuing fieldwork in metropolitan
Melbourne and rural Victoria as part of the Dimensions of Spoken Australian
English project in the Department of Linguistics at Monash University (since
The chapter begins with a preliminary characterisation of plosive, tapped,
fricated and glottalled/glottalised realisations of /t/, and an introduction to
consonantal reduction, in Section 2. A discussion of some of the information
available on the historical development of AusE and AusE /t/ is given in
Section 3.1. Results from previous studies of AusE /t/ are summarised in
Section 3.2. This is followed in Section 4 by an outline of the methods used
in collection of data for the current survey. Illustrations of /t/ “reductions”
from the current survey are presented in Section 5 in standardised format,
using International Phonetic Alphabet (IPA) symbols, taking into account a
range of phonological, morphological and prosodic contexts. A table of quantified auditory findings is presented in Section 5, and a discussion of the
auditory results for voiced /t/-tapping (flapping) in Section 5.1, /t/-frication in
Section 5.2 and glottalisation in Section 5.3. The principal auditory findings
are assessed in terms of speaker attitudes and possible future directions of
change in Section 5.4.
It is important to view the patterns of phonetic realisation of /t/ in the
wider context: /t/ typically reduces in many accents of English (e.g. British
and General American varieties), and so comparison is drawn with nonAustralian examples where this may be enlightening.
Instrumental (acoustic phonetic) techniques have been used to verify and
supplement the auditory description. This is discussed in Section 6. The
method employed is outlined, and brief acoustic profiles of voiced tapped /t/,
fricated /t/ and glottalised variants of /t/ are supplied in Sections 6.1, 6.2 and
6.3 respectively. This includes a discussion of some of the practical problems
involved in investigating /t/ sounds both auditorily and instrumentally. The
chapter concludes with a brief consideration in Section 7 of the relevance of
the findings to sociolinguistic and phonological analyses, and to the forensic
analysis of speech samples in the Australian context.
Realisations of /t/
In the articulation of plosive [t], pulmonic air is compressed behind an
alveolar closure (obstruction) formed between the tip of the tongue and the
upper alveolar ridge and side teeth, with the vocal cords normally widely
abducted (pulled open) for voicelessness. The air escapes with plosion on
cessation of alveolar closure. The precise supralaryngeal (oral) configuration
for [t], and the posture of the lips during [t], are determined by the flanking
segments. Plosion of [t] is influenced by a range of factors: it is hindered when
the airstream is occluded by a secondary closure (e.g. catkin); velar lowering
causes the airstream to be directed thought the nasal cavity (e.g. witness); in
lateral release, tongue tip contact is maintained (e.g. cutlass). If word-initial [t]
is followed, for example, by [e], as in ten, then co-occurring with the alveolar
closure will be a tongue body constriction appropriate for [e]. For pre-pausal
released [t], the last audible correlate will be the release phase of the stop. This
is heard as weak voiceless frication which varies in extent.
The vocal cords are usually abducted at the start of the alveolar occlusion
and then adducted (brought back together) during the opening transition (in
other words, they start to move towards the posture for voicing). As soon as
the air pressure in the oral tract has dropped sufficiently, due to the dissipation
of the burst of air, the vocal folds may be set into vibration again for the
voicing of a following vowel.
There are in fact three glottal states in speech production. Glottal spreading produces voicelessness, and glottal closure produces the glottal stop [ʔ]. If
neither is present, the glottis presents an alternative adjustment which promotes voicing. This is of relevance in the auditory analysis of glottalised
variants of /t/ (see Section 6.3).
Contemporary AusE, in common with most other forms of English,
exhibits aspirated and unaspirated variants (allophones) of /t/. Aspirated variants are typically found preceding a stressed vowel (for example, tar [tha:]).
Aspiration (a relatively unimpeded egressive, voiceless puff of air) is produced after the tongue tip contact is released and before the vocal folds begin
vibrating again for voicing for a following vowel. Unaspirated variants lack
the puff of air, and are found following [s] (for example, star [sta:]). Variants
preceding an unstressed vowel are either weakly aspirated or unaspirated (for
example, water). The plosive may be unreleased in pre-consonantal environments (as in sit down, catflap) or in final position before a pause. In these
forms there is alveolar occlusion and air pressure build-up as for [t], but the
occlusion is not released and there is no audible stop burst.
Plosive [t] is usually the only realisation of /t/ in word-initial position in
AusE,4 but in intervocalic medial, intervocalic final, pre-consonantal medial,
pre-consonantal final and pre-pausal final positions, a number of other variants may be found, formed by reduction processes. These variants (along with
a range of other consonant and vowel features5) add to the character of AusE:
voiced taps (flaps) [ɾ], fricatives [ts], and glottalised variants (traditionally
transcribed as [ʔ], [ʔt] and [tʔ]: see discussion in Section 3.2 below). There is
considerable phonetic variation within each of these broad categories; their
distribution is conditioned by phonological, prosodic and morphological factors, and their frequency is governed in part by social factors (see Section 5).
Consonantal reduction (lenition or weakening) processes give rise to a
weakened articulation, and affect a range of segments. There are three main
types: spirantisation (where a plosive becomes a fricative), vocalisation
(where the product is a glide or a liquid) and debuccalisation (loss of articulation in the oral cavity). Each of these disparate types of reduction are illustrated by AusE /t/.
Plosive [t] (in common with the other voiceless obstruents) is described
as a fortis (“hard”) sound, since the articulatory effort involved is greater than
for voiced obstruents which are lenis (“soft”). This means that the voiced tap
is, in articulatory terms, a “weaker” sound than the plosive, and the alternation
between [t] and [ɾ] can be viewed as a vocalisation (e.g. Harris 1994: 121).
Fricated forms, meanwhile, lack the period of full occlusion associated with
plosive forms. Instead, they are formed by a narrowing of the aperture between the tongue tip and the alveolar ridge, so as to cause turbulent air flow.
Thus /t/-frication is a spirantisation process. Glottalled forms (where /t/ is
realised as [ʔ]) involve total loss of the supralaryngeal articulation; these may
be described as debuccalisation. Glottalised variants (formed by the association of an alveolar closing gesture with a glottal closing gesture, varying in
terms of the relative timing of the two gestures) are, strictly speaking, not true
debuccalisation processes, since the supralaryngeal gesture is not lost, but
acoustically hidden. They may be viewed as forming part of the lenition
continuum, however (see Section 6.3).
Before turning to a full account of the results of the current survey, the
historical information available on the development of AusE and AusE /t/ will
be briefly assessed.
Australian English /t/: historical information
3.1. The development of Australian English
It is important to start with two assumptions which relate to general principles
of language variation and change. The first assumption concerns the historical
development of AusE, which began with a number of landmarks in Australia’s
recent history: the transportation of convicts from Britain in 1788; the advent
of free settlement, mainly by British migrants, in 1793; and in the early
nineteenth century, the mass migration of the Irish.6 Geographical separation
from Britain and Ireland meant that there was little linguistic feedback (in
either direction). The varieties of English spoken by the early British and Irish
migrants underwent development and divergence, and, according to Mitchell,
a process of local levelling is likely to have been responsible for the establishing of a form of “Broad” AusE (see below) in New South Wales as early as the
1830s (see Yallop, this volume). AusE has continued to develop from these
early forms to produce the varieties we hear today.7 This leads to the conclu-
sion that recent Australian developments, such as the /t/-reduction processes
to be discussed below, are separate developments from otherwise phonetically
identical processes which can be identified in contemporary British and Irish
The second assumption reflects a general fact about language variation
and change. AusE pronunciation is a continuum, rather than a number of
discrete accents (see Yallop, this volume). “Australian English” then refers to
one or more of the varieties that fall on this continuum. Put another way, each
AusE speaker’s pronunciation will fall somewhere on this line of continuum.
This said, it is often convenient to conceive of general categories at the
extremes of the continuum, and at one or intermediate points: for AusE the
general categories which are generally accepted are “Broad”, “General” and
“Cultivated” following Mitchell and Delbridge (1965) (see also Collins and
Blair, this volume). These categories were largely determined by plotting
variation in the realisation of /i/, /e/, /oυ/, /u/, /a/ and /aυ/ along a sociolectal
continuum. While the indications are from the Dimensions of Spoken Australian English project at Monash University that this characterisation cannot be
precisely applied to contemporary AusE, it does provide a useful point of
reference for the middle and older generation speakers.
Section 3.2 examines the evidence from previous studies, and from
literature written in popular dialect, pertaining to the development of /t/
sounds in AusE.
3.2. Previous studies
It was noted above that there are a number of clearly identifiable categories
of phonetic variants (allophones) of /t/ in modern AusE. Yet there is relatively
little information available on /t/ from the descriptions supplied in the surveys
by Baker (1945), Mitchell (1946), Mitchell and Delbridge (1965) and Turner
(1972). Baker remarks explicitly that “The dental plosives [t] and [d] are rarely
modified” (1966: 450) and that the “glottal plosive [ʔ] (as found in the Cockney
and Scottish rendering of [bɑʔl] for [bɑtl], i.e. bottle) is practically unknown in
Australia — in fact I have heard it clearly only twice in more than 30 years from
Australian speakers” (1965: 451). Certainly, in the example bottle to which
Baker refers, /t/ occurs in pre-syllabic /l/ position, where glottalling/
glottalisation is not attested in the current survey. Nor are glottalled/glottalised
variants attested in initial position e.g. time, before /r/ e.g. train, or in
intervocalic medial contexts such as in letter (importantly, native speakers are
particularly sensitive to glottalled/glottalised variants in intervocalic medial
environments).9 However, as we shall see, glottalled/glottalised variants of /t/
do turn up frequently in other contexts in AusE — specifically before syllabic
nasals, in final and pre-consonantal contexts (where native speakers are
notoriously insensitive to glottalled/glottalised forms), and in intervocalic final
contexts. It might be unsafe, therefore, to conclude that final and pre-consonantal glottalling/glottalisation was absent from the AusE of 30 or so years ago as
suggested by the documentation; rather that in these environments they might
simply have gone unperceived. Another apparently contradictory finding is
Wells (1982) where the absence of /t/-glottalling in AusE is explicitly reported.10
There had been little interest in AusE until Mitchell’s work in 1945 and
Mitchell and Delbridge’s sociolinguistic survey in 1965. The 1965 survey
focused on vowels, and did not report tapped, fricated or glottalled/glottalised
variants of /t/ in the discussion on consonants. It is now hard to date tapping,
frication and glottalisation developments. There is no mention of a tapped
variant until the 1980s, but, as for glottalisation, this does not conclusively
deny its presence in AusE before this time. Tollfree (1996) has the first
documentation of fricative /t/, but whether it has always been present in AusE
(which may explain why it has apparently gone unnoticed by Australians and
linguists alike), or whether there has been a recent frication development, is
hard to ascertain. Examining the historical evidence available, we find that
Modern Irish Gaelic has a “voiceless alveolar slit fricative” as an allophone of
/t/ (e.g. Wells 1982: 429; Harris 1984: 121). Assuming that this “special
allophone” was also present in the English spoken by Irish settlers in Australia
in the eighteenth and nineteenth centuries, and given that Irish immigrants
formed a large proportion of the incoming settlers by the 1830s, it is possible
that fricated variants of /t/ were established in AusE as early as the nineteenth
century.11 The lack of comment on fricated variants in the older (and indeed in
the contemporary) literature may then reflect the fact that, on the whole, native
speaker-hearers are insensitive to these forms.
Prior to considering a possible dating for the emergence of glottalised
forms in AusE, it is necessary to consider a point of terminology which has led
to considerable confusion in the past. In the sociolinguistic literature, the term
“glottalling” has typically been used to refer to the variable process which
results in the hearer’s percept (impression) of [ʔ] (a glottal stop) in English /t/
contexts such as in knit, accelerate, catkin, hitman, Saturday, getaway and
tomorrow. This process has been recorded in a wide range of accents of
English such as London, Edinburgh, Glasgow, Newcastle, rural South East
England and East Anglia in Britain, and General American in the USA,
although the distribution and frequency of glottalisation in these different
forms of English is very diverse. However, on closer analysis of the data, it
turns out that “glottalling”, strictly a process whereby /t/ completely loses its
oral articulation and [ʔ] prevails, is in fact only one of a range of causes of an
auditory percept of a glottal stop.
The use of “glottalling” as it is traditionally used, to mean “glottal
replacement of /t/ by [ʔ]” (i.e. a process where the product is a canonical [ʔ]),
is theoretically loaded. The assumption of a replacement or substitution process
entails allowing for underlying representations located at a phonologically
deeper level. Such a concept of phonological organisation is an integral part of
many modern models which have emerged from the Generative tradition, e.g.
Lexical Phonology and Morphology (e.g. Kiparsky 1982; Mohanan 1986;
Booij and Rubach 1987). However, a number of recent developments in
phonological theory reflect a concern with simplicity and with the problem of
excessive abstractness (distance from stored to surface forms); they retreat from
the generative notion of a multi-levelled, level-ordered phonology. Two such
approaches are Declarative Phonology (Coleman 1992), a phonetic exponence
theory assuming that surface forms are simultaneously retrieved from the
lexicon and articulated, and the universalist approach called Government
Phonology (Kaye, Lowenstamm and Vergnaud 1985, 1990). Substitutions or
replacements are not licit procedures in models such as Declarative Phonology
and Government Phonology, and, pre-theoretically at least, it is then preferable
to use phonetically accurate descriptions to refer to the data.
“Glottalling” is not only a theoretically loaded term, but is also phonetically inaccurate. “Glottalling” implies that the phonetic product is always a
canonical glottal stop. As has been mentioned, there are in fact a continuous
range of products of “glottalling”, all of which promote the auditory impression of a glottal stop, but very few of which resemble a canonical glottal stop
either articulatorily or acoustically (see Docherty and Foulkes 1995, on
Tyneside English). The acoustic results from the current survey (see Section 5) identify no lone glottal stops, but instead a range of glottalised variants, involving a glottal gesture in addition to an alveolar closure gesture for
plosive [t] (see Section 6).
The relative timing of the glottal gesture and alveolar closing gesture in
glottalised variants is highly variable: the glottal gesture may precede, coincide with, or follow the alveolar gesture. Where the glottal gesture precedes
the alveolar, it hides the alveolar closure phase. Where the gestures coincide,
the glottal gesture may hide the alveolar gesture. Where the glottal gesture
lags behind the alveolar, the release phase of the alveolar occlusion will be
hidden. Gestural overlap is gradual, not discrete, thereby allowing in principle
for an infinite array of possibilities. These variants are difficult to describe
using standardised IPA format, since IPA notation implies discrete states of
affairs. However, [ʔt] has typically been used for pre-glottalised forms, [tʔ] for
post-glottalised forms, and [ʔ] for a canonical glottal stop.
The range of processes which result in an auditory percept of [ʔ] are thus
probably best referred to as “glottalisation” processes, thereby allowing for a
range of products which can then be analysed phonologically in a nonconclusion driven manner. Henceforth in this chapter, the term “glottalisation”
will be used for convenience for processes which produce a glottal percept.
Glottalisation of the other voiceless stops /p/ and /k/ in similar reduction
processes can be heard in many dialects of English (in London, for example,
stopcock may be pronounced [stɑʔkhɑʔ] or [stɑʔpkhɑʔk]: see Tollfree, 1999).
There may be a visual cue (i.e. a labial gesture) in the case of /p/ to distinguish
it from /t/ and /k/, but typically pairs such as pack-pat will be rendered
homophonous when the /k/ and /t/ are replaced (as opposed to reinforced) by a
glottal gesture. Glottalisation of /p/ and /k/ also arises in AusE: results from
the current survey showed that glottalisation of /k/ was frequent in a number of
items, and categorical, or near-categorical, for /k/ in the item like in the speech
of the focal (15–16 year old) informants. Occasional glottalisation of /p/ was
also found, e.g. in couple.12
Estimating a dating for the emergence of glottalised forms of /t/ in AusE
is extremely hard, given the relatively sparse documentation of this variety of
English compared to forms such as British English. In Britain, [ʔ] was first
recorded as replacing /t/ in Scotland by Alexander Melville Bell in 1860, in
the North of England by A. J. Ellis in 1908, and in London by Daniel Jones in
1909; in America, [ʔ] was recorded in New York by E. H. Babbitt in 1896;
meanwhile glottally-reinforced /t/ (as opposed to glottally-replaced /t/) was
first recorded in Glasgow (Scotland) by Henry Sweet in 1908 (see the comprehensive survey in Andrésen 1968). In addition to information in abundance in
the linguistic literature, there is plenty of supporting evidence for glottalised
forms of /t/, especially for London accents, from the literature written in
dialect, for example by George Bernard Shaw, G. Noël-Armfield and Julian
Franklyn. In Britain, [ʔ] was traditionally stigmatised in contexts where it is
noticeable by speaker-hearers, but is now more acceptable for some groups or
regions (see Rosewarne 1994 on “Estuary English” and Mees 1987 on Cardiff
English; see also Milroy, Milroy and Hartley 1994).
Meanwhile, in Australia, there is no explicit mention of glottalised forms
in linguistic works until Ingram (1989). There is also no special information
available from the literature written in dialect, e.g. by the Australian popular
poet C. J. Dennis, who was writing around the turn of the century. Dennis
clearly indicates features such as /h/-loss, /ŋ/ → /n/, /nd/ cluster reduction and
vowel reduction by orthographic means, but suggests no dialect-specific
realisations of /t/ in any context:
Life’s wot yeh make it; an’ the bloke ‘oo tries
To grab the shinin’ stars from out the skies,
Goes crook on life, an’ calls the world a cheat,
An’ tramples on the daisies at ‘is feet.
From “The Mooch O’ Life”, Tales of a Sentimental Bloke, circa 1900.
By contrast, glottalised forms were so prevalent in Britain by the latter part of
the nineteenth century that writers had begun to reflect the “glottal catch” in
their writings (for example, with “' ” by Bell in 1867 for Scots English, with
“;” by A. J. Ellis in 1875, and with “x” by Henry Sweet in 1877: Andrésen
1968). It is highly likely that if glottalisation were in evidence in AusE of
Dennis’s time, he would have exploited this orthographically in his works.
In sum, there can be no precise dating of a glottalisation development in
AusE. Stronger predictions can be made on phonetic bases, however. It has
already been noted that there are two main mechanisms producing voicelessness in /t/; one is vocal fold abduction which produces [t], and the other is
vocal fold adduction which produces glottalised variants. The choice of
devoicing strategy for segments such as /t/ is not exclusively under the control
of the speaker, as might be expected, but is phonetically conditioned. For
instance Ohala (1983) and Ohala and Ohala (1993) argue that glottals emerge
adjacent to sonorants since an adduction strategy for devoicing is favoured
over that of abduction in the context of nasals, in particular (and possibly
laterals and other sonorants).13 In addition, glottalisation arises not only in the
sonorant context in AusE, but in a number of other environments, suggesting
that (assuming it arose in the sonorant context) the process has spread across
phonological environments. Further, native speaker-hearers of AusE are
largely insensitive to glottalised variants in all of these contexts. Given these
facts, it is likely that glottalised forms have been an established part of AusE
for considerably longer than is represented by its earliest documentation.
In the more recent linguistic literature, however, phonetic variants of /t/
have been more closely documented. Voiced tapping (a well established
process in General American, for example) is recorded for AusE by Wells
(1982) (he describes it as a variable lenition of intervocalic /t/ to a “devoiced
/d/”: we shall see in Section 6.1 that this is phonetically inaccurate); Harris
(1984: 121) reports “vocalisation to tap ɾ”. Horvath (1985) also documents a
[ɾ] variant for /t/ in Sydney, and suggests that tapping is more typical of male
speech, especially in “Broad” varieties. Tapping is also recorded by Clark
(1989), Ingram (1989) and Haslerud (1995), who, in her sociolinguistic study
of (t) in Sydney adolescent speech, reports a high occurrence of [ɾ] as a nongender-linked informal variant.
Frication/affrication of /t/ is discounted in the surveys in Mitchell (1946)
and Turner (1972), and goes unmentioned in Mitchell and Delbridge (1965),
Baker (1966) and Wells (1982). Horvath (1985) records an affricated
(plosive) variant in Sydney, notably in speech of females, and those of Greek
ethnic background. She suggests that affrication is a recent development in
Sydney speech. A small proportion of affricated variants are recorded by
Haslerud (1995) who, like Horvath, assumes that affrication is a recent innovation in Sydney. She proposes that the affricated variant emerged in the
formal register speech of higher status girls, and remains a female and highstatus variant in Sydney.
As noted, the first report of glottalised variants in AusE is in Ingram
(1989). Haslerud (1995) finds a large proportion of glottalised variants in
Sydney, and describes [ʔ] as a recent, low-status, non-gender linked development, and as the least formal variant of /t/.14
The following section outlines the method used to obtain the data base of
the current survey.
Data base
The data were obtained from two surveys in Australia. The 1993–4 survey
was carried out for the purpose of another project (Tollfree 1996) in which
tape-recorded conversations and Labovian-type interviews were obtained
from informants from urban Adelaide, Canberra, Melbourne and Sydney. The
Melbourne material formed part of a pilot study for the current survey. The
bulk of the material, however, was collected in 1996–8 as part of the largescale Dimensions of Spoken Australian English project currently under way in
the Department of Linguistics at Monash University. Fieldwork was carried
out both in metropolitan Melbourne and in the Western District of Victoria.
Tape-recordings were made of two conversations between a researcher and a
focal informant of 15–16 years, and self-recorded conversations were made
between the focal informant and same age friends and between the informant
and members of their family. One of the interviews with the researcher
included a reading task involving words and sentences containing sociolinguistic variables of a phonetic nature, amid a range of decoy items. There are
55 focal informants in the sample (all of whom satisfy a criterion of having
had most of their schooling in Victoria), and in addition around 5 friend or
family informants per focal informant. The focal informants were selected
from nine schools located in Melbourne metropolitan districts of varying
socioeconomic status, boys’, girls’ and co-educational, state, Catholic, Anglican/Protestant schools, and a Jewish and a Greek Orthodox school.15
The current survey focuses on parts of Victoria, and the patterns of /t/
behaviour observed cannot be assumed to hold precisely for other regions (for
example, in the earlier survey, marked differences were found to exist between Victoria and South Australia in both the frequency and distribution of
/l/-vocalisation: Tollfree 1996).
The principal auditory results are detailed in Section 5 below.
Phonetic variants of /t/: auditory analysis
The material was transcribed auditorily in the first instance (see Kerswill and
Wright 1990 for an evaluation of this technique). A sub-part of the recordings
were re-analysed after an interval to provide a measure of consistency. The
material supplied tokens of the target sound in great excess of the minimum
number required to investigate each variable, which is 30 (Guy 1980; Milroy
1987). This minimises the chance of random fluctuation. Full-scale statistical
analysis was not undertaken for this project, but a sub-part of the data base was
analysed and the patterning of variables in conversational style expressed
quantitatively. The number of tokens of each variable were supplied as a
percentage of the total possible occurrence. Results were taken for /t/ in four
contexts: intervocalic medial, intervocalic final, pre-consonantal final and
pre-pausal final, in two speaking styles, conversational (informal) and wordand sentence-list (formal). For each context, 36 tokens were taken, from each
of 2 groups of focal informants (15–16 year olds) of differing socioeconomic
background, A and B (one lower socioeconomic group and one middle
socioeconomic group). Thus the total for each group is 144, and the overall
total is 288 tokens of /t/. In the case of pre-syllabic contexts, for which there
were fewer tokens available, results were compiled in terms of absolute
A range of each of the variants were scanned spectrographically (the
results are presented in Section 6). In addition, where the auditory percept
was ambiguous or unclear, which is frequently the case with fricated/fricative
and glottalled/glottalised variants, more detailed acoustic analysis was carried
out to confirm/disconfirm the auditory impression. Some important points of
disparity were found between auditory and acoustic analyses of the variable
data. These, and other practical problems associated with the analysis of /t/,
are considered in Section 6.
The auditory phonetic results obtained from the current survey for
tapped, fricative and glottalised variants of /t/ are now presented in Section 5.
Selected quantitative results are given in Tables 1 and 2; the full range of data
is assessed in Sections 5.1, 5.2 and 5.3.
In the following Tables, figures in bold type indicate data from formal
style as opposed to conversational style.
Table 1. Group A (lower socioeconomic group)
% auditory
Pre-pausal final
Table 2. Group B (middle socioeconomic group)
% auditory
Pre-pausal final
5.1. Voiced tapped /t/: auditory analysis
Results from assessment of the full data base showed that the tap [ɾ] was blocked
in foot-initial onset positions (that is, in prosodic positions where the preceding
syllable is less prominent than the following), e.g. Tasmania, television,
attempt, guitar. The tap was, however, a frequent realisation of /t/ in non-footinitial onset contexts generally (i.e. where the preceding syllable is more
prominent than the following). It occurred in intervocalic final contexts e.g. lot
of, get up, and medial contexts, e.g. bitter, mutter. A number of lexicallyspecific items were identified (these are items where, for many speakers,
tapping occurs categorically), e.g. attitude, beauty, beautiful, theoretical,
automatic, data, city. Preceding syllabic laterals (bottle, subtle) and syllabic
/n/ (mutton, baton) there were two variants, the plosive [t] and [ɾ]; [ɾ] was found
to be near-categorical preceding syllabic laterals in younger speech.
Statistical results showed that for Group A (lower socioeconomic group
15–16 year olds), in conversational style, tapping was high (93%) in (nonfoot-initial onset) intervocalic medial contexts (decreasing in the formal style
to 75%). The incidence was lower in intervocalic final position (59%), but still
preferred over glottalling (41%) in this position. Tapped initial /t/ occurred in
what is probably part of a restricted lexical set of to items: a number of
instances of this were found, (for example go to), apparently restricted to the
conversational style. Because of the low incidence of the pre-syllabic contexts, these results are expressed in terms of absolute values. In the 7 tokens of
/t/ in pre-syllabic /l/ contexts, the two expected variants were found: [t] (1
instance) and [ɾ] (6 instances). In the 24 tokens of pre-syllabic /n/ contexts,
only [ɾ] was found.
For Group B ( middle socioeconomic group), tapping occurred in 89% of
(non-foot-initial onset) intervocalic medial positions (this incidence was reduced in the formal style to 60%). Tapping occurred at 25% in intervocalic final
contexts (the remaining 75% being glottalised). This increased dramatically to
66% in the formal style, where the remaining 34% were realised as [t]. In the 12
tokens of /t/ in pre-syllabic /l/ contexts, there were 3 instances of [t] and 9 of [ɾ].
In the 6 tokens of pre-syllabic /n/ contexts, again, only [ɾ] was found.
5.2. Fricated /t/: auditory analysis
Results from assessment of the full data base indicated that the fricative
[ts] was most frequent in older informants’ speech, and in intervocalic and prepausal contexts, especially for slower speech rates. The auditory percept was
that of a true fricative (rather than a plosive with fricated release), and the
extent of audible frication heard was highly variable (see Section 6).
Statistical results showed that for Group A (lower socioeconomic group
15–16 year olds) the fricative was found only in formal speech in pre-pausal
final contexts, where its incidence was 13% (with 74% glottalised and 13%
plosive forms).
For Group B (middle socioeconomic group), the incidence of fricated
forms was 22% in pre-pausal final contexts (where the remainder were plosive
[t]), but the fricative was also found at 12% in intervocalic medial contexts
(increasing in formal styles to 20%).
5.3. Glottalised /t/: auditory analysis
Results from assessment of the full data base showed that glottalisation
(determined by an auditory percept of [ʔ]) was governed by prosodic relationships: it was blocked in foot-initial onset positions generally (as was tapping)
e.g. Tony, attack. Glottalisation was found to be high in pre-consonantal
position generally (regardless of morphology/syntax): in medial contexts with
no internal morpheme boundary (“unanalysable” or “underived” items), e.g.
cutlass; in medial contexts with internal morpheme boundary (“analysable” or
“derived” items), e.g. hitman. Glottalised forms were occasionally attested in
intervocalic final contexts, e.g. lot of, get out (where it was of higher incidence
in younger speech than in older), but were apparently blocked from occurring
in intervocalic medial contexts, e.g. bitter. Glottalised forms were attested
before a pause, and in post-resonant position, notably preceding consonantinitial items, e.g. different, Celt from, pelt with, paint that. Glottalisation was
either categorical or near-categorical in pre-syllabic /n/ contexts, e.g. Martin,
cotton, smitten. However, for /t/ in pre-syllabic /m/ and pre-syllabic /l/ contexts, glottalisation was not attested.
Statistical results showed that for Group A (lower socioeconomic group
15–16 year olds) glottalised forms of /t/ were categorical (100%) in pre-pausal
final position, but their incidence was reduced to 13% in the more formal style
(where the same incidence of fricated forms arose: the majority 74% were the
plosive [t]). Glottalisation was near-categorical (98%) in pre-consonantal final
position (with [t] at 2%); the balance shifts slightly in the formal style to 86%
glottalised forms (14% [t]). The incidence of glottalisation was 41% in
intervocalic final contexts (the other variant in this position being the tap which
is slightly preferred at 59%). Out of the 51 total possible tokens of /t/ following
/n/, 3 instances were of [t] and the vast majority, 48, were of glottalised forms.
For Group B (middle socioeconomic group), glottalisation was categorical in pre-consonantal final position, but reduced to 41% in formal style
(where the remaining forms were [t]). Equal proportions of [t] and glottalised
variants were found in pre-pausal final contexts (with the balance shifting in
the formal style to 78% [t] and 22% [ts]). Out of the 38 total possible tokens of
/t/ following /n/, 9 instances were of [t] and 29 were of glottalised variants.
5.4. Discussion of principal auditory findings
This section discusses the main trends of the data and speakers’ attitudes to the
data, and provides some suggestions for possible future change. For the 15–16
year age group generally, glottalised variants appear to be competing with the
tap in intervocalic final position (this confirms the findings of the survey in
Tollfree 1996). The fricative was not found in intervocalic final contexts (as
was the case for older informants in Tollfree 1996), being heard only in prepausal final, and intervocalic medial positions. Many informants’ speech did
not feature fricative variants at all. The age-related material suggests that
fricated forms are losing ground in Australia.
In regard to speaking style, the most dramatic difference between the
conversational and formal styles was found for Group B (middle socioeconomic group) in pre-consonantal final contexts, where glottalisation dropped
from 100% in the conversational style, to 41% in the formal style. This
suggests that speakers have at least some control over the mechanisms involved. Another noticeable pattern for Group B was that intervocalic final
tapping was favoured over glottalisation in the formal style, suggesting that
the tap might be the more acceptable variant in formal (as well as conversational) speech styles.
It is apparent that speakers are largely insensitive to alternation between
[t] and the fricative [ts], and, even after they have had the variation pointed out
to them, are not well equipped to discriminate between them. As mentioned
above, speakers are also typically insensitive to glottalised forms, even in the
intervocalic final context, whilst they readily recognise [ʔ] in items such as
butter as spoken by London (and other) English speakers.
Informants from the 1993–4 survey (Tollfree 1996) were also asked to
provide information on their percept of the range of /t/ variants they used (this
was conducted at the end of the interviews to avoid biased results). A general
pattern of speaker-hearers’ preferences became apparent, suggesting a formality hierarchy of: plosive [t] (most formal, “correct”) > fricated [ts] > voiced
tapped [ɾ] > glottalised forms. A number of informants reported an association
of [ɾ] with a “street-wise”, laid-back image (a fact which, if substantiated,
might maintain it). Two informants reported that [ɾ] is “normal” in many
items, especially those which they called “technical” (such as theoretical,
data, automatic etc.). Perhaps the popularity of the tap reflects the apparent
increasing prestige of American (as opposed to British) characteristics in
attitude, dress and behaviour in the young. Since the tap is apparently
favoured over glottalised forms, and since the processes occur in the same
environments, it seems unlikely that there will be a general development to
intervocalic medial glottalisation in AusE.
AusE speakers are in general insensitive to the effects of frication, and
even more so to the effects of glottalisation.16 Typically, speakers are not well
placed to discuss frequencies and distributions of the range of variants of /t/,
although they are sometimes aware of voiced tapping, describing it as a sort of
“d” sound, and can supply examples where the tap is attested in their own
Section 6 discusses the results of the acoustic analysis of tapped,
fricated/fricative and glottalised variants of /t/ in the survey.
Phonetic variants of /t/: acoustic analysis
An acoustic profile was constructed of each of the range of auditorily
perceived tapped, fricated/fricative and glottalled/glottalised variants of /t/ in
medial and final position in items from both the formal and informal speech
styles. Signals from the tape recordings were digitised at a sampling rate of
20 000 Hz using a Kaye Computer Speech Laboratory Program. The signal
was analyzed into wide band spectrographic format and inspected visually.
The results for tapped, fricative and glottalised variants are given in Sections
6.1, 6.2 and 6.3 respectively.
6.1. Voiced tapped /t/: acoustic analysis
The alveolar tap [ɾ] is produced by an essentially momentary contact
between articulators: the tongue tip swiftly flicks or taps the alveolar ridge,
and immediately moves away. Instrumentally, a picture somewhat similar to a
voiced alveolar stop might be expected: a decrease in the formants (F1 and F2)
entering the tap slot (indicating alveolar closure, as for [t] or [d]), a gap in the
pattern (corresponding to complete occlusion), followed by a sharp beginning
of formant structure (as for voiced stops, rather than the burst of noise
expected for voiceless stops). Vertical striations would be expected throughout (corresponding to vocal fold vibration for voicing). A tap is likely to be
characterised by a somewhat shorter onset medial and offset phase than might
be expected for a voiced stop.
Analysis of a range of tapped variants in the survey provided acoustic
displays similar to this description. Typically, there was only very slight
lowering of the formants entering the tap, an evident gap phase, and a clear
finish and start to the formant structure of adjacent vowels. There was evidence of voicing throughout the tap in each case. The display for some taps,
however, did not clearly indicate alveolar articulation. In other words, some
displays lacked clear evidence of lowering of the formant transitions into and/
or out of the tap slot.17
The only source of noise during articulatory closure, for aerodynamic and
acoustic reasons, is voicing. Voicing is aerodynamically impeded, though not
prevented, during stop closure, and often fades away by the end of the closure
phase in stops. However, there was no evidence of fading in the taps examined, possibly because the medial (contact, or “tap”) phases were too brief.
This means that Wells’ (1982) impression of tapping as “variable lenition of
intervocalic /t/ to a devoiced /d/” is inaccurate (see Section 3.2 above).
6.2. Fricated /t/: acoustic analysis
The aim of the acoustic analysis of perceived fricated forms was to
determine whether these variants were characterised by a clear closure and/or
release phase (i.e. plosives with fricative release), or by the lack of a closure
phase (“true” fricatives), or both (a mixture of plosives and fricatives). The
acoustic results showed that there was a mixture. There were plosives with a
fricative release phase on the release of the alveolar obstruction, and fricatives
characterised by the absence of a period of complete closure. The fricatives
generally showed a continuation of formant structure throughout the majority
of the /t/ slot, and also sometimes had voicing throughout (for example, in prepausal Ballarat). There were no indications that the phonological or morphological context conditioned the choice between plosive with fricated release
and fricative /t/.
6.3. Glottalised /t/: acoustic analysis
Very few acoustic analyses of glottalled/glottalised /t/ are found to reveal a
canonical glottal stop (see Section 3.2 above). Thus it was perhaps not
surprising that none of the variants in the Australia survey resembled a
canonical glottal stop on acoustic analysis. The analysis of perceptually
“glottalled” /t/s revealed instead a range of glottalised variants. The most
frequent of these were characterised by an absence of formant transitions
entering or leaving the stop, indicating that they either involve no oral gesture,
or if there is an oral gesture, it is hidden by a reinforcing glottal gesture. The
displays were ambiguous: there was no stop release burst to confirm the
presence of an alveolar gesture (and, therefore, a build-up of air pressure), but
this could be due to masking by a glottal articulation.18 These variants had an
identifiable stop gap, but one which was never entirely silent. The difficulty in
identifying the tongue tip contact in glottalised variants of alveolar /t/ is well
known (e.g. Pierrehumbert 1995).
Many of the variants with no clear indication of an oral gesture also
showed evidence of laryngealised voicing throughout the stop gap (these are
seen as irregular peaks, produced by increased glottal tension or constriction).
Laryngealisation often occurs as an intermediate stage between voicing and a
glottal stop (e.g. Catford 1977), and it is difficult to distinguish the striations
pertaining to intervals of irregular fold vibration. Unfortunately, even where
this is possible, it is often not possible to distinguish between the different
types of vocal fold articulation giving rise to laryngealisation.19 Docherty and
Foulkes (1995) report that in their study of glottally-reinforced stops in
Tyneside English, a high number have laryngealised voicing throughout the
interval where the “glottal articulation” is perceived — in some cases all that is
required to provide a percept of glottal articulation is one or two pulses of
voicing which are slightly irregular with respect to neighbouring pulses: this is
likely to be the case in at least some of the Melbourne examples.
Concluding remarks
The auditory and acoustic survey of /t/ identified, in addition to plosive /t/, a
high quantity of reduced forms common to everyday Victorian AusE speech.
A small number of fricated and fricative variants of /t/ were found, but there
were strong indications from the age-related material that these are losing
currency. The most prolific reduced variants were tapped and glottalised
forms, which were found to be conditioned by phonological, morphological
and prosodic context. In contexts where both tapping and glottalisation arose,
tapping was strongly favoured over glottalisation.
Clearly the growing corpus of variable material from the Dimensions of
Spoken Australian English project can be usefully employed in a number of
areas of sociolinguistic and phonological analyses. It is already being put to
use in the development of techniques in forensic speaker identification,
speaker profiling, content determination and related procedures in forensic
phonetics and linguistics, with special reference to the Victorian context. It is
assisting in the investigation of some important issues, for example the auditory and acoustic phonetic parameters of inter- and intra-speaker variation,
especially in the case of perceptually similar voices, and in establishing a data
base of the distribution of values amongst the local population. It will also be
valuable in tracking future change in AusE.
I am indebted to the Leverhulme Trust for the Study Abroad Studentship which has
enabled me to carry out research in Australia in 1996–1998.
I am grateful to the British Academy for a Major State Studentship which financed the
British-based pilot study of AusE speakers resident in Britain. The Australian fieldwork
(1993 -1994) was in part funded by the Smuts Memorial Fund (University of Cambridge).
The Departmental project has benefited greatly from financial assistance from the
Monash Research Fund.
With one exception: tapping of initial unstressed /t/ is sometimes found when the /t/ is
intervocalic and where the preceding syllable bears greater prominence than the following, e.g unstressed to (see Section 5.2).
AusE also exhibits variable /h/-loss, use of [f] and [v] for /ð/ and /θ/, glottalisation of /p/
and /k/ (see 5.4), glide formation including linking- and intrusive-/r/, velarisation/
pharyngealisation of /l/ and /l/-vocalisation (see Borowsky, this volume, for a treatment
of AusE /r/ and /l/).
Whilst there is evidence to suggest that Irish English has had some influence on AusE
(see Section 3.2), there is little indication that Irish Gaelic has been spoken en masse or
maintained in Australia by the second and later generations. It is also likely that AusE had
already been established before the main influx of Irish English speakers in the 1830s,
thereby minimising the linguistic effect Irish English could have had on the extant
Other factors, such as migration by speakers of languages other than English, have had an
increasingly important influence on the development of modern AusE (see, for example,
Clyne, Eisikovits and Tollfree, this volume).
In a similar way, the /l/-vocalisation in contemporary London and Australian Englishes
must be viewed as separate innovations (Tollfree 1996).
For example, London English has glottal variants of /t/ in word-initial contexts e.g.
tomorrow (and other members of a restricted lexical set beginning with to-, [thɘ]), before
syllabic laterals e.g. bottle and in intervocalic medial contexts, e.g. water. In these, more
noticeable environments, they may be stigmatised for some speakers (Tollfree 1996, and
in press). Before syllabic nasals, however, speakers are typically insensitive to glottalised
variants and the stigma is minimal or absent, e.g. button, Martin.
Wells (1982: 594) comments that AusE lacks /t/-glottalling and /l/-vocalisation. However, /t/-glottalling/glottalisation and /l/-vocalisation were reported as widespread in
Australia not long after Wells’ survey (Tollfree 1996).
There is some counter-evidence to this suggestion. Features of Irish English, such as the
presence of dental variants of /t/, have pervaded Liverpool English since the mass
migration of Irish to Merseyside in the 1840s. AusE, by contrast, does not have dental
variants of /t/. The assumption that AusE /t/-frication is Irish in origin fails to explain the
phonetically similar variants of /t/ in regions of, for example, Britain, which have no
special history of Irish immigration, such as Tyneside.
Statistical and acoustic analyses of /p/- and /k/-glottalisation are currently being carried
out as part of the Monash University Dimensions of Spoken Australian English project
(Tollfree, in preparation).
Note that preference for adduction as a devoicing strategy implies that a glottalised form
would necessarily precede [ʔ] in glottalling developments in these contexts. However,
there is substantial counterevidence for this from a range of language varieties, e.g.
Tyneside English (Milroy, Milroy and Hartley 1994), West Yorkshire English (Harris
1994) and Malay (Farid 1980).
See also Holmes (1994) on voiced flapping of /t/, and Holmes (1995) on glottalisation of
word-final /t/, in New Zealand English.
Some notable differences between “mainstream” AusE and Greek AusE are discussed in
Clyne, Eisikovits and Tollfree (this volume).
Notice that the use of laryngealisation is more widespread than this. Glottal onsetting in
pre-vocalic contexts and “creaky voice” are frequent characteristics in the speech of
many of the 15–16 year old informants in the Dimensions of Spoken Australian English
project: the frequency and distribution of these, and speakers’ attitudes towards them, are
currently being investigated (Tollfree, in preparation).
Investigation of the relative duration of onset, medial and offset phase of these taps, and
of the tapping of fricated variants, is currently under way as part of the continuing
Dimensions of Spoken Australian English project.
Care should be taken in interpreting the burst: both glottally reinforced (ejective) stops
and oral stops exhibit a burst phase.
Many of the variants showed evidence of laryngealised voice on the vowel following the
/t/ slot — this and glottal onsetting are noticeable features of the speech of a number of
the focal informants.
Andrésen, B. S. 1968. Pre-glottalisation in English Standard Pronunciation. Oslo: Norwegian Universities Press.
Baker, Sidney J. 1945. The Australian Language. Milson’s Point, New South Wales:
Currawong Press.
Booij, G. and J. Rubach. 1987. “Postcyclic versus postlexical rules in Lexical Phonology”.
Linguistic Inquiry 18: 1–44.
Catford, John Cunnison. 1977. A Practical Introduction to Phonetics. Oxford: Clarendon
Clark, John E. 1989. “Some proposals for a revised phonetic transcription of Australian
English”. In Peter Collins and David Blair, eds. Australian English. The Language of a
New Society. St Lucia: University of Queensland Press, 205–214.
Coleman, John. 1992. “‘Synthesis-by-rule’ without segments or rewrite rules”. In G. Bailly
and C. Benoit, eds. Talking Machines: Theories, Models and Applications. Amsterdam:
Elsevier, 43–60.
Docherty, Gerry and Paul Foulkes. 1995. “Acoustic profile of glottal and glottalised
variants of English stops”. Proceedings of the XIIIth International Congress on Phonetic Sciences I, 350–353.
Farid, M. O. 1980. Aspects of Malay Phonology and Morphology: a Generative Approach.
Bangi: Universiti Kebangsaan Malaysia.
Guy, Gregory R. 1980. “Variation in the group and the individual: the case of final stop
deletion”. In William Labov, ed. Locating Language in Time and Space. New York:
Academic Press, 1–36.
Harris, John. 1994. English Sound Structure. Oxford: Blackwell.
Haslerud, Vibecke C. D. 1995. “The variable (t) in Sydney adolescent speech: A sociolinguistic study of phonological variation”. Cand. Philol. dissertation, University of
Bergen, Norway.
Holmes, Janet. 1994. “New Zealand Flappers: an analysis of T Voicing in New Zealand
English”. English World-Wide 15: 195–224.
———. 1995. “Glottal stops in New Zealand English: an analysis of variants of word-final
/t/”. Linguistics 33: 433–463.
Horvath, Barbara. 1985. Variation in Australian English: The Sociolects of Sydney. Cambridge: Cambridge University Press.
Ingram, John C. L. 1989. “Connected speech processes in Australian English”. Australian
Journal of Linguistics 9: 21–49.
Kaye, J., J. Lowenstamm and J.-R. Vergnaud. 1985. “The internal structure of phonological
elements: a theory of charm and government”. Phonology Yearbook 2: 305–328.
———. 1990. “Constituent structure and government in phonology”. Phonology 7: 193–
Kerswill, Paul and Susan Wright. 1990. “The validity of phonetic transcription: limitations
of a sociolinguistic research tool”. Language Variation and Change 2: 255–276.
Kiparsky, Paul 1982. “Lexical Phonology and Morphology”. In I.-S. Yang, ed. Linguistics
in the Morning Calm. The Linguistic Society of Korea. Seoul: Hanshin, 3–91.
Mees, Inger. 1987. “Glottal stop as a prestigious feature in Cardiff English”. English WorldWide 8: 25–39.
Milroy, Lesley. 1987. Observing and Analysing Natural Language. Oxford: Blackwell.
Milroy, James, L. Milroy and S. Hartley. 1994. “Local and supra-local change in British
English: the case of glottalisation”. English World-Wide 15: 1–33.
Mitchell, Alexander G. 1946. The Pronunciation of English in Australia. Revised 1965.
Sydney: Angus and Robertson.
——— and A. Delbridge. 1965. The Speech of Australian Adolescents: A Survey. Sydney:
Angus and Robertson.
Mohanan, K. P. 1986. The Theory of Lexical Phonology. Dordrecht: Reidel.
Ohala, John J. 1983. “The origin of sound patterns in vocal tract constraints”. In P. F.
MacNeilage (ed.). The Production of Speech. New York: Springer Verlag, 189–217.
Ohala, John J. and M. Ohala. 1993. “The phonetics of nasal phonology”. In M. K. Huffman
and R. A. Krakow, eds. Phonetics and Phonology 5: Nasals, Nazalisation and the
Velum. New York: Academic Press, 225–250.
Pierrehumbert, J. 1995. “Prosodic effects on glottal allophones”. O. Fujimura and M.
Hirano (eds.). Vocal Fold Physiology 8. San Diego: Singular Press. 39–60.
Rosewarne, David. 1994. “Estuary English–tomorrow’s RP”. English Today 37. 3–9.
Tollfree, Laura. 1996. “Modelling phonological variation and change: evidence from
English consonants”. PhD dissertation, University of Cambridge.
———. 1999. “South East London English: discrete vs continuous modelling of consonantal reduction”. In Gerry Docherty and Paul Foulkes, eds. Urban Voices — Variation and
Change in British Accents. London: Arnold, 163–184.
———. In preparation. “The relationship between laryngealised voice and glottalisation of
/p, t, k/ in Australian English”.
Turner, George W. 1972. The English Language of Australia and New Zealand. 2nd
edition. London: Longman.
Wells, John. 1982. Accents of English. 3 Volumes. Cambridge: Cambridge University
The vocalisation of dark l
in Australian English
Toni Borowsky
University of Sydney
This paper is concerned with the phonological restrictions on the vocalisation
of dark /l/ in Australian English (henceforth, “AusE”).
In a preliminary analysis of the vocalisation of /l/ in AusE based on data
collected in Adelaide, South Australia, Borowsky and Horvath (1997) use
Optimality Theory to explain the variable linguistic patterns. There it was
proposed that the variation can be understood as a conflict between the
faithfulness constraints that require that the phoneme /l/ must be a consonant
in all instantiations, and markedness constraints (particularly syllable harmony constraints) which insist that segments in nuclear positions should be
The aim of our current /l/ vocalisation project is to extend the previous
study and consider additional factors in the linguistic and social patterning of
this change in progress across Australia and New Zealand. Our original
project has been extended to include a great deal more data from many more
places, including Brisbane, Melbourne, Sydney, Hobart and Mount Gambier.1
The data collection materials have also been extensively redesigned to include
phonological environments not considered in the earlier study. The range of
phonological factors which we considered is shown in Table 1 below.
The results reported in this paper are the first results of the variable rule
analysis of the phonological aspects of this study. Results from the social and
geographical side of the study have been reported in Horvath and Horvath
(1997). On the basis of the facts which emerge from this analysis we propose
an extension of our earlier explanation for the variants of l in AusE on the
basis of the interaction between the phonological representation of l and its
realisation in different syllabic environments.
Table 1. Phonological factors
dorsal+syll /l/
high/back V+/l/
high V+/l/
/l/ is clustered
long V+/l/
central V+/l/
front V+/l/
back V+/l/
coronal+syll /l/
/l/ is syllabic
short V+/l/
mid V+/l/
low V+/l/
/l/ is coda
low/front V+/l/
labial+syll /l/
milk, hulk
pickle, haggle, fickle, wiggle
cool, fool, school
boil, toil, foul, aisle
field, feel
feel ##sorry
hulk, girls, bald
hall,cool, meals
bottle, puddles, whistle, needle
horrible etc.
small, feel, Malcolm
bottle of
help, film
people, horrible, ripple, incredible
felt, field, bald, halt
Bi-gestural l and syllable harmony
The production of a lateral involves narrowing and lengthening the tongue
blade produced by retracting the tongue body, so that openings on the sides
may occur. This is combined with a closure forward of the retraction, usually
apical or palatal. Thus the consonant l has been described by phonologists as
well as phoneticians as a bi-gestural2 articulation involving a tongue tip or
apical gesture and a tongue body gesture. (See Ladefoged and Maddieson
1996; Walsh Dickey 1997; Blevins 1994; Sproat and Fujimura 1993.) This is
shown informally in diagram (1) where CORONAL is the tongue tip gesture
and DORSAL is the tongue body gesture.
Sproat and Fujimura (1993) is an articulatory study of the dark versus light l
allophony, in which it is proposed that the primary difference between the
light and dark l s in English is the timing of the dorsal gesture in relation to the
apical gesture. When a dark l is produced the dorsal gesture occurs earlier
than, and is maintained longer than the coronal gesture.
The timing of each gesture depends crucially on where in the syllable the
l occurs. Each of the gestures has a “strong affinity” (Sproat and Fujimura
1993: 291) for different parts of the syllable. The tongue body gesture is
inherently vocalic and therefore has an affinity for the syllable nucleus and the
tongue tip gesture is inherently consonantal and hence prefers the syllable
onset. If the /l/ is in the nucleus or near the nucleus of the syllable the tongue
body gesture precedes the tongue tip gesture considerably, with the effect that
the tongue body part of the articulation is enhanced or maximised while that of
the tongue tip is minimised: the result is a dark l. If the l is in the syllable onset
the timing of the coronal /tongue tip gesture is maximised while the dorsal /
tongue body gesture is minimised, hence the l is light.
Minimisation of the tongue tip gesture can be so great that it results
ultimately in the loss of the coronal gesture. This is what occurs in vocalisation. So the change from consonantal dark l to a vowel is due to the special
“affinity” of each gesture for different parts of the syllable. We formalise this
in terms of the notion of syllable harmony in Optimality Theory (henceforth,
OT; see Prince and Smolensky 1993, for example).
Prince and Smolensky 1993 propose a hierarchy of markedness of syllable types. The optimal syllable is that with the best (most consonantal) Onset
and the best (most vocalic) Nucleus, probably something like ta. Any other
syllable incurs some marks against it in the markedness hierarchy. The measure along which consonants are judged to be more or less consonantal is
sonority and stricture. The liquids r and l are thus less consonantal than stops,
say, or fricatives, because they are more sonorous and lack stricture. Vowels
are judged more or less vocalic by their sonority.
We add to this the idea that some articulatory gestures are more consonantal than others, as suggested by the Sproat and Fujimura account of l
allophony: the coronal (tongue tip) gesture is more consonantal than the dorsal
(tongue body gesture). This allows us to understand the Sproat and Fujimura
claim in terms of syllable harmony. The best nucleus is that which contains the
most vocalic element; and clearly in the circumstances in which a dark l
occurs in a syllable nucleus the dorsal gesture will be favoured since it is the
more vocalic gesture. The coronal gesture is marked; it is too consonantal for
a nuclear position. On the other hand the best onset is one with the most
consonantal gesture; thus the coronal gesture is less marked here while the
dorsal gesture is more marked in this environment.
Codas are disfavoured in syllables generally: ta is universally a better
syllable than tan (see Prince and Smolensky 1993; McCarthy and Prince
1993). In Borowsky and Horvath (1997) we assume that this is due to a
constraint which asserts that codas must be as sonorous as possible. Thus
codas which are less consonantal, and thus more sonorous, are favoured over
those which are less so. Clearly no coda at all is best by this measure. In the
case in which an l occurs in a coda a dark l is better than a light l and a
vocalised l even better. If the l occurs in a coda cluster the pressure to become
a vowel is even greater than in a simple coda, because it is shorter and closer to
the nucleus and subject to more stringent sonority restrictions.
We formalise these notions with the constraints below:
*dorsal/onset >>*coronal/onset
*coronal/coda (or nucleus) >> *dorsal /coda (or nucleus)
What these constraints say is that it is better to be a coronal in an onset or a
dorsal in a coda or nucleus than it would be to be a dorsal in an onset or a
coronal in a nucleus or coda. This is the formalisation of the Sproat and
Fujimura hypothesis.
When these constraints conflict with the faithfulness constraint IDENT l
which ensures that an underlying /l/ corresponds with an surface [l], vocalisation may occur. When IDENT l outranks the markedness constraints there is
no vocalisation. If the markedness constraints outrank the faithfulness constraint vocalisation occurs. See Borowsky and Horvath (1997) for a comprehensive explanation of variation in these terms.
For our purposes we assume the ranking of IDENT l below the two markedness constraints to explain the vocalisation cases. The same ranking holds for
the markedness constraints governing the timing of the features of the complex l when in an onset position; that is, the markedness constraints outrank
the faithfulness constraint and in such circumstances vocalisation does not
take place. In the form below (filly) the l is categorically an onset and
vocalisation does not occur. The overall ranking which results in this pattern is
given in (5a). The tableau (5b) illustrates how the correct output is selected by
the grammar for an onset l.
a. *place/ons hierarchy> IDENT l >*place/nuc or coda hierarchy
b. /fli /
We will show that the maximisation/minimisation of the articulatory gestures
is promoted in environments which share the appropriate gesture. Furthermore
we see that vocalisation is inhibited in those environments which independently promote the non-prime gesture. Place interacts crucially with syllabic
position, thus supporting the Sproat and Fujimura hypothesis.
Results and analysis3
In this section we present the preliminary results of our cross-city Goldvarb
analysis, simultaneously describing the phonology of l vocalisation.
The results of our preliminary Goldvarb analysis support in large part the
analysis of our previous study of Adelaide English in which we identified
three clear groups related to syllabic affiliation. There was however one
important difference: in the first Adelaide study the clustered environment
(eg. help) showed the least vocalisation while the syllabic environments
showed the most. This study has controlled for the factor which we believe
inhibited vocalisation in coda clusters in that study.4 We have found, as we
expected, that the coda cluster environment is the most likely position in
which vocalisation will occur, followed almost equally by the syllabic positions and coda l.
Table 2. The three groups
syllable factors
1. coda cluster
2. syllabic
3. coda
% vocalised
The coda l environment given above is deceptive, since it fails to distinguish
between coda l following a long vowel or diphthong from coda l following a
short vowel. The two environments correspond in fact to potentially different
syllable environments which have an important effect on vocalisation frequencies as shown in the table below. Coda l following long vowels or
diphthongs is vocalised more frequently than coda ls following short vowels.
Table 3. Preceding vowel length
coda l
Following long vowels, l tends to be syllabic. This is the breaking environment — see Borowsky and Horvath (1997) and McMahon et al. (1994). This
environment could be grouped together with the syllabic environment, with
the result that we have more frequent vocalisation in syllabic (i.e. pickle and
feel) environments as opposed to coda (i.e. full ) environment. After short
vowels ls are merely coda consonants, and so do not get the additional
promotion effect of the nuclear position.
Table 4. The three groups revised
syllable factors
coda cluster
coda A.
coda B
% vocalised
If we leave aside for the moment the clustered5 cases and consider only the
remaining environments, we find a clear distinction between those cases in
which l is, or has the potential to be, nuclear (coda A in 2 and syllabic in 3),
and those in which l is unlikely to be nuclear (coda B in 4). The former show a
greater propensity for vocalisation than the latter. Here we have clear support
for the notion that the vocalic gesture of the l is maximised or enhanced in the
nuclear syllable position while the consonantal gesture is minimised.
In each of these groups two additional factors play a role in vocalisation.
These were the following environment (whether it was followed by a consonant, a vowel or a pause) and the place of articulation of the adjacent segment
(whether it was labial, dorsal or coronal, or backness in vowels). We consider
each of these in turn.
3.1 Adjacent place of articulation
In both coda clusters and syllabic ls, place of articulation of the adjacent segment
is the primary factor in vocalisation. Recasens (1996: 71) reporting on various
acoustic and articulatory phonetic studies of preconsonantal /l/ observes:
…that /l/ should be darker (and thus ought to vocalize more frequently) when
coarticulated with following labials and velars as opposed to apicals and
palatals. Indeed, F2 of /l/ should be lower before velars than before apicals
and palatals…6
This is exactly what we find in our data for preconsonantal l summarised in
Table 5 below.
Table 5. Following consonant place
clustered /l/
What we see here is paralleled in the history of English. In the early Modern
English period [l] was lost between au and ou and following labials and
dorsals: talk, half, balm, folk etc. The clusters lt,ld are the only ones that still
occur after these vowels: e.g. halt, bolt, fold. In words of French origin l
developed and still remains only in words where it is followed by a coronal:
fault, assault, soldier.
The pattern is different for postconsonantal l though it should be noted
that the dorsals continue to exhibit a powerful promotory effect.
Table 6. Preceding consonant place
syllabic /l/
The tables show that while preceding coronals are weak promotors, following
coronals inhibit vocalisation. Preceding labials inhibit but following labials
promote vocalisation. Dorsals are very strong promotors of vocalisation
whether they precede or follow.
Let us consider now each of these two positions in turn. We begin our
discussion with the clustered l facts as shown in Table 5 above.
Gliding of l before a consonant in a cluster is due to the inherent
weakness of coda segments. A consonant in a coda is prone to weakening by
becoming more sonorant and losing stricture. We pointed out above that the
preference is for codas to be as sonorous as possible; so vocalisation, or
gliding, is a response to this pressure. The consonantal gesture is dispreferred
in these environments and minimised. A consonant in a cluster is phonetically
short which means there is less time for the articulators to reach their targets.
This environment is a prime one for vocalisation, even independent of place.
We therefore expect interesting interactions when combined with place.
Velar consonants are produced with a high back closure which causes an
adjacent /l/ to have a retracted tongue body, rather like that of the vowel [υ].
Even an onset l is ‘’darker” when flanked by high back vowels (Sproat and
Fujimura 1993). The combination of the surrounding dorsality along with the
syllabic environment leads to vocalisation.
The relationship between the dorsal and coronal gestures in a sequence
l +velar stop is schematised below. The boxes represent a gesture by the
articulator concerned. The arrangement of these boxes is intended to show the
relative timing of the gestures with respect to each other and to the syllable
structure. Dorsal is used for the tongue body gesture, Coronal for the tongue
tip gesture.7
Where a dorsal consonant follows the l, the dorsal gesture is maintained
through both segments. The coronal gesture of the l may or may not be
articulated by a speaker because the l is in a nuclear environment where the
consonantal gesture is minimised and the dorsal gesture is maximised. If the
coronal gesture is minimal or not effected at all we assume the l is “vocalised”.
These gestural diagrams serve to illustrate in an informal way the articulatory
timing relationships of the gestures concerned. We proposed above a phonological interpretation of these facts which translates the natural “affinity” of
the gestures for different positions in the syllable into ranked markedness
constraints as shown in the tableau in (3). The diagram above can be considered to be a phonetic realisation of the tableau.
We must now build into the picture the assimilatory effect of adjacent
shared place of articulation. To the tableaux above we add another constraint
which values shared place of articulation in codas.
SHARE/place: share place of articulation
If this constraint is ranked above our others, both the promotory and the
inhibitory effect of adjacent consonant place are captured. Consider first
adjacent dorsality. When this is shared between the l and the following
consonant the decision falls to the syllable markedness features which mark
coronality in this position.
*coronal/nuc *dorsal/nuc IDENT l.
Thus the frequency of vocalisation we observe with dorsals is due to the
coarticulatory effect in conjunction with the fact that the l is close to the
syllable nucleus and hence in a position in which it is preferentially sonorous.
The timing of the dorsal gesture is maximised for these two reasons with the
result that vocalisation is highly preferred in this environment.
Following coronals inhibit vocalisation. The articulation of a following
coronal requires a positive gesture of the coronal articulator. The only coronal consonants that can follow liquids in syllables in English are: n, s, t/d. Each
of these has a positive coronal gesture: a closure or almost complete closure. It
is clear that these requirements will inhibit the vocalisation of the preceding l
since anticipation of the following coronal closure will presumably mean
prolonging the closure of the l. To vocalise the l here would require careful
timing of the start of the following coronal gesture and no coarticulation.
However, since dark l is a complex segment with a coronal articulation of its
own, it is less likely that the coronal gesture would not be articulated at all
because of the anticipation of the following segment, even if the gesture is
minimised because of its syllable position. Thus we see here that the nonprime
gesture (i.e. the coronal part of the dark l in a coda position) is maximised for
assimilatory reasons, and this stops vocalisation.
Phonologically we can translate this as follows: SHARE knocks out the
candidate which does not share the coronal gesture.
☛ felt
Following labials8 neither inhibit nor promote vocalisation. Here vocalisation
is permitted because the labial articulations do not overlap with the articulation of the lateral. The syllable harmony factor promotes vocalisation unhindered. If a labial consonant follows the l, the timing of the dorsal gesture will
be maximised and that of the coronal gesture minimised because the l is close
to the nucleus. Vocalisation may occur and is unaffected by the adjacent
consonant which shares no place features with it.
In OT terms the two competing forms tie on the highest ranked share
constraint and the decision falls to the syllable harmony constraints below.
☛ fiwm
*coronal/nuc *dorsal/nuc
The tongue retraction and lowering, which is characteristic of the dark l before
consonants, is expected before all consonants except coronals where the
shared coronality inhibits loss of the coronal gesture. Hence we expect vocalisation before labials. These forms provide our strongest evidence for the effect
of syllable position on vocalisation, which shows that the vocalisation is not
merely assimilatory. Frequent vocalisation occurs in these clusters where the l
is least preferred as a consonant and the dorsal gesture is enhanced but where
there is no additional effect of the adjacent consonant.
In sum, vocalisation is most frequent in clusters before dorsals, because
here the promotory effect of the dorsal is added to that of the cluster environment which independently maximises the dorsal gesture. The coronal gesture
is minimised in these environments and vocalisation occurs. If a coronal
consonant follows the l this will have a negative impact on minimising the
timing of the coronal gesture, because the speaker is anticipating the full
consonantal gesture following. When a labial consonant follows it has no
effect whatever and syllable position alone is responsible for enhancement of
the vocalic dorsal gesture.
We now turn to a discussion of the interaction of the preceding consonant’s
place of articulation and vocalisation. Table 6 above showed that dorsals are
once again strong promotors. Preceding coronals are weak promotors in
contrast with following coronals, which, as we saw, inhibit vocalisation.
Preceding labials, surprisingly perhaps, inhibit vocalisation.
The postconsonantal laterals are syllabic: they occupy a nucleus position.
According to our notion of syllable harmony, this is a strong vocalising
environment. When the dorsal gesture of the l occurs in a strongly dorsal
environment (i.e. both sharing dorsal features with adjacent segments and in a
nuclear or coda position) this promotes its maximisation, hence vocalisation.
Where a dorsal consonant precedes, the dorsal gesture is maintained
through both segments. The coronal gesture of the l may or may not be
articulated by a speaker because the l is in a nuclear environment where the
consonantal gesture is dispreferred and the dorsal gesture is maximised.
The gestural diagram above can be translated into the tableau below:
(10) /pkl/
To explain the other two places of articulation we must think more about their
articulatory properties. Compare now the abstract relationship of the gestures
of l to each other when preceded by a coronal consonant.
There are two possible pronunciations of the coronal: [t] or a flap [ɾ] and the
choice between these two has an effect on the frequency of vocalisation.
The dorsal gesture of the l can occur simultaneous with the production of
the coronal consonant which precedes it. The vocalisation effect has to do
with the timing of the release of the coronal gesture. The speaker has the
option of not releasing the central coronal closure at all, or releasing it laterally
straight into the l. Maintaining the central coronal gesture will clearly inhibit
vocalisation which requires loss of the coronal closure. If the coronal closure
is released the l can vocalise. After flapped t : [bɒɾ], vocalisation is less likely
and it seems to have to do with the fact that the flap is in these circumstances
not released centrally. Instead it is laterally released and the coronal closure is
maintained. If the central closure of the t is released vocalisation may occur:
[bɒtυ]. This explains the figures for preceding coronals. The preceding
coronal can inhibit if it is laterally released but when it is centrally released it
has a promotory effect.
The picture is entirely different with labial consonants because the labial
gestures are independent of the coronal and dorsal gestures. The gestures for
the l may overlap or even be simultaneous with the labial stop and not be
perceived. The two articulations are independent of each other.
We would have expected the frequency for preceding and following
labials to be similar. Labials should neither promote nor inhibit vocalisation
because they do not share any gestures. The inhibitory effect of preceding
labials must be due to the fact that the articulation of both the labial consonant
and the following l can be uttered simultaneously. This means a speaker has
plenty of time to articulate both of the gestures which constitute the dark l.
Nevertheless, vocalisation, while less likely, is still possible because being in
a nuclear position has a promoting effect. A following labial has a different
effect because it in fact shortens the articulation of the preceding l and thus
leaves the speaker less time to articulate both gestures. Furthermore, we must
also take into account the fact that the cluster ls also have the promotory
effects of the preceding vowels while the preceding labial+ l environment do
not. The only other factor influencing these forms is the following environment. Both of these additional factors are discussed below.
3.2 Following environment
Following environment proved to have one of the strongest effects on the
frequency of vocalisation in our study. These environments include those
which may cause changes in the affiliation of the l in the syllable structure and
hence provide interesting support for our position.
A following consonant had the strongest effect in promoting vocalisation
for all classes of Coda l. Following consonant does not change the affiliation
of the l except to push it closer to a preceding nucleus which in turn has the
effect of shortening its articulation. Pause was almost as effective a promotor
as a following consonant. This is a position which does not shift the l out of its
coda position and it is not clear what phonetic effect of pause would trigger
vocalisation. In contrast, a following vowel inhibits vocalisation because it
causes a crucial change in syllabic affiliation. Where the l is required as a
syllable onset the timing of the “consonantal” coronal gesture must correspondingly be maximised and the segment will surface as [l]. Where it is
unequivocally in the coda of a syllable it may vocalise or not.
Table 7. Following environment — coda
coda /l/
Following consonants promote vocalisation because of the coda effect
discussed above. This accounts for the predilection of ls to vocalise both when
followed by consonants and when followed by pause. The requirement that l
be an onset when followed by a vowel inhibits vocalisation since in this
environment the more consonantal coronal gesture is enhanced.
The results for syllabic l were at first glance very puzzling. Pause was a
strong promotor, as above for coda l. In prepausal position a syllabic l is
definitely a nucleus. However, while following vowel was still an inhibiting
factor it was much less so than with the codas, and the role of following
consonant is problematic because it is inhibiting where we expect it to promote.
Table 8. Following environment — syllabic
syllabic /l/
We consider the problem of the following consonant first. A careful investigation of our data led us to the discovery that it was skewed in favour of other
inhibiting factors. If we correlate the results given in Table 6 showing the
effects of consonant place with these data involving consonants which follow
in another word, the explanation emerges. A quick check of the place of
articulation of these following consonants led to the discovery that there were
no following dorsal consonants in the corpus at all; there were five instances
of following coronal consonants and three of following labials. In contrast the
coda group in the data is followed by an even spread of following places of
articulation: four following dorsals, six coronals and five labials. It is probably
this gap in the syllabic l data which is the reason for the peculiar differences
between these ls and coda l. We know that dorsals are strong promotors while
coronals are strong inhibitors. Having no promotors and many inhibitors will
result in a false outcome.
The discrepancy observed in prevocalic position between the two sets of
data has a straightforward explanation. When followed by a vowel-initial
word a final l becomes ambisyllabic in order to provide an onset for the
following vowel. A syllabic l functions as the nucleus of its own syllable as
well as the onset of the following one. The syllable harmony injunction (to be
as vowel-like as possible in a nucleus and as consonant-like as possible in an
onset) remains in force and results in a conflict for these cases. In the
nonsyllabic cases, which form the majority of coda ls, the l does not suffer this
conflict because the syllable they are shared with has a vowel nucleus.
Therefore the inhibition effect we saw so strongly for coda l is less strong for
these syllabic ls because of the conflict between the syllable harmony requirements which syllabic ls incur when they are ambisyllabic: they need to be both
vocalic (as nuclei) and consonantal (as onsets).
3.4. Preceding vowel place
Finally we consider the effects of the adjacent vowels on l vocalisation. We
found that height, backness and length of the preceding vowel played a
significant role in vocalisation of ls. We have seen the effects of vowel length
above in Table 3, so we shall turn immediately to the other vowel features.9
On the frontness/backness parameter we found that frontness10 is a strong
inhibitor of vocalisation while central and back vowels promote. According to
Recasens (1996) the preceding vowel context is an ambiguous one. Arguing
on perceptual grounds he points out that the environment in which the vocalised form should be most perceptible is after front vowels since the cue,
which is a longer and lower second formant, would be more salient here. This
is however not the case either for the Romance languages he discusses in
which it is the back vowels which favour vocalisation not the front vowels, or
for AusE. Our results correlate with those of Recasens and support our
proposal that the coarticulatory effects combine with the syllabic to promote
or inhibit maximisation of a gesture.
Table 9. Preceding vowel backness
Vowel backness
clustered /l/
coda /l/
In the clusters the preceding central vowels had a very powerful effect followed
by the back vowels. The formant transitions for the lower and the backer vowels
closely resemble those of the dark l. In the coda situation the same pattern of
distribution occurs.11 For both sets front vowels inhibit vocalisation.
One of the strongest factors promoting vocalisation is preceding vowel
height. Mid vowels weakly disfavour, while low vowels inhibit vocalisation.
Table 10. Preceding vowel height
vowel height
clustered /l/
coda /l/
no low Vs
Here our results show a marked discrepancy from those of Recasens. He
found that in Romance languages the low vowel [a] is the most frequent
vocalising environment. In fact there are no l clusters with this vowel remaining in English. They have vocalised historically (e.g. palm, calm) which
suggests that this is/was also a strong vocalising environment in English. Our
low vowel environment in this data set is the front vowel [æ] and we did not
have any example of a clustered [æ] (as in talc) and all the coda [æ]
sequences were names (Sal, Mal, Malcolm).
In our data high vowels strongly favoured vocalisation. These results may
be skewed by the vowel length factor; the figures above do not separate out the
long high vowels from the short high vowels. In our previous study we found
that the combination of length and height led to vocalisation. Borowsky and
Horvath (1997) argue that l is syllabic after a long vowel. If syllabic it is under
pressure to vocalise because it is a nucleus. It is not possible at this time to
separate out the length and height factors for our new data so this must remain
speculation for the time being. Suffice it to say that if our previous observations were correct the same will be the case here, which will explain these
We also found that certain words were overwhelmingly vocalised, for
example milk and real (in the context real cool); again we can surmise that
forms like these skew the data. For this reason we put forward our vowel
figures even though it is clear that further work is required to determine the
facts of the situation.
In conclusion, we have shown that vocalisation of l in AusE can be
described as a response to two things: adjacent backness of both consonants
and vowels, in combination with syllable position. Non-onset positions seem
to require more vocalic segments and this is promoted when the segment is in
a back environment and inhibited when minimisation of the coronal gesture of
the l does not take place because of assimilation. We proposed a unified
account of these two things by making use of the bi-gestural representation of
l and the Sproat and Fujimura hypothesis that some gestures are more intrinsically consonantal and others are more intrinsically vocalic, and that these are
enhanced in the appropriate syllable positions.
This study was funded by a small ARC grant awarded to to the author and Barbara
Horvath on the Institutional Grants Scheme at Sydney University. Special thanks must go
to Barbara Horvath with whom I work on L vocalisation and who did all the varbrul
analysis for me. I would also like to thank our research assistants, Vicki Walker, Megan
Jones and Q. Chi Luu, for their assistance in data collection and coding.
Browman and Goldstein (1990) define a gesture as “a coordinated articulatory movement” as well as an “abstract, discrete dynamic linguistic unit”. Our understanding of the
term includes both. We encode the abstract in our phonological account and the phonetic
in our attempt at an articulatory description.
The results were coded using “The Language Coder” by a research assistant who was
conservative about judging a word to contain a vocalised /l/. Vocalisation is often
extremely difficult to hear in recordings and therefore the rate of vocalisation is probably
See Section 2.1 below. The inhibiting factor was the accidental exclusion of noncoronal
clusters in the Adelaide study.
Clustered cases are subject to a number of other factors promoting vocalisation which we
discuss below. The phonetic effects of clustering include shortening of the duration of the
consonants in the cluster, which has an effect on gestural timing. This would mean in this
environment that the consonantal (i.e. coronal gesture) would be even shorter than usual.
Clustering also pushes the l closer to the nucleus which we assume has an effect on
maximisation of the vocalic gesture.
Recasens (1996) goes on to show that vocalisation takes place in Romance languages
before apicals but not before velars and labials and proposes a dissimilatory mechanism
to explain it.
Note that the gesture of the tongue tip need not in fact be complete closure; presumably the
tongue tip can make the apical gesture without actually effecting a closure and the sound
will remain a lateral. It is an interesting question for future research to find out when the
coronal gesture is no longer large or long enough to effect a salient lateral sound.
We had only two words of this type in our data and they showed a large discrepancy in the
amount of vocalisation observed: film vocalised frequently: 60%, as compared to help:
17%. We wondered whether the sonority of the following consonant played a role but we
can do no more than speculate at this stage. Sonority did not have any obvious effects
with the dorsal or coronal clusters in our data.
The vowel classification system we used is given below:
feel (long)
fill (short)
file (diphthong)
sail (long)
dull, girl(stress) Arnold,
sell (short)
horrible, (unstress)
soil (diphthong)
Mal (length not applicable)
pool (long)
pull (short)
owl (diphthong)
old (long)
We note here the single exception to this was the word milk which vocalised 60% of the
time. We must assume that the promotion effect of the following dorsal consonant
outweighed the inhibitory effect of the preceding front vowel.
Recasens offers an interesting explanation of these facts. He makes the same assumption
as we do about vocalisation, arguing that the timing of the dorsal gesture occurs earlier
than the apical one. This results in a longer and very salient F2 transition which he
suggests may be considered to be a separate segment by a listener. After a back vowel the
listener assumes that the vowel formant transitions to a very dark l constitute a separate
segment — thus cool [ku]> [ku] and possibly also: [ku]. (Presumably after front
vowels the acoustic salience of the vowel transitions makes them more clearly vowel
transitions. Thus when there are clear and long transitions one perceives a dark l.
Certainly our intuitions suggest this to be the case. We seem to “hear” ls here with more
certainty than after back vowels.)
If a consonant follows the l there does not appear to be a stage where a glide is
perceived as well as an l. Here Recasens (1996: 85) suggests there is some articulatory
overlap of the gestures of the “segment corresponding to the vowel transitions and the l
segment due to the fact that /VwlC/ is unstable both on articulatory and perceptual
grounds”. See also Borowsky and Horvath (1997), who show that it is unstable also on
phonological grounds, specifically it is a bad syllable.
Blevins, Juliette. 1994. “A place for lateral in the feature geometry”. Journal of Linguistics
30: 301–48
Borowsky, Toni and Barbara Horvath. 1997. “L-Vocalization in Australian English”. In F.
Hinskens, R. van Hout & W. L. Wetzels, eds. Variation, Change and Phonological
Theory. Amsterdam: Benjamins, 101–123.
Browman, Catherine and Louis Goldstein. 1990. “Tiers in articulatory phonology, with
some implications for casual speech”. In John Kingston and Mary E. Beckman, eds.
Between the Grammar and Physics of Speech. Cambridge: Cambridge University Press,
Horvath, Barbara. 1985. Variation in Australian English: The Dialects of Sydney. Cambridge: Cambridge University Press.
Jespersen, Otto. 1961 A Modern English Grammar on Historical Principles. Part 1:
Sounds and Spellings. London: Allen and Unwin.
Jones, Charles. 1989. A History of English Phonology. London: Longman.
Ladefoged, Peter and Ian Maddieson. 1996. The Sounds of the World’s Languages. Oxford:
McCarthy John and Alan Prince. 1993. Prosodic Morphology I: Constraint Interaction and
Satisfaction. RuCCS Technical Report 3. Piscataway, NJ: Rutgers Center for Cognitive
Science, Rutgers University.
McMahon, April, Paul Foulkes and Laura Tollfree. 1994. “Gestural representation and
lexical phonology”. Phonology 11: 277–316.
Prince, Alan and Paul Smolensky. 1993. Optimality Theory: Constraint Interaction in
Generative Grammar. RuCCS Technical Report 2. Piscataway, NJ: Rutgers Center for
Cognitive Science, Rutgers University, and Boulder, CO: Department of Computer
Science, University of Colorado.
Recasens, Daniel 1996. “An articulatory-perceptual account of vocalization and elision of
dark /l/ in the Romance languages”. Language and Speech 39: 63–89.
Sproat, Richard and Osamu Fujimura. 1993. “Allophonic variation in English /l/ and its
implications for phonetic implementation”. Journal of Phonetics 21: 291–311.
Tollfree, Laura. 1996. “Modelling phonological variation and change: evidence from
English consonants”. Ph.D., University of Cambridge.
Walsh Dickey, Laura 1997. “The phonology of liquids”. Ph.D., University of Massachusetts, Amherst.
Hypocoristics of place-names
in Australian English
Jane Simpson
University of Sydney
They were all getting off at Goondiwindi, another dragon for my tongue. As
with Dirranbandi, some chose the first syllable, others the second; some
rounded the Goondi and others flattened it to Gundy. I found again that the
locals called it Gundy, which comes easiest to the tongue and is unmistakeable to the ear. There and then I learned that Australia is a great place for using
the diminutive suffix. Thus, Brissie was Brisbane, Gundy was Goondiwindi;
Dirranbandi became Dirran or Douran (blackfellow), and even Australia was
Aussie, or Ossie. (Smith 1976: 61)
English speakers have many ways of making hypocoristic1 forms of words or
names, which share part of the same form, have the same denotation (Gundy:
Goondiwindi), but have different connotations. Australian English (henceforth, “AusE”) speakers, as the quotation from Smith shows, have many
hypocoristics of place-names. Questions about the meaning, form and use of
these include:
1. Is the use of place-name hypocoristic forms a marker of AusE?
2. What are the meanings associated with using a place-name hypocoristic?
3. Is there the one hypocoristic form albeit with allomorphs?
4. What are the means for creating hypocoristic forms of a place-name?
5. Are there systematic relations between the phonological and morphological shapes of the full form of the name and the hypocoristic?
6. Is there variation as to use of hypocoristic and of form used?
The use of hypocoristics for common nouns and proper nouns is a feature of
the speech of many Australians. It is often mentioned in popular works on
AusE (Keesing 1982).2 Linguists have paid attention to the meanings, uses
and phonological properties of these forms (Dabke 1976; Dermody 1980;
Mühlhäusler 1983; Wierzbicka 1984, 1986; Taylor 1992; McAndrew 1992;
Robinson 1998). We build on their work to consider one kind of hypocoristic,
those used for place-names.
The data for this paper consists of 346 alternant names used in Australia
for places which are mostly in Australia. The referents of these toponyms
include countries (Oz: Australia), states (Tazzie: Tasmania), towns (Adders:
Adelaide), suburbs (Sevvo: Seven Hills), sports grounds (The WACA), schools
(Stannies: St Stanislaus School), buildings (Wenty: Wentworth Building),
pubs (The Wello: The Wellington). We exclude names for denizens of places
(Troony: Duntroon cadet).
The data3 forms part of a study of hypocoristic words in AusE.4 Our
original interest was the phonological and morphological relation between the
hypocoristic and the base word. We collected forms from written sources,
from listening to radios, conversations and asking people. We did not collect
systematically with a view to which hypocoristics are used where, by whom,
and for what conversational purpose, or to distinguishing between those of
wide currency and those used by small groups. But even so, this unsystematically collected data raises interesting questions for regional and social studies
of AusE.
Before discussing the data, a couple of problems must be mentioned. The
ambiguity of English spelling results in many variant spellings. Not all spelling variants cause ambiguity (whether of base placky or plakky: plastic, or of
endings, such as ie or y or ey, o or oh, a or er). But in some cases the ambiguity
makes it difficult to use examples from written sources. For example ‘ng’
stands for both [ŋ]5 and [ŋ]. Without hearing Singo, an alternant of the place
Singleton, we would not know that the alternant differs from the original in
lacking the [g]. Likewise the spelling rego can stand for [rεdoυ] from
registration, and [rεoυ] from regulation.
Compounding the problem is the fact that when writing the hypocoristic,
people tend to maximise the resemblance of the alternant to the original; thus
in Aussie: Australian the retention of the ‘s’ obscures the correspondence
between [s] in the original and [z] in the hypocoristic form. Vowel changes
also occur, Mazza: Mary; although consonants may be doubled to show this:
Ammo: Amy. In both the hypocoristic has a short vowel [æ]. Finally, sometimes spellings are changed — some people write trany for transvestite, transgender, or trans-sexual, to contrast it with tranny: transistor (Adam Schembri
We list the ways of forming hypocoristics of toponyms in Table 1. With
one peculiarity (‘the place-names’, names formed using the definite article:
The Loo: Woolloomooloo), hypocoristic names for places are formed in the
same way as hypocoristics of common nouns (forkie: forklift operator) and
personal names (Hendo: Henderson). Personal names and place-names are
subject to punning, and idiosyncratic creation.6
Table 1. Ways of forming hypocoristic names for toponyms
Target form
Forms without The
one syllable plus /i/
one syllable plus /o/
one syllable (usually
one syllable plus /a/
One syllable plus /as/
Two syllables9
One syllable plus /i-s/
One syllable (usually
first) plus /s/ ([s] or [z])
Two syllables plus /s/
([z] or [s])
Two syllables plus /o/
no. of
no. of
examples7 toponyms
example of
example of full
Crows Nest
Baulkham Hills
Port Melbourne
Macquarie Uni.
Port Lincoln
Ursula College
St Ives
Target form
no. of
no. of
Forms with The10
topographic descriptor
one syllable11
one syllable plus /i/
one syllable plus /o/
one syllable plus /a/
one syllable plus /i-s/
two syllables12
OTHER: e.g.
consonant distortions,
acronyms, other
example of
example of full
The Mount
The Hills
The Gong
The Steyne
The Curry
The Impy
The Wello
The Alice
Mount Gambier
Surry Hills
North Steyne
The Imperial Hotel
The Duke of
Wellington Pub
Mount Isa
The Exhibition
(event, building)
The Tradesmen’s
Union Club
Alice Springs
Sing Sing
Port Augutta
Kangaroo Island
Port Augusta
Civic Centre14
The Isa
The Ekka,
The Ekker
The Tradies
The data includes six alternant names for overseas place-names, but excludes
forms we have not found used in Australia. We also exclude work-place
specific alternant names unless they have become widely used, such as the
three-letter codes used for airports (LAX: Los Angeles), and abbreviations for
Victorian library branches (BLA: Blackburn, Simon Clegg p.c.).
In the first part of this paper we discuss some aspects of the meaning and
usage of the hypocoristics, and speculate on regional variation. In the second
part of the paper we discuss the ways of forming hypocoristics.
Use and meaning
Place-names vary as to whether they have widely accepted hypocoristics (as
shown by use on the national broadcaster, The Alice: Alice Springs and
Wagga: Wagga Wagga), or ephemeral hypocoristics (Lajas: Lajamanu), or
even as to whether they have hypocoristics at all. The type of place is relevant;
pubs often have hypocoristics. Some hypocoristics of place-names appear
primarily in derived place-names: thus several people gave Thebbie Town
Hall: Thebarton Town Hall but hesitated about Thebbie: Thebarton for the
Moreover, knowledge of hypocoristics, and use of hypocoristics, varies
widely,15 both regionally and within social groups. Not all Australians use
hypocoristics of place-names. We discuss geographical variation first, and
then social variation, emphasising that our data in this area is only preliminary.
Table 2. Distribution of hypocoristics according to State
Location of places
Total number of places
recorded in State
New South Wales (NSW)
(excl. Sydney)
Victoria (Vic)
Queensland (Qld)
Western Australia (WA)
South Australia (SA)
Australian Capital Territory
Northern Territory (NT)
Tasmania (Tas)17
General (Oz: Australia)
Number ending in /o/16
8 (one used by
3 (one used by
The difference in numbers reflects in part the difference in sources of information; we had good access to data from New South Wales, and almost no access
to Tasmanian data. However, we believe that for the states where we have
more data (NSW, WA, SA, Qld and Vic) it also reflects a difference in use.
The differences almost certainly reflects differences in use within social
networks, but there are two striking regional differences.
First is the difficulty of finding hypocoristics in South Australia. Seventeen South Australians (including a postman, a taxi-driver and three linguists),
from seven social networks, could only come up with 27 hypocoristics of
toponyms, whereas four linguists with associations with Western Australia
came up with 33 hypocoristics of place-names. A suggestive difference comes
from the official Air Traffic Services call-signs (ATSC 1998) for the towers:
Archer Tower at Archerfield Airport (Queensland), but Parafield Tower for
the tower at Parafield Airport (South Australia).
Second, the greater number of place names ending in /o/ in Sydney, the
ACT and Western Australia is also striking. We believe this is a change in
progress. The ending /o/, elsewhere recorded on personal names and common
nouns, is moving on to place-names, possibly via new places and compound
place-names, as the Victorian Shoppo: Shoppingtown, Seppo Street: Separation Street, Packo: Packington Street.
In terms of social variation, some networks of friends and relations use
hypocoristics for toponyms; others do not. Use of familiar names may be an ingroup activity; people sometimes commented on how odd outsiders sounded
using such names. One farming family in Parkes, New South Wales, in 1998,
could think of only seven hypocoristics for towns (such as Cooka:
Cookamidgera) and none for names of farms or properties,18 whereas they
could think of many dozen places. This contrasts with more than 40 hypocoristic place-names from five Sydneyside linguists who knew each other.
While we have occasionally recorded more than one form of hypocoristic
for a place (The Matta/Parra: Parramatta; the Been/Northie/ North: North
Narrabeen; Wonnie/Thaggie: Wonthaggi), this is uncommon. Even more
rarely have we found a single speaker admitting to using more than one form
for the same place-name.19 Variant forms of place-name hypocoristics are
likely to occur in different social networks.
To conclude, there is variation as to whether a hypocoristic is used for a
place, and variation in what form of hypocristic is used, according to region
and social network. Many places do not have accepted hypocoristics, and
hypocoristics of place-names vary considerably in how widely they are known
and used, and how ephemeral they are.
Wierzbicka (1986) has argued for differences in meaning based on speaker’s
attitude between hypocoristics of common nouns ending in /i/ (mozzie: mosquito), or ending in /o/ (journo: journalist). We might expect that, given that
place-name hypocoristics are formed in nearly the same ways, the differences
in meaning would carry over. However, the variation in whether or not a
hypocoristic is used, and if it is used, what form is used, makes it hard to
justify positing differences in meaning between the different ways of forming
hypocoristic toponyms. There is no evidence20 that the Sydneysider who says
the Gong: Wollongong thinks of Wollongong in the same way that they think
of the Loo: Woolloomooloo, but contrasting with how they think of Ballie:
Balmain, and both ways in turn contrasting with Werro: Werrington.
Most of the hypocoristics we have found act as alternative names, as
ways of expressing familiarity with a place, perhaps “I want you to know that
I know things about this place”. Or, “You live there, you’re proud of it, and
you want to show you’re familiar with it.” (Anna Choy p.c.). Exceptions are
joking names, Brizvegas: Brisbane (used by people in Lismore, Miriam Corris
p.c.) which are used not only to show familiarity but also to make a joking
point. Another exception is the /as/ ending. We first heard it on certain well
known tourist and shopping destinations: Honkers: Hong Kong, Bangers:
Bangkok.21 We understood it at the time as a kind of boastful familiarity “I
don’t think most people would know the kind of things I know about this
place.” It was later used ironically, as a kind of disparagement of the place:
“the kind of place that people boast about going to. I am not that kind of
person”. However, in the last few years we have heard it used on other places:
Sydders: Sydney; Tuggers: Tuggeranong; Lajas: Lajamanu, apparently without disparagement.
What is needed then, is the study of the use and meaning of hypocoristic
names within social networks. It is already clear that there are regional
differences, both in the way hypocoristics are formed and in how widespread
the use of hypocoristics is. But there will probably be significant differences
between people as to use and meaning. A tantalising hint of this comes from a
talk back radio caller22 who gave eight hypocoristics of beach names used by
Sydney surfers, all of which ended in /i/ (Palmie: Palm Beach, Whalie: Whale
Beach etc), thus going against the trend in Sydney to use /o/ for place-names.
How does the hypocoristic relate to the full form? Traditionally, the relation is
seen in terms of truncation of an original (Bree: Brewarrina) or as suffixation (Crowie: Crows Nest), or both (Monty: Montmorency). But these approaches have little to say about common pairs such as Pirie: Port Pirie, or
Archer: Archerfield, or Condo: Condobolin, in which the truncation results in
a form that ends phonologically with the same sound as the suffix.23 They also
have little to say about the couple of toponyms which conform to the shape
through loss of a consonant: Minno: Minto; Packo: Pascoe Vale.
Following the approach of McCarthy and Prince (1986) (see also Ito and
Mester 1997), we propose instead to see the relation as the aligning of the
original with one of a set of templates which may be one or two syllables long.
Thus Bree results from aligning Brewarrina with a one-syllable template;
Monty, Crowie and Pirie result from aligning Montmorency and Port Pirie
with a two-syllable template ending in /i/. Archer and Condo result from
aligning Archerfield and Condobolin with two-syllable templates ending in /a/
and /o/ respectively. That is, hypocoristics are formed by aligning the full form
to a template and making whatever changes are necessary (truncation, suffixation or change) for the form to fit the template.
4.1 Templates
Below, we go into detail about how these matches are made. We start with
one-syllable templates, then two-syllable and three-syllable templates, and
conclude with a discussion of the different endings, and the the place-names.
4.1.1 One syllable template
To match the one syllable template, normally the first syllable is taken. A few
examples have the last syllable: (Wang: Wallerawang, Scray: Footscray). A
few compound place-names with a monosyllabic first word have that word as
the hypocoristic: Glen: Glen Innes.
Bris| bane
The hypocoristic name has to conform to the word structure principles of
English, and thus the following are impossible: *Brizb: Brisbane (Briz);
*Kalg: Kalgoorlie (Kal); the *Dfern: Redfern (the Fern). Changes are sometimes needed to satisfy the stressed vowel requirement; thus initial schwa
vowels in the base may correspond to other vowels in the hypocoristic: Bree:
Brə’warrina; Tat: Tə’tura.
What constitutes the first or last syllable of the word for the purpose of the
template? The one-syllable hypocoristics are mostly formed by matching the
template to the first part of the original that can be construed as a syllable.
Thus the hypocoristic name does not have to maximise the coda: The Cry: The
Criterion Hotel, not The *Crite; Mount Vic: Mount Victoria, not Mount *Vict;
Oz: Australia, not *Aust (the voicing of the ‘s’ is discussed later). The few
hypocoristics formed from the last syllable of a word: the Gong: Wollongong,
the Wheel: Camooweal maximise the onset, subject to the possible word
constraint: *the Eel: Camooweal. But we have only one last-syllable example
in which more than one consonant could be taken without violating the word
structure: Scray: Footscray.
The match between the first syllable of the word and the one syllable
template can be modified in several ways:25 adding a consonant from another
part of the phrase: the Forest Lodge Hotel: the Flodge/the Slodge, adding a
final /s/ Scarbs: Scarborough Beach, or even both: Slens: St Leonards.
4.1.2 Two syllable template
The two syllable templates normally require the second syllable to be /i/, /o/,
/a/ or /as/, and require primary stress on the first syllable.26 A few exceptions
to the ending requirement exist, Warrack: Warracknabeal; Balak: Balaklava,
as well as names used in compound designators: the Crossing: Fitzroy Crossing; Alice/the Alice: Alice Springs; Hedland: Port Hedland; Tennant: Tennant
Creek, and five words of four or more syllables in which the second syllable
has a nasal coda: Dirran: Dirranbandi, Mullum: Mullumbimby; Tan.gam:
Tan.gambalanga. Murbah: Murwillumbah is an unusual form including the
first and last syllables.
The original can match the two syllable template in one of six ways:
by reduction to the first syllable (or part thereof) and adding an
Strad| broke Island
i → Straddy
by adding one of the endings to a monosyllabic word:
Palm| Beach
i → Palmie
by keeping the first two syllables of a word whose second syllable
matches one of the endings: Bilo: Biloela; Coota: Cootamundra;
Mordie: Mordialloc. This strategy also covers taking a whole word
of the right shape in a multi word name: Pirie: Port Pirie, as well as
altering the form of a two syllable word: Packo[pækoυ]: Pascoe
Vale; Bondy [bɔndi]: Bondi [bɔnda].
Mord i  alloc → Mordie
taking the last two syllables of a word: the Curry: Cloncurry; Gatha:
Leongatha, Thaggie: Wanthaggi, Quinty: Uranquinty. Most of the
examples are of the place-names. This strategy also covers taking a
whole word of the right shape in a multi word name: the Isa: Mount
Clon curr y → The Curry
by blends and acronyms: Wrackers: building that was once WRAAC
accommodation (Moore 1993); the Wacka: the WA Cricket Ground.
as → Wrackers
changing part of a word by introducing an ending inside a morpheme boundary (one jocular example: Muckyville, Mackyville:
[i] ville → Mackyville
The alternant name and the original do not always correspond exactly in sound.
Just as in the one syllable templates, stress shifts result in initial schwa vowels
in the original corresponding to other vowels in the template: Dennie:
De’niliquin. Other hypocoristics show a similar pattern, e.g. lino: li’noleum;
devvie: de’veloper; connie: con’ductor; mozzie: mo’squito; tom sauce: to’mato
sauce. There is a slight preference for the first vowel of the template to be short.
Thus occasionally a long vowel is changed to a short vowel: [æ] Scabbie: [a]
Scarborough (also Scarbs). Compare: [æ] cranno (=haircut): [e] cranium;
vag: vagrant; vagrancy, and chinga, ching (Chinese).
Sometimes a consonant is taken from elsewhere in the phrase; the examples often involve “Saint”: Snorbans: St Albans; Snadders: St Andrews.
Just as with the one-syllable template, the major questions arise from
determining the length of the match between the original and the template. In
the one-syllable examples we saw that the match was basically the first
complete syllable working from left to right. But the two syllable examples
show some variation determined by several parameters. The first parameter
concerns faithfulness to morpheme boundaries.27 This has several aspects.
(a) Modification of monosyllabic originals is rare: free: freebie, Kim: Kimbo;
Jim: Jimbo; Sambo: Samantha being exceptional non-toponym hypocoristics.
(b) If the original matches both syllables of the template, Cabra: Cabramatta,
then modifications of the medial consonants are rare, even though the br
cluster is not possible word-finally.
(c) If the word appears to be a compound, then the word is broken up at the
putative morpheme boundary: The Dicko: the Dickson pub, regardless of
whether meanings can be assigned to both parts: Arno: Arncliffe; Birchie:
Birchgrove; Clevo: Cleveland St. Some compounds include old morphemes
that only occur nowadays in place-names: Strathy: Strathmurton; Rockie:
Rockhampton. A rare exception: Newc: Newcastle.
If the original has one consonant separating the two syllables, and the first
vowel is short, then that consonant is taken: Katty: Katamatite. If the first
vowel is long, then the consonant is usually taken, Kybie: Kybybolite, but
exceptions exist: Freo: Fremantle, Bowie: Beaumaris, and, among the other
hypocoristics: cuie: cucumber.
Questions arise when the original has two consonants separating the first
two syllables. Faithfulness is a key principle. Other factors include:
(a) The Sonorant Ranking Hierarchy, with liquids highest and stops lowest.
The more sonorous the first consonant, the more likely the second consonant
is to stay.
(b) For the endings /i/, /a/ and /as/, the medial consonants must be capable of
appearing word-finally: Gillie/Gil: Gilgandra, not *Gilg or *Gilgie. Moreover, if the final consonant is more sonorous than its predecessor, then it is
dropped: Tazzie: Tasmania, (/zm/ is possible finally, chasm). For /o/ these
constraints are less strong: aggro: aggressive.
(c) The quality and place of articulation of the consonant immediately following the first vowel, and whether the cluster is homorganic.
The data to show this is given below, summarised in Table 3, and then
organised first in terms of how many medial consonants follow the first vowel,
and then in terms of manner of articulation, subdivided according to place of
articulation. We make comparisons to non-place-name hypocoristics, where
there are not enough place-name hypocoristics to allow patterns to emerge.
Table 3. Consonant clusters following the first vowel
C2 labial
C2 velar
C2 alveolar
C1 Liquid
C1 Nasal
C1 Fricative
C1 Stop
V C1 i
V C1 C2 V
V C1 i
V C1 C2 V
no examples
V C1 i rare
V C1 C2 V
V C1 V
V C1 C2 V
V C1 V
V C1 C2 V
V C1 V
V C1 V
V C1 C2 V rare
V C1 V
V C1 C2 V rare
C1 V
V C1 C2 V rare
V C2 V
(2 examples)
V C1 V
V C1 C2 V rare Two consonants
a. liquid initial clusters:
C1C2: Baulko: Baulkham Hills; Belco: Belconnen; Colbo: Colbinabin
C1: Gillie/Gil: Gilgandra; The Marlie: The Marlborough Hotel; Ballie:
These examples suggest a pattern we shall see more of: that is, the ending /o/
tends to preserve more of the original’s consonants, regardless of whether the
VCC sequence can occur word-finally. Examples from other hypocoristics
include: Salvo: Salvation Army member; galvo: galvanised iron.
The examples ending in /i/ all have VCC sequences that do not occur
word-finally: [alb], [ælm], [ilg]. Comparing with other hypocoristic words
shows a similar pattern: dallie: dalmation; Sallie(s):Salvation Army;28 since
[ælm] and [ælv] do not occur word-finally. Even if the cluster can occur wordfinally, with an /i/ ending the second consonant is often excluded: Thellie:
Thelma. Examples where the final consonant is kept and the sequence is
possible word-finally include: alkie/alko: alcoholic; alphie order: alphabetical order; bolshie: bolshevist; dissolvo: dissolving stitches; Wilko: Wilkinson.
The sequence /ly/, realised phonetically by some as [ ʎ ], by others as
[y], consistently is realised as /l/, obeying the word-final constraint: Willie:
nasal-initial clusters
In general, homorganic nasal-stop clusters keep both C1 and C2.
C1C2: nasal-stop: Cumbo: Cumberland College; The Impy: The Imperial
Hotel; Tumba: Tumbarumba; Wamby: Wamberal Beach; Cambo:
Compare: C1C2: ambo: ambulance/ambulance officer; imbo: imbecile, umpy: umpire; champagne: champers; Campo: Campese
(surname); compo: compensation
bommy: bombora (submerged reef)
For nasal-lateral clusters, we have no toponym examples, but retention of the
first consonant only is shown in gremmie: gremlin. (Note that [ml] is not a
word final sequence, unless the [l] is syllabic).
Here, several possibilities are found, with little to choose between them.
[ŋg] with the [g] deleted:
Singo: Singleton; Rangers: Rangoon
sanger: sandwich; Ingas/Ingy: Ingram
[ŋg] with the [g] kept:
Longgie: Longueville
sanggo: sandwich; linggo: language
[ng] with the [g] deleted:
Tonnie: Tongala
[ŋk] with the [k] deleted:
Bangers: Bangkok
[ŋk] with the [k] kept:
Honkers: Hongkong
Singapore allows the expected Singers with [ŋ], or Sinkers, with /k/ introduced, perhaps by analogy with its fellow shopping destination Honkers.
iii. palatal
There are no good examples of toponym hypocoristics with nasal-palatal
affricate clusters. Other hypocoristics are rare, and keep the nasal and affricate: engo: engineer; Rangie: Range Rover; conchie: conscientious objector,
with the possible exception of winnie: windcheater.
iv. alveolar
As with bilabials, homorganic nasal stop or fricative clusters usually keep
both the nasal and the stop. Some voiceless stops or fricatives are dropped.
C1C2 voiced: Condo/Condie: Condobolin; Bundy: Bundaberg; Gundy:
Goondiwindi/Gundagai; Yuenders: Yuendumu; Kenso: Kensington
C1C2 unvoiced: the Contie: Continental hotel in Broome;
BUT C1 (C2 unvoiced): Clonnie’s: Clontarf Beach Cafe; Lonnie:
Launceston; Wonnie: Wonthaggi29
fricative-initial clusters
With fricatives, the tendency is to drop one of the two consonants.30
C1: Brizzie: Brisbane; Chezza: Chester Hill; the Dizzo: district court (C2
C1C2: Austy: Austinmer, Husky: Huskisson
C2: Erko: Erskineville; Packo: Pascoe Vale
Clusters with voiced fricatives show variation in other hypocoristics:
• deletion of C2: wizzie: wisdom tooth, Bozza/Bozzie: Bosnich (surname)
• deletion of C1 (two examples), hubby: husband and the surname
Throbbers: (Margaret) Throsby (partly a pun)
• variation in C2 deletion: (two examples)
Prezbie/Prezbo/Prezzie/prezzo: Presbyterian;
lezbo/lezzo/lezzie/lez: lesbian
Clusters with voiceless fricatives tend to lose C2 and voice the fricative.
/ft/ (one example) arvo/arv: afternoon. But also aftie
/sp/ (two examples) hozzo: waves that “hospitalise you”; sarz: sarsaparilla [saspərlə]; But sus: suspect, as in “it’s a bit sus”.
/st/ (many examples):spazzie: spastic; fantazzo: fantastic; wizzie:
wisteria; cozzies: bathing-costume; plazzo: plastic nappie; Juz:
Justine; prozzie/prozzo: prostitute. But also prostie. Contrast also
voiceless /s/ in susso: sustenance = the dole, Chrissie/Christine;
Hessie: Hester (surname)
/sk/ clusters show variation: mozzie: mosquito; bikkie: biscuit; Bosko:
Boskovic; Rasko: Raskopoulos; Esky: Eskimo = food and drink
The voicing of the fricative probably relates to the AusE tendency to voice /s/
word- or syllable-finally, demonstrated in the [z] realisation of acronyms:
FAS (First Assistant Secretary) [fæz], BOS (Board of Studies) [bɔz], ASTEC
[æztεk], SCSI (Small Computer System Interface) [skzi].31 Notice that faithfulness to the original takes precedence when /i/ is added to a monosyllabic
morpheme: Westie: Western suburbs denizen,32 hostie: air hostess.
Interestingly, fricative-stop clusters in the original sometimes correspond
to /k/ in the hypocoristic: plakky: plastic; spackers: spastic (= drunk); lakky:
elastic. Note go berko: go berserk; spekky: special;33 Arko: Ken Arthurson.
There are two examples of fricatives or affricates followed by /y/: Ursies:
Ursula College, in which the original is realised phonetically by a voiceless
palatal fricative or a [sy] cluster, but the hypocoristic has an alveolar /s/ only,
and Cazho: Casuarina; in which the original and hypocoristic have a voiced
palatal fricative. Contrast Ramsay (1977)’s cazzo: war casualty.
d. stop-initial clusters
With these, the second consonant is almost always dropped, regardless of
whether the cluster can appear finally.
C1: (C1C2 impossible or rare word-finally) Rotto: Rottnest; Sydders:
Sydney; Matto: Matraville; The Aggy: The Agricultural Hotel;
Maggy: Magnetic Island
C1: (C1C2 possible word-finally, C2 more sonorous than C1) Watto:
Watson’s Crags; The Ekka/The Ekkie: The Exhibition (event, building)
An exception is Oodna: Oodnadatta, created by truncation to two syllables, and so more constrained by faithfulness. Other hypocoristics show the
pattern of dropping the second consonant:34
C1: (C1C2 impossible word-finally) gippo: Egyptian; Woodies:
Woodruffes (softdrink manufacturer); maggie: magpie/magneto;
joggers: geography; piccie: picture; lekkie: lecture; secko: section
C1: (C1C2 possible word-finally, C2 more sonorous than C1) secko:
sexual pervert; ackers: acne; preggo/preggers: pregnant; dippo:
diplomat; Sibbers: Sibley (surname); pubbo: public school kid;
techie: technician/technofreak
C1: (C1C2 possible word-finally, same sonority) Bappo: Baptist; seppo:
septic tank=Yank; ocky strap: octopus strap
Exceptions in which the second consonant is kept usually have the ending
/o/ and /r/ as the second consonant: aggro: aggressive; appro: approval. Three consonants
The few examples of originals with three or four consonants following the
first vowel all drop the final consonant, but vary as to whether the second
consonant is also dropped.
C1C2 (stop-initial) Maxie: Maxwellton35
C1: (stop-initial) Straddy: Stradbroke Island; Gladdy: Gladstone;
Chaddy: Chadstone; Gippy Highway: Gippsland Highway
Compare: C1 ecca/eckie: ecstacy tablet; lekkie: electricity, electrical
engineer; Yabbers: Yabsley (surname)
C1C2C3 Ruxto: Ruxton (surname)
C1C2 (nasal-initial) the Sando: the Sandringham hotel; Sandie: Sandringham; Wentie Falls: Wentworth Falls; Wenty: Wentworth/
Wentworthville; Pendo: Pendle Hill; Monty: Montmorency
C1 (nasal-initial) Wennie Falls: Wentworth Falls; Wenny: Wentworthville; Dunners: Duntroon; Brunnie: Brunswick
Compare: C1 Demo: Dempster; Simmo: Simpson; Thommo: Thompson; Super Connie: Super Constellation (plane); blunnies:
Blundstone boots
C1C2 tanty: tantrum; Anders/Andy: Andrew; Lindy: Lindwell
(surname); The Simpo: the Simpson Desert Land Claim, a spur
of the moment form we once heard
C1C2C3 anthro: anthropologist; intro: introduction
4.1.3. Three syllable template
We have only found one three syllable place-name hypocoristic: Maroochy
Tower: Maroochydore Control Tower. However, they do occasionally occur
as hypocoristics of other words, often of words with unstressed short light first
syllables: delinko: delinquent; dissolvo: dissolving stitches; emaco: immaculate; sophisto: sophisticate. But the first syllable is not always light: fantazzo:
fantastic; or unstressed: colourie: coloured glass marble; commonoh: baked
clay marble; royalies: member of Royals football club; bullocky: bullock team
driver; rabbit-0: rabbit-seller, South Sydney rugby team; rubbishy: garbage
collector (WA word: Mark Donohue p.c.).
4.2. Endings
The major endings in order of frequency are: /i/, /o/, /s/ or /as/ and /a/.
The most common hypocoristic ending on place-names is /i/. Such names are
found all over Australia, and in other dialects of English (Cincy: Cincinatti,
Philly: Philadelphia). They are so common that Dermody (1980) tentatively
suggested that /i/ was the way of forming hypocoristics of place-names. This
is perhaps a function of the fact that /i/ is an ending with many meanings:
babytalk (doggie); occupation nouns (fuelie: aeroplane refueller; stackie:
library stack officer; firie: fire-engine officer etc); adjectives: (nutty), including “denizen of place-name” (Bankie: Bankstown denizen; Newfie: Newfoundlander). This adjectival use perhaps accounts for the fact that some
hypocoristic place-names in /i/ are recorded only, or primarily, in attributive
uses: the Goodie Showgrounds: the Goodwood Showgrounds but not Goodie
for the suburb Goodwood (SA); the Tuggie Parkway: the Tuggeranong Parkway but not Tuggie for the suburb Tuggeranong (ACT).
Irish English may well be the source (Taylor, this volume) of the /o/ hypocoristic ending on common nouns and personal names,36 which shows the
same freedom to violate the word-final constraint: Hanro: Hanrahan. However apparently it is not found on Irish place-names (Maire NíChíosain p.c.).
Nor have we found examples of /o/ on hypocoristics of place-names in the US
or Canada and only one example in the UK, Wanno: Wandsworth Prison.37
The apparent lack of /o/ on toponyms in Irish English is perhaps not
surprising, since /o/ is geographically restricted in Australia too. By far the
majority of such place-names occur in Sydney. So, for example, the longestablished Paddo: Paddington in Sydney contrasts with Paddington in
Brisbane which never used to be referred to as Paddo (Kevin Keeffe p.c.). The
Brisbane suburb contains places with hypocoristics of Paddington used
attributively: the Pad pub, the Pad shops (Bill Palmer p.c.), but now the pub is
called the Paddo. Compare also the English Padder: Paddington Station
(Jespersen 1961). Indeed, Dermody, writing in 1980 in Melbourne, considered
the /o/ of Paddington in Sydney to be a unique example of a hypocoristic placename, although he recognised /o/ as forming other hypocoristics, and recorded
/o/ on 18 surnames (Dermody 1980). Finally, the /o/ ending appears to have
taken stronger hold in Sydney in recent years (Brian Taylor p.c.).
As mentioned earlier, the /o/ ending, in contrast with the the other
endings, appears to maximise the medial consonants matched. A neat example
of this occurs in the following passage,38 in which the /o/ form takes the full
[kst] cluster; the /i/ form takes only the [ks].
[…] Bruce Ruxton made his way onto the Strassman set […] every time I
mention the Ruxto boy in dispatches […] Look, Ruxy is as modern as a
parliamentary opening. [Tony Squires, Sydney Morning Herald 14/11/98
Spectrum p.24]
The /s/ ending is now quite common on personal names (Gabs: Gabrielle;
Shopes: Shopen; Cuts: Cutler; Deeks: Di Castello; Mares: Mary, Pabs:
Pablo) and shows the [s], [z] allomorphy of the English morph /s/. But, as
mentioned above, there are only three place-names with this ending, as
opposed to eleven with the /as/ ([əz]) ending. However, the fact that Scarbs:
Scarborough (Beach) was given by first year students in Perth in 1998,
whereas in the same year a person in their mid-30s gave Scabbie for the same
beach, may indicate a rise in its popularity.
When this ending appears on forms like Pashes: Pashley [surname] it cannot
be distinguished from the previous ending, as it fits the phonetic allomorphy
of the English plural. However it also appears on forms where [z] or [s]
allomorphs would be expected: preggers;pregnant; chockers: chock-full;
champers: champagne, and so on. We therefore must posit a separate template
ending in /as/, which covers forms such as Bangers: Bangkok, Dunners:
Duntroon, Hongkers: Hongkong, Rangers: Rangoon; Sydders: Sydney. Some
forms have two possible sources: the /as/ template (with ending [əz] attached
to a one syllable truncation), or the /s/ template (with the /s/ morph attaching to
a two syllable vowel-final truncation, and hence being realised as [z]): Lajas:
Lajamanu; Singers: Singapore; Adders: Adelaide; Tuggers: Tuggeranong;
Yuenders: Yuendumu. Thus Lajas is derived from either Laj+as, or Laja+s.
The ending /a/ (usually written ‘er’) is used agentively in English to form
deverbal nouns: bite: biter, but has also extended its use, thus in AusE
chockfull: chocker; guts: to come a gutser, to fall off, and the modern imma:
immature (Adam Flynn p.c., Sydney 1999). The postvocalic /r/ is retained in
Irish English, “Eric her mentler son” (Doyle: 1994: 206) from “mental”, and
the ending occurs on surnames: Doyler: Doyle; Fitzer: Fitzpatrick/Fitzgerald;
Ryaner: Ryan (Maire NíChíosain p.c.).
In AusE the post-vocalic /r/ is lost, and so the ending is realised as [ə].
Taylor (1992) traces this ending’s occurrence on personal names in Sydney
AusE in the early part of this century: from Jimma: Jimmy, Micka: Mick, to
occurring with insertion of /k/ as a way of making hypocoristics from names
with medial /r/: Ecker: Eric, Wocka: Warwick, Mocka: Maurice, to the present-
day use with insertion of /z/: Ezza: Erin; Shazza: Sharon; Wozza: Warwick/
Warren; Bazza: Barry (Robinson 1998). These in turn can be truncated: Shaz,
This ending is recorded on British English place-names in student slang:
the Bodder: The Bodleian Library, the Radder: the Radcliffe Camera,
(Jespersen 1965), and occurs on present-day Irish place-names: Croker:
Croke Park (the national Gaelic games stadium); Dalyer: Dalymount Park
(Dublin soccer stadium). However, in Australian English it is most common in
two syllable truncations: Cooka: Cookamidgera; Lapa: La Perouse; Meeka:
Meekatharra. This is quite common in words of four or more syllables from
indigenous languages with initial syllable stress. As added ending, /a/ occurs
only on a few structures: The Royza: the Royal Hotel; The Ekka: The Exhibition (event, building), Macker: Macquarie University (Pam Peters p.c.).
4.3. Place-names in the
Forming place-names with the probably derives from use of a topographic
descriptor to substitute for a compound name which includes that topographic
descriptor, thus the Port for Port Adelaide or the Mount: Mount Gambier/
Mount Morgan. Such names reflect the perceived uniqueness of that port or
mountain. The topographic descriptor can be the first word, but more commonly is the last word: the Hills: Surry Hills; the Creek: Julia Creek/Hall’s
Creek; the Bay: Byron Bay/Batemans Bay/Holdfast Bay; the Cross: Kings
Cross/Southern Cross. Such topographic descriptors can in turn be shortened:
the Prom: Wilsons Promontory.
This strategy is then extended to other place-names. The simplest extensions are by taking a part of the word homophonous with a topographic
descriptor: the Reach: Longreach. Since the most common compound structure is modifier+head=topographic descriptor, it is not surprising to find heads
taken which are not topographic descriptors — as in pub names: the Rose: the
Native Rose Hotel; the Star: the Evening Star Hotel, but also other places: the
Isa: Mount Isa. Sometimes the initial word is taken: the Alice: Alice Springs.
Another extension is to take the last part of the word, and treat it as a head:
the Loo: Woolloomoolloo; the Gong: Wollongong/Mittagong; the Wheel:
Camoweal; the Donga/Dong: Wodonga; the Cutta: Tarcutta. Such forms are
most common in Queensland and New South Wales. The focus on the last
syllables of the original instead of the first syllables is a noticeable difference
between the place-names and other hypocoristic place-names.
We now turn to the questions we posed at the start of this paper, which have to
do with meaning, form and use of hypocoristics of place-names.
On the question of form, we have argued that the morphological relation
of the full form to the hypocoristic name can be considered as the full form
having to fit a one- or two-syllable word template which has stress on the first
syllable. The template approach allows us to represent the similarity in end
result, which would be obscured by concentration on the process (truncation
or addition) achieving that result. We thus can consider forms like Mordy:
Mordialloc, Pirie: Port Pirie, and Crowie: Crows Nest, which have similar
morphological structures, as having similar meanings, regardless of the different processes that relate the original to the hypocoristic.
On the question of meaning, we have argued that, with the exception of
the /as/ ending, the fact that there is normally only one hypocoristic form for
any given toponym suggests that the different templates for hypocoristic
forms of toponyms do not as yet represent a shared system of different
meanings. That is, while the /o/ template and /i/ templates for hypocoristics of
common nouns may represent different meanings (speaker attitude, say), there
is no evidence that the same templates on toponyms represent different
On the question of use, we have found some regional variation, both as to
whether hypocoristics are used at all, and as to the distribution of particular
forms, notably the /o/ form. We suggest that study of toponym hypocoristic
use by different social networks will reveal interesting variation.
Alternative words and names are often discussed under names such as ‘shortenings’,
‘clippies’, ‘abbreviations’, ‘truncations’, which focus on one means for creating them, or
under the name ‘hypocoristics’ which labels the meanings, or under the name ‘diminutives’ which labels one kind of meaning and, implicitly, one aspect of the form. Since
there are several means for creating these alternative forms, we cannot use names like
‘clippies’. We adopt with some reluctance the label ‘hypocoristic’ which is more general,
if less transparent.
The process receives enraged attention from time to time in letters to the editors, as in a
letter to The Age, 9/1/99 from Sally Graham, Malvern. ‘Why do Australian journalists
persistently rely on such childish abbreviations as “wharfies, “greenies”, “bikes” and
“yachties”?’ (Kate Burridge found this example).
This paper is derived from work David Nash and I have been doing on hypocoristics in
AusE, originally inspired by Anna Wierzbicka’s work. We thank the following people for
providing AusE data: our families, Brett Baker, Pauline Bryant (and the listeners to
Angela Catterns’ programme on ABC National Local Radio Network 20/8/98, especially
L. Ziegeler), Kate Burridge (and the listeners to Terry Laidler’s programme on 3LO 4/11/
98, especially Simon Clegg), David Bradley, Peter Campbell, David Carrick, Rachel
Dallas, Alan Dench, Mark Donohue, Mark Durie, Ian Green, Arlene Harvey, Andrew
Ingram, Jason Johnston, Kevin Keeffe, Mary Laughren, Bill McGregor, Daphne Nash,
Bill Palmer, Bruce Rigsby, Adam Schembri, Brian Taylor, David Wilkins and Monty
Wilkinson. For American English data we thank Judy Kegl and Beth Levin. For Irish
English data we are grateful to Maire NíChíosain. None of these people are responsible
for our recording of their data or for the use we have made of it. For discussion of the
theoretical consequences in earlier work we thank Toni Borowsky, John McCarthy,
Heather Robinson, Donca Steriade, Anna Wierzbicka.
The data, with IPA transcription, is available from the author, or on the web page of the
Department of Linguistics, University of Sydney.
I use the transcription of The Macquarie Dictionary, Second Edition.
Thus Durry High: Turramurra High School, a pun on durry=cigarette. We have only
looked at those names which fit regular shape patterns, like Scrambo: Scrambletown:
Campbelltown; and so Hurtville: Hurstville, Suffering Christ: Southern Cross are not
discussed. Augments for personal names, most typically used in playful intimate contexts
have been found on place-names: Rosypops: Roselands shopping centre (Poynton 1984);
Addles: Adelaide, but seem to be ephemeral playful inventions.
This total includes hypocoristics of common nouns, personal names and place-names.
We use /a/ to refer to the ending usually spelled ‘-er’, which is pronounced as a mid
central vowel in AusE, or as a syllabic “r” in Irish or American English. Since we decided
late in data gathering to include forms such as Victor: Victor Harbor (on the grounds that
we were including the converse: the Bay: Holdfast Bay), this is not a true reflection of the
popularity of these forms.
This target type excludes two syllable truncations which end in a vowel homophonous
with one of the endings.
These forms are almost exclusively used for place-names, although they are occasionally
found in proper names used for reference: the Don: Donald Bradman.
This target type excludes topographic descriptors.
This target type excludes topographic descriptors and two syllable truncations which end
in a vowel homophonous with one of the endings.
We have included only a few acronyms, e.g.: FNQ: Far North Queensland, TI: Thursday
Island, and states’ names: WA: Western Australia and so on. Since we did not collect
these systematically, the number here is an underestimate.
This ending [ɔs] was reported as used by a group of working-class young people in
Canberra in 1998, mostly on personal names (Rachel Dallas p.c.). ‘Civic’ is the heart of
They are rarely recorded in dictionaries or gazetteers.
This includes both the place names and other place-names.
No native informant.
Big properties in north Australia sometimes do have short names: thus in the Northern
Territory: VRD: Victoria River Downs, Banka: Banka Banka Station, Rocky: Rockhampton Downs Station. Complex names with ‘Downs’ are usually shortened: Alroy:
Alroy Downs.
Dermody investigated the suffixes /i/, /o/ and /a/ on about 200 words, and notes
“Out of some sixty regular, and forty ‘casual’ (i.e. occasional) informants, I encountered less than 1% disagreement on which suffix was
‘correct’ for place-names, Christian names, and surnames.” (Dermody
1980: 19)
Given the variety of personal name hypocoristics available, this probably reflects a
general tendency for a person to have a single preferred form of nickname used by a given
person, and probably within a particular social network. This is a tendency — not an
absolute; different diminutive names may be used for the same person, both regularly and
playfully. A Charles may be Charlie, Chaz, Chazza, Chazzie, Chazzo, or in Sydney at
least Chilla.
This is not an argument against Wierzbicka’s analysis of /i/ and /o/ on common nouns;
hypocoristic formation could apply differently to common nouns and proper names.
Remmers: Remuera, an Auckland suburb, has similar connotations (Philippa Horton
Pauline Bryant’s spot on Angela Cattern’s ABC radio show 20/8/98.
Poynton (1984) makes a similar point, suggesting that the final [i] of Ashley and Shirley
helped their extension to names of women from their use as surnames, or names for men.
This is to be read as follows: ‘A one-syllable template σ is linked ∆ to the first part Bris|
of the word “Brisbane” to derive → a hypocoristic form “Briz”.’
We have not found place-name hypocoristics with the /r/~/z/ alternation (Robinson 1998)
found in personal names: Ezza/Ez: Erin. Compare the places: Gerro: Geraldton, Werro:
Werrington. We have found one /l/~/z/ alternation: The Royza: the Royal Hotel.
A counterexample to the stress rule: Utopes: Utopia (Utopes is probably ephemeral).
Faithfulness to morpheme boundaries holds in other hypocoristics: rare exceptions are:
gumbies: gumboots; strawbs: strawberries, winnie: windcheater. Monosyllabic prefixes
are often exceptions: repo: repossession; reno: renovation; preemie/premmie: premature baby.
Some consultants consider Salvo Australian and Sallies British.
Compare other hypocoristics: connie: conductor; Tinnie: Tindale (surname), but contie:
Continental aeroplane engine, convo: conversation. Tranny: transistor [nz] is a rare
example of dropping a voiced fricative.
If C2 is nasal it is predictably dropped, as more sonorous: compare the toponyms Mozzie
Park: Mosman Park; Norzie: Norseman [nɔzmən] with a non-toponym Chrissy: Christmas.
British English shows this to some extent too: witness Wiz for Wystan Auden, and Gazza:
Gascoigne [sportsman’s surname] (Sheila Watts, HistLing posting 30/10/1998).
This contrasts with German forms such as ‘westerner’ Wessie: West and ‘easterner’
Ossie: Ost, in which the /t/ is lost in the derived form.
Spectacular is another possible source.
Shelties: Shetland Island seamen/ponies/dogs (also found in British English) has unusual
Perceived morpheme boundaries are relevant in these; thus the /ks/ may be kept because
of the similarity with the name Max.
Common nouns include: alco: alcoholic (Doyle 1994: 169); journo: journalist (Maire
NíChíosain p.c.). Proper names include: Sherro: Sheridan (Doyle 1994: 99); Charlo:
Charles (Doyle 1994: 270); Henno: Hennessy (Doyle 1994: 2) Jayo: Jason (Maire
NíChíosain p.c.).
John Hoskison. Inside. One man’s experience of prison. John Murray ca 1998, reviewed
Peter J. M. Wayne. Times Literary Supplement. October 9 1998 p.30.
Taylor (1992) notes that journalists often employ short forms of people’s names, regardless of whether any friend of the person uses them.
Blank, Claudia. 1992. Language And Civilization: A Concerted Profusion Of Essays And
Studies In Honour Of Otto Hietsch. Frankfurt am Main: Peter Lang Publishers.
Dabke, Roswitha. 1976. Morphology of Australian English. (Ars Grammatica, Band 6).
Munich: Wilhelm Fink Verlag.
Dermody, Anthony C. 1980. “Word abbreviation and suffixing in Australian English”.
B.A. (Hons), La Trobe University.
Doyle, Roddy. 1992. Paddy Clarke, Ha-Ha-Ha. New York: Viking.
Heller, L.G. & J. Macris. 1968. “A typology of shortening devices”. American Speech
Ito, Junko, and Armin Mester. 1997. “Sympathy theory and German truncations”. To
appear in V. Miglio and B. Moren, eds. Proceedings of the Hopkins Optimality
Workshop/Maryland Mayfest 1997. Baltimore: University of Maryland Working Papers
in Linguistics 5.
Jespersen, Otto. 1961. A Modern English Grammar On Historical Principles. London:
George Allen and Unwin.
Keesing, Nancy. 1982. Lily On The Dustbin. Ringwood, Vic.: Penguin Books.
Kreidler, Charles W. 1979. “Creating new words by shortening”. Journal of English
Linguistics 13:24–36.
Lightner, Theodore. 1983. Introduction to English Derivational Morphology. (Lingvisticae
Investigationes: Supplementa. Studies in French and General Linguistics Volume 6)
Amsterdam/Philadelphia: John Benjamins.
McAndrew, Alex. 1992. “Hosties and garbos: a look behind diminutives and pejoratives in
Australian English”. In Blank, ed. 1992, 166–184.
McCarthy, John J. and Alan S. Prince. 1986. Prosodic morphology. Unpublished ms,
Department of Linguistics, Harvard University.
Mester, Armin. 1990. “Patterns of truncation”. Linguistic Inquiry 21:478–485.
Moore, Bruce. 1993. A Lexicon Of Cadet Language: Royal Military College, Duntroon In
The Period 1983 To 1985. Canberra: Australian National Dictionary Centre, Australian
National University.
Mühlhäusler, Peter. 1983. “Stinkiepoos, cuddles and related matters”. Australian Journal
of Linguistics 3:75–91.
Neufeldt, Victoria. 1995. “A civil but untrammeled tongue: spontaneous creativity in
language”. Dictionaries 16:19–31.
Poynton, Cate. 1984. “Names as vocatives: forms and functions”. Nottingham Linguistic
Circular 13:1–34.
Ramsay, Jim. 1977. Jim Ramsay’s Aussie Slang. Sydney: Allegheny News Service.
Robinson, Heather. 1998. “Nicknames in Australian English”. Paper presented at the
Annual Meeting of the Australian Linguistics Society, Brisbane.
Smith, G.A.W. 1976. Once a Green Jackaroo. Sydney: Seal Book (Rigby).
Sundén, K.F. 1910. “On the origin of the hypocoristic suffix -y (-ie, -ey) in English”. In
Festskrift tillegnad K.F. Johansson. Göteborg: 131–170.
Taylor, Brian A. 1992. “Otto 988 to Ocker 1988”. In Blank, ed. 1992, 505–536.
Weeda, Donald. 1992. “Word truncation in prosodic morphology”. Ph.D., University of
Texas at Austin.
Wierzbicka, Anna. 1984. “Diminutives and depreciatives: semantic representation for
derivational categories”. Quaderni di semantica 5(1), (June): 123–130.
Wierzbicka, Anna. 1986. Is language a mirror of culture? Evidence from Australian
English. Unpublished 42pp. ts, Australian National University.
Syntactic features and norms
in Australian English
Mark Newbrook
Monash University
Australian English (henceforth, “AusE”) is for the most part very “mainstream”: it does not in general differ markedly — at any linguistic level or at
any level of standardness — from the equivalent contemporary usage in other
major English-speaking countries (Trudgill 1986: 143). Given the relatively
recent European settlement and the likely levelling and mixing of usage at that
time (Trudgill 1986: 129–146), this is not surprising.
This applies even to the most obviously non-standard AusE. Almost all
the forms which do appear markedly divergent can be explained in terms of
non-nativeness or Aboriginal background. Most of the other non-standard
features identified as rural Tasmanian by Pawley (1995) or as typical of cities
and country towns by scholars such as Eisikovits (1989, etc) and Shnukal
(1989) are all familiar from around the world. There are very few syntactic
non-standardisms which are clearly more prevalent in Australia than elsewhere, or appear both in Australia and elsewhere but are deemed clearly nonstandard in Australia but clearly standard elsewhere (the reverse pattern is a
little more common). The main exceptions involve cases where Australia
follows standard British usage which is non-standard in the United States —
or perhaps, less commonly, vice versa (because of increasing American
influence; Sussex 1989; Taylor 1989). Even here, however, there is obviously
no distinctively Australian non-standard form.
This pattern is repeated for features of “disputed/debatable/divided” usage, such as those discussed by Collins (1989). While details of patterning and
national preferences for particular variants differ, every feature examined by
Collins is familiar (as well as marginal) usage in British English (henceforth
“BrE”) and/or other native speaker varieties, as Collins himself makes plain.
The pattern occurs again, even more obviously (and not surprisingly), in
respect of careful, educated, clearly standard AusE (assuming that there is
indeed a distinctively Australian standard variety of English in respect of
syntax), and especially for formal styles/registers: printed works cannot usually be identified as being written or edited by Australians from their grammar
(though lexis may betray them). The few relevant grammatical features —
none of which is categorically diagnostic — are discussed below.
Syntactic norms
The extreme similarity of AusE syntactic usage at this level and equivalent
usage elsewhere does, however, raise the issue of the separate existence of a
distinctively Australian standard variety of English (for syntax), and hence of
the existence of distinctively Australian syntactic norms. Until relatively
recently Australia took its linguistic norms very largely from England/the UK.
Obviously this has now altered for lexis; and for syntax, style guides and
manuals produced in Australia by both linguists and others reveal that there
are indeed a number of forms in respect of which the advice offered in such
texts differs from what is offered in equivalent works published elsewhere —
although by no means all the relevant features are actually discussed very
It is also possible to judge the status of usage from the features of
published AusE, which has typically been edited by at least one educated
AusE user other than the author. Forms deemed clearly non-standard would
presumably have been modified, especially in more formal genres or contexts.
Some features which are either unusual or non-standard elsewhere survive this
process and are presumably deemed fully acceptable in Australia. Not all of
these are mentioned in style guides. (Of course, style guide writers may not
feel obliged to identify usage which is deemed fully acceptable locally, even if
they realise that there is an issue; their main business is to identify usage felt to
be palpably non-standard, especially if dangerously liable to be used in formal
contexts, and thus explicit focus is usually reserved for disputed usage).
In any event, it is clear from discussion with style manual writers and
other educated AusE users that most syntactic features typical of formal styles
which are in fact characteristically Australian (at least statistically) are not
generally recognised as such by Australians. Surprise is often expressed at the
information that some such feature is rare or absent, or would be deemed nonstandard, in (especially) BrE. In consequence, the standard variety of AusE, in
respect of syntax, currently constitutes more of a de facto than a de iure set of
norms; it is explicitly codified apart from other (mainly British) norms only to
a limited degree, and is not generally recognised by its users as a separate body
of norms. The issue is not as serious for Australia as for some other countries,
where de facto English norms in Asia differ much more markedly from the
supposed de iure norms, but it is still potentially important for Australians as
they seek to forge a new identity less closely associated with the British
heritage. However, standard AusE (despite now sharing some AmE features)
is still much less syntactically distinctive from BrE than AmE and BrE are
from each other. In Trudgill and Hannah (1994), the set of AusE versus BrE
grammatical differences occupies two pages; the equivalent set for AmE
versus BrE fills 27.
Given all this, the syntactic distinctiveness of standard AusE is hardly
obvious. Furthermore, some characteristically AusE features are largely confined to certain genres (including less “serious” genres) and may be infrequent. In addition, many forms which are characteristically Australian are
used in Australia alongside equivalent forms with wider distribution. These
“Australian” forms occur only in a certain percentage of the instances where
they could have occurred. This reduces their salience, both for Australians
themselves and for overseas commentators. An example is the case of concord
with sports team names, discussed below. In Australia both singular and plural
concord are normal, whereas in England only the plural is found. The presence
in AusE of the plural variant reduces the frequency and salience of the more
distinctively local form.
Putative syntactic features of standard AusE
Collins (1979, 1988, 1989) is perhaps the most perceptive of the few scholars
to have investigated these matters. I myself have listed possible syntactic
features of a de iure Australian standard variety of English which appear to be
(at least statistically) characteristic of Australian usage or at least shared only
with a few other varieties (Newbrook 1992, 1993). The list of features which
follows is drawn from these papers, and the account of each feature is further
informed by the results of subsequent investigation (by myself and others) and
by discussion with various colleagues, notably Michael Clyne, Peter Collins,
Jane Curtain, Edina Eisikovits, Geoffrey Pullum, Brian Taylor and Joanne
Winter. I include here only the better attested and more interesting phenomena. The list is divided according to the apparent status of each variable and of
its key variants, and the geographical (national) distribution of the variants.
In addition, I refer below to some leading recent style manuals: Hudson
(1993), Renton (1994) and Peters (1995). Of these, Renton (1994) is the most
traditional and prescriptivist, though the advice given is usually helpful;
Hudson is more discursive and more tolerant of forms which he himself would
avoid. Peters (1995) is clearly the most descriptivist and linguistically informed of the three. Both Peters and I have conducted surveys of the data in
the Australian Corpus of English (“ACE”) compiled at Macquarie University,
and some results are reported below.
3.1 Disputed usage (specific variants)
3.1.1 usedn’t to
There is some (rather marginal) evidence that usedn’t to, the preferred negative of used to, as in (1), has held its own more successfully in Australia
against didn’t use(d) to, as in (2, 3), than in other English speaking countries
(Newbrook 1992: 4–5, 1993: 52; Trudgill and Hannah 1994: 19). Collins
(1979, 1989: 144), however, found a preference for the latter form (albeit by a
smaller margin than would probably apply elsewhere). In ACE all three forms
occur, albeit infrequently (all feature informal -n’t). Items (2) and (3) are not
readily distinguishable in speech; (2) is a more plausible formulation of the
usual phonetics, but (3) construes more readily.
She usedn’t to do it.
She didn’t used to do it.
She didn’t use to do it.
The more formal variant used not to (4) complicates the issue further.
She used not to do it.
There is little consensus among style manuals. Renton (1994: 143) endorses
only the formal variant with not (but interestingly is especially critical of
didn’t use to). Peters (1995: 783) paraphrases Collins and adds remarks on
the preference for contraction, especially with did. Hudson (1993: 422) focuses instead on the equivalent interrogatives, suggesting that Australians
again favour the treatment of used to as an auxiliary (5) rather than the use of
do as in (6) or (7). However, while (5) is certainly found, it does not in fact
appear to be even as common as forms with usedn’t to; and Peters makes this
Used she to do it?
Did she used to do it?
Did she use to do it?
Hudson ascribes the form(s) with did to AmE specifically and is critical of this
usage (if more tolerant than Renton).
3.1.2 Different from/than/to and hypercorrect from
It is widely held (Newbrook 1992: 3, 1993: 52) that Australian usage favours
to as a preposition following the lexical comparative adjective different (and
its derivatives) to introduce the object of comparison, as opposed to conservative/British from and American-influenced than. See (8–10).
(8) Cricket is different to baseball.
(9) Cricket different from baseball.
(10) Cricket is different than baseball.
Trudgill and Hannah (1994: 76–77) discuss the (statistical) British-American
divergence involving from and than. Recently than has become common in
BrE (Crystal 1984). Trudgill and Hannah also note that to does occur in the
UK; and of course both from and than occur in Australia. The account given
by Follett (1966) suggests that to is (or was then) not widely accepted in the
USA. One structural factor favouring the adoption of than appears to be its
greater syntactic versatility: it freely functions either as a preposition or as a
subordinator introducing comparative clauses, whereas to and from cannot
have the latter function in standard English. Elsewhere (Newbrook 1989: 55–
63, Newbrook and Yio 1987: 104–132) I discuss this and other aspects of the
variable patterning of different (including possible correlation between use of
than and intensification with much).
Hudson (1993: 113–114) obviously prefers from but suggests tolerance
of both to and than; he accepts than as more elegant (if not, to him, natural)
where a subordinator is required. Renton (1994: 308) upholds from over to
and ignores than completely. In fact, it appears that — despite the local
flavour of to and the versatility of than — from has survived relatively well in
Australia, especially in more formal styles. Collins (1989: 147) found, not
surprisingly, that Australians accepted to more readily in informal contexts
than in formal. Peters (1995: 203–204) suggests that than might be expected
to occur most often as a subordinator in Australian usage (see above), but she
admits that the evidence is lacking, because no tokens at all of this particular
construction appear in ACE. Across ACE, from outnumbers to by around 6:1
(Peters 1995: 203). In a sampling of the newspapers section, the breakdown
across the relevant 107 tokens was as follows:
different to
different from
different than
These results suggest that than has made little ground and that to is still in a
minority in printed material, though by different margins in different genres.
In Australia and elsewhere, tokens occur — in both speech and writing —
of what may be hypercorrect from after morphological comparatives, perhaps
prompted by confusion/concern over usage with different.
(11) There were better instances from this.
3.1.3 Less and hypercorrect fewer
Collins (1989: 148) and Newbrook (1992: 3, 1993: 52) draw attention to the
use of less (12) rather than fewer (13) with plural countable nouns, on the
analogy of its use with non-count nouns as in (14).
(12) We have less students this year.
(13) We have fewer students this year.
(14) We have less money this year.
The use of less with plurals is apparently stigmatised to some degree in all
countries (Partridge 1947; Follett 1966) but 72% of Collins’ subjects and 53%
in an earlier Australian survey conducted with teachers by Watson (1978)
accepted it, which makes a contrast with 35% of the British subjects used by
Mittins (1969). The usage may be more acceptable in Australia; though of
course British attitudes too may be more tolerant today.
Renton (1994: 139) endorses the traditional pattern, accepting only fewer.
Hudson (1993: 149–150) also endorses fewer, but observes that on occasion
less may appear required by the semantics. He recommends restructuring in
such cases; but sometimes less may indeed be judged appropriate, as in (15),
where the sense is ‘fewer than 28 days’, not ‘one/two/three complete week(s)’.
(15) Less than four weeks later this actually happened.
The existence of such cases and a degree of awareness of the issue seem to be
behind occasional tokens (some Australian) of what appears to be hypercorrect fewer for standard less, as in (16) = (15) (Newbrook 1992: 23, 1993: 52):
(16) Fewer than four weeks later this actually happened.
Peters (1995: 276–277) suggests that less is becoming acceptable with plurals, particularly in informal styles. She points out that ACE has a number of
instances of less with plurals, and also a few tokens parallel with (15), though
apparently none resembling (16).
Specifically Australian usage (at least statistically)
The phenomena discussed here are specifically Australian usage (at least
statistically, and arguably standard (or potentially standard) in Australia.
3.2.1 Concord with sports team names
As noted, AusE (and New Zealand English), in both print and speech, permits
both singular and plural concord with the names of sports teams. English
English permits only plural concord here (though singular concord is common
with the same words understood as club names). See Newbrook (1992: 5–6,
1993: 54); and see below on AmE. The usage is associated with the Australian preference for singular concord with collective common nouns such as
team, government etc; here, however, English English does permit the singular, although it is apparently less common in England than in Australia.
Australian style guides generally do not refer to this specific phenomenon. They usually acknowledge that either type of concord is possible with
team etc. often suggesting that the choice involves the precise sense intended
(e.g. Renton 1994: 102; but see also pp. 124–125); some (e.g. Peters
1995: 31) refer to the prevalence in Australia of singular concord. Peters does
also draw attention to the possibility of plural concord with team names,
which is interesting but represents a somewhat odd focus given the dialectal
This phenomenon may be instantiated by (17) and (18); only the latter is
standard (or usual) in England, both are quite normal in Australia.
(17) North Melbourne is playing well.
(18) North Melbourne are playing well.
More distinctively Austral(as)ian is the use in this context of singular pronouns and nouns. Item (19) is quite possible in Australia; compare the English
English and alternative Australian version, (20).
(19) North Melbourne, which was the premier in 1996, is winning its
matches easily.
(20) North Melbourne, who were the premiers/champions in 1996, are
winning their matches easily.
AmE favours singular verbal concord (Trudgill and Hannah 1994: 72) but
sometimes avoids singular nominal concord even to the extent of switching
number between adjacent clauses to avoid a singular pronoun:
(21) Detroit is winning, aren’t they?
In contrast, AusE sometimes (although not very frequently) switches to singular number in such circumstances, or even within a clause, for instance where
a grammatically plural team nickname is used as subject:
(22) The Kangaroos [= North Melbourne] must improve its percentage.
In a small survey of Australian newspaper reports taken from ACE, the form
Collingwood (referring to cricket and Australian Rules football teams) appeared with singular concord (verbal and/or nominal) on 19 occasions and
with plural concord on three. There were no cases here of mixed concord. In a
comparable sample of British reporting the form Liverpool (referring to the
soccer team) was always grammatically plural (25 tokens).
3.2.2 Possessive pronouns with same
A number of tokens have been found of possessive pronouns (my etc.) with
same, in the sense ‘my (etc.) X, which is/was/will be the very same as…’
(23) Can I keep my same phone number?
Item (23) appeared in a television commercial (see Newbrook 1993: 52). The
usage appears unfamiliar to most non-Australians but uncontroversial to many
Australians. The style manuals do not seem to refer to the matter. There are
nine tokens in ACE, featuring her (four), your (two), his (two) and their (one);
they are predominantly of recent date.
3.2.3 Backshifting
There is some evidence (Newbrook 1992: 9–10, 1993: 53) that Australians
are less inclined than some other English speakers to apply the syntactic rule
traditionally called sequence of tenses but perhaps better named backshifting
(see Huddleston 1988: 72, 82). The simplest type of case involves preferences for unshifted verb forms in indirect statement clauses after past tense
verbs of saying (etc.), as in (24), over shifted forms as in (25).
(24) Kim said she has a bad cold.
(25) Kim said she had a bad cold.
In the (native) English speaking world generally, (25) is preferred wherever
possible; it is always used unless “Kim” still has her cold at the time of the
utterance, when (24) might — but still need not — be selected instead. In
contrast, Australian students often report that they prefer (24) over (25)
wherever the sense permits. Patterns in student writing offer some limited
support for such claims.
Backshifting is also a feature of the protases of remote conditional
constructions; and a second phenomenon involving avoidance of backshifting
is the use of the singly-shifted simple past (26) rather than the doubly-shifted
past perfect (27) in the protases of remote (counterfactual) past conditionals.
Though this point has not yet been examined in any detail, it appears that this
usage arises more commonly, in more formal styles and at higher educational
levels where the verb in the protasis is negated, as in (26) and (27).
(26) If that didn’t happen, I would have been surprised.
(27) If that hadn’t happened, I would have been surprised.
Clauses such as the protasis of (26) are, of course, familiar in remote (unlikely) present/future conditionals, again as a result of (single) backshifting as
is usual in all remote conditionals (28).
(28) If that didn’t happen, I would be surprised.
If the pattern manifested in (26) is adopted, pairs of sentences such as (26) and
(28) can, obviously, be distinguished only by their apodoses.
Usage such as (26) is certainly common in the UK and probably in other
English-speaking countries. However, it does not appear to be as common
outside Australia in printed usage or very careful speech; compare, for instance, its markedly non-standard use in Reading (Cheshire 1982: 50). It may
be more overtly stigmatised in other countries; in contrast, my Australian
students often express surprise at the information that any stigma might attach
to it.
In a sample across the newspapers section of ACE, 13 out of the 70
relevant protases contained simple past verbs (mostly negated) rather than past
perfect; the usage clearly does reach the print media.
A third manifestation of the avoidance of backshifting involves the use of
unshifted may for shifted might in a number of counterfactual constructions,
notably in perfective may have (29) rather than might have (30) used in the
apodoses of counterfactual conditionals (or in isolated main clauses with a
counterfactual protasis assumed). Interestingly, while the usage per se certainly occurs elsewhere (if less frequently), it appears much more strikingly
unfamiliar to many non-Australians when the verb is negated (compare the
previously discussed phenomenon).
(29) (If he’d missed that kick,) they may not have won.
(30) (If he’d missed that kick,) they might not have won.
In other varieties, it appears that verbs incorporating may (not) have can
generally refer only to past events the outcome of which is unknown (31).
(31) I haven’t heard the final scores: for all I know, they may not have
As in the previous case, the Australian usage creates an ambiguity in respect
of the relevant clauses where there is no supporting context: the final clause
which is shared by (29) and (31) has quite different senses in the two
In a survey of ACE, six instances were found of perfective may have
where in my judgement might have might have been expected, compared with
18 instances of might have in such circumstances. Not all of the six involved
(32) Lewis agreed he may have been able to play if he had delayed his
decision until Sunday.
There were also six parallel instances of may for expected might used in the
apodoses of non-perfective remote conditional constructions and preceding be
(lexical, passive or progressive); there were 21 corresponding cases of might.
(33) If we found out why these things happen, prevention may be
In fact, for many younger Australians the semantic contrast between may and
might (in their non-deontic senses) no longer involves backshifting or, indeed,
any issue associated with time or the time-related significance of tense (the
notion that might could be interpreted as a past tense form of may often
surprises my students). The distinction appears to be understood largely in
terms of the perceived probability of the action/event reported (especially after
verbs of saying); this is taken to be lower if might is used (Hudson 1993: 243–
245). This factor is, of course, relevant in all varieties, at least in some
constructions. Another factor here is degree of politeness, which is obviously
relevant in some constructions and may, as David Blair has pointed out, have
assumed a wider importance for some Australians. Collins (1988: 284) reports that might is, in any case, the least common of the modal verbs in
These issues are not discussed a great deal in most style manuals. However, Hudson (1993: 243–245) observes (with some dismay) the increased
use and acceptance of may have for might have in remote past apodoses,
though without suggesting how far this might be especially characteristic of
Australia. Renton (1994: 94–95), Hudson (1993: 371–372) and Peters
(1995: 684–685) discuss backshifting generally but do not raise these matters
3.2.4 Special use of the superlative with since
It has been noted (Newbrook 1992: 10–14, 1993: 51) that the construction
involving a superlative adjective + since (etc.), has, in addition to its more
widespread sense (which is also common in Australia), a second interpretation
which is very familiar to Australians and is perceived by most as quite
uncontroversial. However, it appears to be particular to Austral(as)ia; at any
rate, no instance has yet been found elsewhere, and most non-Australasian
judges seem to find the usage almost incomprehensible at first. Elsewhere, the
item following since in this construction is always described as having the
quality expressed by the superlative to a higher degree than the item actually
described by the superlative, as in the cricket example (34). The most usual
contexts are sports and weather reports, where quantitative records are common.
(34) His score of 200 was his highest since he made 250 in 1995.
The new mark of 200 is here the closest approach to the old mark of 250 that
has been achieved since the latter was reached, but not an improvement upon
it. The old mark may or may not have been the cricketer’s best ever score; for
instance, he may have made 300 in 1990, but that may no longer be deemed
relevant to his current scoring.
In Australian reporting, however, it is very common for this construction
to be used (in some instances) to report (a) the passing of an old mark, which
(b) typically is the standing record — as in (35), where the player’s old mark
of 175 is usually understood as being his best ever score prior to this new score
of 200.
(35) His score of 200 was his highest since he made 175 in 1995.
Elsewhere, this would be expressed with a completely different construction,
for instance as (36).
(36) His score of 200 was a personal best, beating his previous record of
175 set in 1995.
This phenomenon is, of course, very specific indeed, and appears not to have
attracted the attention of any other commentators.
3.2.5 Relative clause punctuation
There is some evidence (Newbrook 1992: 17–19, 1993: 53) that some Australians (at least in Victoria) have learned from secondary school teachers a
reversed form of the traditional rule determining the presence or absence of
commas before and after relative clauses. This evidence involves the reports
both of the students themselves and of mature university students who are
current or former teachers. This is likely to involve ad hoc advice from
individual teachers, since the relevant curriculum does not seem to address such
matters and any textbooks consulted would give the traditional rule (if any).
This traditional rule prescribes commas both before and after non-restrictive relative clauses (37) and the non-use of commas around restrictive relative clauses (38).
(37) Joanne and Jane, who had finished, left the hall.
(38) Any students who have finished may leave the hall.
There is, of course, considerable variation in this respect more generally.
Many writers of English (everywhere) sometimes use only one of the two
potential commas, which was at one time quite normal in written English but
which has more recently been deemed non-standard. This usually involves
restrictive relative clauses (which are much the more common), and most
usually (though by no means always) involves the use of the second comma
(39) Any students who have finished, may leave the hall.
The more specifically Australian pattern (if genuine) is represented by (40)
and (41).
(40) Joanne and Marie who had finished left the hall.
(41) Any students, who have finished, may leave the hall.
Those for whom this is a consistent pattern are obviously liable to interpret
sentences such as (42), encountered in reading, with a restrictive sense,
whereas if the writer has followed the traditional canon the sense is intended
as non-restrictive. Conversely, they may take (43) as non-restrictive, whereas
it is probably intended as restrictive.
(42) All the students, who had finished, left the hall.
(43) All the students who had finished left the hall.
The style manuals (e.g. Hudson 1993: 356; Renton 1994: 44, 48) mostly
rehearse the traditional rule, acknowledging that it is not absolute. Peters
(1995: 650) points out that some non-restrictive relative clauses, whose sense
is less obviously parenthetical, do not seem to require the commas so urgently;
although this proviso obviously does not apply to all non-restrictive relatives.
For more discussion, see Newbrook (1992), and for more on variation in
relative clause punctuation more generally see Newbrook (1988, 1992, 1997,
1998a) and references listed there.
Usage shared with AmE but not with (conservative) BrE
3.3.1 Epistemic have (got) to
It appears that younger Australians (like some younger British people) find the
epistemic use of have (got) to as in (44) — which is felt by me and many other
British people to be an Americanism — quite unremarkable. The more traditional British pattern is that have (got) to can substitute for must only in deontic
uses of the latter, as in (45). Collins (1989: 142) might be read as suggesting
that this “British” pattern applies in Australia also; but his focus here is upon
another issue (the presence or absence of got) and he now indicates (personal
communication) that his comments referred only to deontic meanings.
(44) Martina has (got) to be the best player in the world!
(45) Martina has (got) to be home by midnight!
3.3.2 Would (have) in the protases of remote conditionals
There is some evidence (Newbrook 1992: 22, 1993: 55) of the use in Australia, in both speech and writing, of traditionally American would (would have/
would’ve in the remote past) + main verb in the protases of remote conditionals, rather than the more usual British/Australian singly backshifted simple
past (remote present/future) and doubly backshifted past perfect (remote past
= counterfactual); see Section 3.2.3 above on these latter. In the American
usage (Trudgill and Hannah 1994: 60) the verb forms in the protases and
apodoses are, of course, parallel, which may be a factor encouraging this
usage. Compare (46) (remote present/future) and (47) (remote past) with
British usage as in (48) and (49) respectively.
(46) If they would do that it would be better.
(47) If they would have done that it would have been better.
(48) If they did that it would be better.
(49) If they had done that it would have been better.
The usage appears to be much more common with have, as in (47), than without,
as in (46). This may involve awareness of the common alternative forms -’d
have/-’d’ve and had have/had’ve (see below), which are associated with would
have/would’ve; there are no equivalent forms for would alone as in (46).
The style manuals do not seem to address this issue. The usage seems to
be not uncommon, especially in speech, and there are a few (but only a very
few) relevant tokens of would have (none, it seems, of would alone) in ACE.
The equivalent, derived/re-interpreted forms with had for would (see below)
are more numerous (though still not very numerous).
As noted, this feature interacts in a complex way with the second type of
“backshift avoidance” outlined in Section 3.2.3 above; it provides a further set
of variants differing from the traditional British usage. A still further possibility
(which is common in Australia but not distinctively or even especially Australian) involves the innovative forms had have/had’ve; these appear to represent
British/Australian re-expansions/re-interpretations of -’d have/-’d’ve, originally mainly American usage and derived from AmE would have/would’ve.
These latter forms are not at all familiar to most British people in these
constructions, and are probably rather unfamiliar to some Australians, despite
their obvious presence in some Australian speech. Yet a further possibility
involves the stigmatised set of forms would/had/-’d of, involving a re-interpretation of reduced have as of; this again is found both in the UK and in Australia
(and probably throughout the English-speaking world). For more on these
further variants, see, eg Newbrook (1989: 49–55), Peters (1995: 331).
3.4 Usage shared with standard varieties other than British or AmE
3.4.1 Epistemic mustn’t
There is considerable evidence (Newbrook 1992: 4, 1993: 54) that mustn’t
occurs quite freely in (some) AusE in its epistemic sense. In this respect,
Australian usage resembles Irish English (the most plausible source) and
regional varieties of English English (including educated varieties) which
have also been influenced by Irish English (notably that of Liverpool), but
differs from both Southern English English (where can’t is usual) and North
AmE (where uncontracted must not is typical, except before perfective have,
where mustn’t is possible). Compare (50) with the equivalent Southern English and North American forms (51) and (52).
(50) She mustn’t be at home; the light’s off.
(51) She can’t be at home; the light’s off.
(52) She must not be at home; the light’s off.
Item (51) is also possible in Liverpool but with a contrasting sense (greater
degree of conviction). It is not clear how widely this may be the case in Australia.
On the dialectological background, see Trudgill (1986: 140), Trudgill and
Hannah (1994: 61).
Australians appear quite markedly divided on the issue of the acceptability of sentences such as (50); but the style manuals do not generally discuss the
issue (possibly because it arises only when the modal is negated). There are
tokens in ACE, but not many.
3.4.2 Sentence initial as well
There is ample evidence in student writing and indeed in published AusE
(Newbrook 1992: 6–7, 1993: 54–55) of the use of sentence initial as well
(not as well as…), as in (53).
(53) As well(,) there are three other cases of this.
This usage is not reported as general in the UK or in the USA (though some
commentators believe that it is not salient to most Americans). It is, however,
reported (Trudgill and Hannah 1994: 78–79) as normal usage in Canada (and
is often found in Canadian publications); there is also some anecdotal evidence that it is common in South Africa. The details of the diffusion of this
usage would be of great interest.
As many as 71 instances were found in ACE: 64 cases with a comma
following, one with a dash and six with no punctuation. On the other hand,
Australians appear quite markedly divided on the issue of the acceptability of
sentences such as (53). Somewhat surprisingly, the style manuals do not
generally discuss the issue.
3.4.3 all/both…not
The interpretation of this usage varies in a dialectologically very complex
manner across the English speaking world. Because of inter-variety differences in respect of the interpretation of negative scope, sentences such as (54)
— and equivalent sentences with both — are susceptible to being taken in two
quite different senses, as equivalent either to (55) (normally taken to imply
(56), by implicature) or to (57).
(54) All the letters didn’t arrive.
(55) Not all of the letters arrived.
(56) Some of the letters arrived and some didn’t.
(57) None of the letters arrived.
For some speakers/writers, the usage appears ambiguous, but for others only
one sense appears possible; for me and many other British people, only (55),
for almost all South-East Asian users of English, only (57). For most of those
for whom (54) is equivalent to (55), it is rare: (55) generally appears a much
more plausible formulation of the message in question, and (56) is probably
still more usual. This is especially the case in writing, because of the associated phonology: in the sense of (55), (54) requires marked stress and intonation, which are not generally reproducible in written form (especially in print,
for stylistic reasons). On the other hand, many of those for whom (54) is
equivalent to (57) use (54) freely; the phonology is here quite unremarkable.
In South-East Asia, indeed, (54) is much the most common way of expressing
the sense of (57), and (57) may even appear unfamiliar to some.
On the basis of my experience with student writing, speech and reactions
when questioned, I have suggested (Newbrook 1992: 16–17, 1993: 55) that
some younger Australians, in particular, tend to interpret (54) as uncontroversially equivalent to (57) (like South-East Asians). This is still to be examined
in a more systematic way. The construction is rare (perhaps predictably; see
above) in ACE; and the style manuals do not discuss the issue.
3.4.4 will in offers, suggestions etc.
Trudgill and Hannah (1994: 97) (along with other sources) draw attention to
the Scottish (and, less saliently, North American) use of will in offers, suggestions etc (58), where most other varieties have shall (this is one of the few
remaining constructions where shall is common). Although they note
(1994: 19) that will is generally preferred to shall in Australia, they do not
identify this specific use of will as Australian, and the style guides do not
explicitly refer to it; however, it does occur quite frequently, and it is very
striking to those for whom will never occurs in such cases, such as most English
people. However, there appear to be no cases in ACE.
(58) Will I close the door?
3.5 Other interesting syntactic features of AusE
Some other syntactic features which appear (to a degree, and on current
evidence) distinctively Australian are not likely to become candidates for
standard status in the short term. The types listed below are also absent from
ACE. However, given the “mainstream” nature of AusE syntax, any such
features are still worthy of note. Some have emerged from the current research
project on AusE being conducted by the Department of Linguistics at Monash
University, and/or from earlier work conducted in Victoria by Monash linguists. They include:
a) the non-use of the definite or, more usually, the indefinite article
(Newbrook 1998b; Clyne and Bouma 1994)
(59) He’s very good batsman.
b) the inflected comparison/superlativisation of polysyllabic adjectives, including participial adjectives (Newbrook 1992: 20)
(60) She’s a much confidenter shooter these days.
(61) He’s the winningest coach in the league.
c) the use of genitives with relative that, as in some American usage
(Newbrook 1990: 107–108, 1992: 22, 1993: 25, 1997: 41, 1998a: 53)
(62) This is the girl that’s book I borrowed.
Although the features discussed above may appear rather few and, in a
number of cases, of doubtful validity, it may be said that — even in respect of
careful, formal usage — AusE is not wholly lacking in distinctiveness at the
syntactic level. Furthermore, some of the relevant features appear to be
perceived as norms, at least by some Australians, and thus to constitute
plausible candidates for inclusion in any more overtly de iure version of
Standard AusE which may be emerging. But there is clearly still much to be
learned: about the forms themselves, about their geographical, social and
stylistic distribution, and about Australians’ attitudes and perceptions regarding them. The way is open to further research along all these lines.
Cheshire, Jenny. 1982. Variation in an English Dialect: a Sociolinguistic Study. Cambridge: Cambridge University Press.
Clyne, Michael and Gary Bouma. 1994. “Talking about one’s life and faith”. Text 14: 167–
Collins, Peter C. 1979. “Elicitation experiments on acceptability in Australian English”.
Working Papers of the Speech and Language Research Centre, Macquarie University. 2
(4): 1–49.
———. 1988. “Semantics of some modals in Australian English”. Australian Journal of
Linguistics 8: 233–258.
———. 1989. “Divided and debatable usage in Australian English”. In Collins and Blair,
eds. 1989: 138–149.
——— and David Blair, eds. 1989. Australian English: the Language of a New Society. St
Lucia: University of Queensland Press.
Crystal, David. 1984. Who Cares about English Usage? Harmondsworth: Penguin.
Eisikovits, Edina. 1989. “Girl talk/boy talk: sex differences in adolescent speech”. In
Collins and Blair, eds. 1989: 35–54.
Follett, Wilson. 1966. Modern American Usage. New York: Longman.
Huddleston, Rodney. 1988. English Grammar: an Outline. Cambridge: Cambridge
University Press.
Hudson, Nicholas. 1993. Modern Australian Usage. Melbourne: Oxford University Press.
Mittins,W.H. 1969. “What is correctness”. Educational Review 22: 51–63.
Newbrook, Mark. 1988. “Relative clauses, relative pronouns and Hong Kong English”.
University of Hong Kong Papers in Linguistics 11: 25–41.
———. 1989. “Some neglected syntactic phenomena in near standard English”. Chinese
University of Hong Kong Papers in Linguistics 1: 46–69.
———. 1990. “Some notes on dialect in recent English literature”. Chinese University of
Hong Kong Papers in Linguistics 2: 105–110.
———.1992. “Unrecognised grammatical and semantic features typical of Australian
English: a checklist with commentary”. English World-Wide 13: 1–32.
———. 1993. “Grammatical features of standard Australian English”. In Pam Peters, ed.
Style on the Move: Proceedings of Style Council 92. Sydney: Macquarie University
Dictionary Research Centre, 46–56.
———. 1997. “Relative that, that isn’t always restrictive”. ITL Review of Applied Linguistics 115–116: 39–60.
———. 1998a. “Which way? That way? Variation and ongoing change in the English
relative clause”. World Englishes 17: 43–59.
———. 1998b. “Teenage talk”. Australian Style 6 (2): 3.
——— and Yio Siew Koon. 1987. “Patterns of usage in comparative and related constructions in Singaporean English”. In Mark Newbrook, ed. Aspects of the Syntax of
Educated Singaporean English: Attitudes, Beliefs and Usage. Bern/Frankfurt: Peter
Lang, 62–152.
Partridge, Eric. 1947. Usage and Abusage. London: Hamish Hamilton.
Pawley, Andrew. 1995. “Some characteristics of the Tasmanian English Vernacular”. Paper
presented at the ALS Conference, ANU, Canberra, 1995.
Peters, Pam. 1995. The Cambridge Australian English Style Guide. Cambridge: Cambridge
University Press.
Renton, Nicholas E. 1994. Good Writing Guide. Melbourne: The Business Library.
Shnukal, Anna. 1989. “Variable subject relative pronoun absence in Australian English”. In
Collins and Blair, eds. 1989: 70–77.
Sussex, Roly. 1989. “The Americanisation of Australian English: prestige models in the
media”. In Collins and Blair, eds. 1989: 158–168.
Taylor, Brian. 1989. “American, British and other foreign influences on Australian English
since World War II”. In Collins and Blair, eds. 1989: 225–254.
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Blackwell.
——— and Jean Hannah. 1994. International English: a Guide to the Varieties of Standard
English. London: Edward Arnold.
Watson, Keith. 1978. “Teachers attitudes to usage”. ALAA Occasional Papers (Applied
Linguistics Association of Australia) 2.
Australian English and indigenous voices
Bruce Moore
Australian National Dictionary Centre,
Australian National University
Gladys Tybingoompa could contain her exuberance no longer. She reached
into her handbag, produced a pair of clap-sticks and whirled into a wild song
and dance of victory. For a moment or two, every face around the normally
sombre precincts of the High Court of Australia appeared to be wreathed in
smiles as Ms Tybingoompa leapt and kicked through the dance she called
‘Shake a Leg’. … She had travelled all the way from Cape York to Canberra
to hear the High Court’s opinion of her people’s rights to their traditional
land. … ‘This is the Christmas present I’ll take home to the Wik people’.1
The quiet despair was obvious on Gladys Tybingoompa’s face. The senior
Wik woman, whose dances of joy have become two of the most enduring
images during the divisive native title debate, was conceding defeat with the
parliamentary process. ‘This is called the walk away, the last time the message is given to you, the Australians. Put your hands together like the sea of
hands to give support to us all,’ she said. She then turned to the handful of
Wik people, who had travelled to Canberra to sit patiently through the Senate
debate for the past two weeks, and called out in their language. Slowly and
purposefully, they walked away from Parliament House. It was a sad day for
the woman who had danced out the front of the High Court when the Wik
decision was announced and, just last week, had coaxed a barefoot Brian
Harradine to join the ‘shake-a-leg dance’ on the lawn of Parliament House.2
These two passages offer interesting evidence for the process of language
creation in contemporary Australia. Before this native title case, most Australians would not have known of the Wik people. They would have known the
word taipan for the snake Oxyuranus scutellatus of northern Australia; they
would probably not have known that the word taipan is a borrowing from the
Wik-Mungkan language, first recorded in English in 1933. As a result of the
native-title case the term Wik is known by all, and has entered Australian
English (henceforth “AusE”), whether in the form the Wik people, the Wik
decision of the High Court, or simply Wik. Similarly, most Australians
would not have known the term shake-a-leg. It is first recorded in Aboriginal
poet Lionel Fogarty’s Yoogum Yoogum (1982, Ringwood: Penguin Books),
and it is described in the journal Aboriginal History (7, 1983: 161): “Shake-aleg is performed solo or as part of a small group with the dancers using an
extreme spread leg movement”. It is included in J. M. Arthur’s Aboriginal
English (1996: 56): “In traditional shake-a-leg dancing, the participants stand
in a ring, and one at a time come forward and perform a dance, which involves
rapid in-and-out movement of the knees. The emphasis in the dance is on the
virtuoso performance of this action.” Until 1996, shake-a-leg was a term
which belonged to the Aboriginal community. But here we witness a moment
of transition, as a term which existed in Aboriginal English (henceforth
“AbE”) moves out into the wider community of AusE.
The passages also add to the evidence that AusE is experiencing a new
influx of words from Aboriginal culture. Dixon, Ramson and Thomas (1990),
drawing largely on the evidence of The Australian National Dictionary, list
some 400 Australian words borrowed from 80 Aboriginal languages. Most of
these words were borrowed in the early days of European settlement, and most
refer to the physical world — words for new fauna and flora, features of the
environment, and implements. Borrowings later in the nineteenth century do
not reflect a continuing interest in Aboriginal culture in settled areas; rather,
they reflect the fact that the frontier was late in moving to some areas. Thus
alcheringa (‘Dreamtime’) is not recorded until 1897, but this merely reflects
the lateness of the work of anthropologists among the Arrernte people. In
addition to the borrowings from Aboriginal languages, there were also many
English words used to describe Aboriginal life. Most of these appeared in the
nineteenth century: bark canoe (1830), country (1843), bullroarer (1848),
point the bone (1884), bush tucker (1895), bark painting (1897), etc. In the
first sixty years of the twentieth century there were very few new terms; the
smattering includes fringe-dweller (1959), sacred site (1933), and x-ray art
The history of the word Moomba sums up the first sixty years of the
twentieth century. Moomba is an annual carnival held in Melbourne since
1955, at the beginning of March. The main event has always been a parade of
floats through the streets of the city. In 1955 the Melbourne Herald (12 March,
1/1) reported: “All was set for the Governor, Sir Dallas Brooks, to open the
Moomba officially”. There was no doubt genuine goodwill in the choice of an
Aboriginal name for the festival. Craig McGregor (1966: 69) writes:
“Melbourne’s Moomba (an aboriginal word meaning ‘Let’s get together and
have fun’) [is] a yearly event during which floats parade through the city”. But
where had this term come from? Barry Blake wrote (1981: 84): “Undoubtedly the most unfortunate choice of a proper name from Aboriginal sources
was made in Melbourne when the city fathers chose to name the city’s annual
festival ‘Moomba’. The name is supposed to mean ‘Let’s get together and
have fun’, though one wonders how anyone could be naive enough to believe
that all this can be expressed in two syllables. In fact ‘moom’ (mum) means
‘buttocks’ or ‘anus’ in various Victorian languages and ‘ba’ is a suffix that can
mean ‘at’, ‘in’, or ‘on’. Presumably someone has tried to render the phrase ‘up
your bum’ in the vernacular”.
At the end of the twentieth century the report card is starting to show a
different story. This is most evident in the Northern Territory. In 1990, Dolly,
a magazine aimed at teenagers, published the results of a reader survey of
contemporary slang (October, 1990: 62). In one part of the survey readers had
been asked to give terms which describe “the guy you’ve spotted at the disco
who looks like a cross between Tom Cruise and Johnny Depp”. More than 100
terms are listed. The most popular term in Queensland was honey, in South
Australia it was gorgeous, and in Tasmania and the ACT it was hunk, but in the
Northern Territory it was budju. This word, from the Larrakia language of the
Darwin area, was a word for women’s genitalia, was then used as a term for
“cunt” in AbE (You budju),3 and then surfaced in teenagers’ slang as a word for
a good-looking person of either sex (a shift of meaning comparable to teenagers’ use of wicked or sick to mean ‘great’). Maluka (also maluga, malaga) is
a word from the Djingulu language of the Northern Territory, where it meant
‘old man’. It entered AbE (via Australian pidgin), where it came to mean ‘boss,
the person in charge’. With that meaning it has now entered AusE in the
Northern Territory. Gammon, originally meaning ‘guile, deceit’, is obsolete in
standard English, but it was retained in AbE in the sense ‘nonsense, bullshit’.
With this meaning it has re-entered the wider community in Darwin. Deadly
came to mean ‘fantastic, awesome’ in AbE, and this sense too has spread to the
wider community. Shame job, meaning ‘an event which causes a person
shame or embarrassment’, is also used widely in Darwin among teenagers,
although this is a term which is used in AbE throughout Australia, and which
has entered the language of teenagers in many parts of Australia.
The large Aboriginal population in the Northern Territory explains the
number of words borrowed into the Territorians’ AusE from AbE. Elsewhere
in Australia the borrowings and new terms have not occurred at the colloquial
level. Yet from the late 1960s there has been a new influx of words into AusE
from Aboriginal languages and Aboriginal culture. This undoubtedly reflects
significant changes taking place in Australian society. The influx parallels the
development of Aboriginal political and cultural activism, but it also goes
hand in hand with an increasing interest in Aboriginal languages and culture
on the part of non-indigenous Australians.
Fighting back
In the first half of the twentieth century it was assumed that Australia would be
a monolingual and monocultural society, and that Aborigines would be “assimilated”. Aboriginal languages would die, and Aboriginal culture would be
remembered through the grainy black-and-white films of anthropologists. But
the culture refused to die, and it fought back.
From an Aboriginal point of view, January 26 1938, the sesquicentenary
anniversary of the arrival of the First Fleet, is a significant date. While a reenactment of the arrival was held in Sydney, a group of Aborigines observed a
Day of Mourning. In 1964 the term land rights is first heard in AusE.4 The
term became especially prominent in 1966 with the beginning of a strike by
Gurindji pastoral workers in the Northern Territory. What began as a strike
over pay and conditions developed into a protracted battle for land rights. In
1965 a group of students from the University of Sydney, including Charles
Perkins among its leaders, conducted a 3,200 km bus tour of northern New
South Wales, protesting against the discrimination levelled at Aborigines.
These became known as the freedom riders. In the late 1960s the outstation
movement began. This occurred in northern and central Australia, with
groups of Aborigines moving away from large centres back to their traditional
countries, although maintaining links with the large centres for supplies and
services. The community itself was called an outstation, by analogy with the
grazing property term, where there was often a subordinate station some
distance from the main establishment. By the mid 1970s outstation and
outstation movement had been largely replaced by the more accurate homeland and homeland movement.
On 26 January 1972 the Aboriginal tent embassy was erected on the
lawn outside Parliament House, drawing attention to the denial of rights to
Aborigines, and called embassy to draw attention to the fact that many
Aborigines felt that they were foreigners in their own land. In the same year the
Aboriginal flag (with the top half black, representing the people, the bottom
half red for the earth, and a central solid yellow circle representing the sun) was
officially adopted by the Aboriginal people. In 1973 the term land council
makes its first appearance in AusE, with a recommendation that two land
councils be set up in the Northern Territory, bodies appointed to represent the
interests of Aborigines with regard to Aboriginal land. In 1974 the term
traditional owner appears to describe an Aborigine who is a member of a local
descent group having certain rights in a tract of land. With the approach of the
1988 bicentenary celebrations, Aboriginal activism increased. The term Day
of Mourning was revived, and a new term to describe January 26, Invasion
Day, was introduced. Other key terms appear. Aboriginality is used from
1977, but its association with issues of identity and pride in being Aboriginal
comes to the fore especially after the bicentennial protests: “Warburton Ranges
principal Wilbur Klein said the desert dust-up, which was in its third year,
aimed to give children a chance to get together and to be proud of their
Aboriginality” (West Australian. 7 Sept. 1990: 1/5). Aboriginalisation in
the sense ‘affirmative action to enable Aborigines to take up positions in
organisations etc., especially those dealing with Aboriginal affairs’ first appears in 1972, but its widespread use is in the late 1980s. The corresponding
verb, Aboriginalise, also belongs to this period. The process of Aboriginalisation meant the creation of terms to describe Aboriginal-designated positions,
such as Aboriginal Education Officer, Aboriginal Field Officer, Aboriginal Police Aide, and Aboriginal Teacher Aide. As these advances were being
made, attention was also paid to the disturbingly high numbers of deaths of
Aborigines, especially young Aborigines, in prisons and youth detention
centres. The mid 1980s saw the introduction of the phrase black deaths in
custody. In 1987 the Royal Commission into Black Deaths in Custody was
In 1992 the Mabo decision of the High Court was handed down, soon
after generally known as Mabo. Eddie (Koiki) Mabo, an Australian Mer
(Murray) Islander in the Torres Strait, had instituted legal proceedings to
establish the Mer people’s traditional ownership of their land. The High Court
judgment of 1992 recognised the continuous possession of their land by the
Torres Strait Mer Islanders before these lands were annexed by Queensland. It
was this judgment of the High Court which brought the term native title into
AusE.5 And in refuting the legal fiction that Australia, prior to white settlement, was terra nullius (‘land belonging to no-one’), the High Court made an
obscure Latin legal phrase, terra nullius, a household word.
In 1994 the then Federal Minister for Aboriginal Affairs, Robert Tickner,
acting on a report which examined the sacred-secrets of a group of
Ngarrendjeri women, placed a 25-year ban on the building of a bridge between Hindmarsh Island and the South Australian mainland. The decision was
overturned on a legal technicality in 1995, and a South Australian Royal
Commission found the evidence to be a “fabrication”. The dispute is a
continuing one, but one of its results was that the term secret women’s
business (sometimes abbreviated to women’s business) came into AusE. By
analogy with this term, the term secret men’s business was created. The terms
can be used in a variety of ways: “Kingswood driving is secret men’s business
— just like pushing a shopping trolley straight is secret women’s business”
(New Idea. 29 Nov. 1997: 20); “Subscribe now & we’ll send Ralph over
every month! Australia’s newest magazine that understands Secret Men’s
Business” (Ralph. Oct 1997: 65). The uses are often jocular, but they depend
for their point on a knowledge of the original term. The basic structure can be
varied: “Talking to boys about boys’ stuff was one thing: discussing sex and
secret girls’ business with an uninitiated group of 30 boys and girls in one
room was enough to make a grown man feel faint” (The Australian. 10 Jan.
1998 (Review): 10/1).
The term stolen generation(s) became widely known in 1995 with the
establishment of the National Inquiry into the separation of Aboriginal and
Torres Strait Islander children from their families.6 One of the recommendations of the Inquiry was that a national Sorry Day should be held each year to
commemorate the history of forcible removal and its effects. 26 May 1998
was the first National Sorry Day, and it led to the production of sorry books.
And finally, in 1996, there was Wik.
Here is a series of culturally and politically significant terms, many of
them attributable to Aboriginal activism, but many of them a result of concern
by non-indigenous people that reparation should be made for the wrongs of
the past.
A living culture
At the same time there was a resurgence of interest in Aboriginal culture.
Typical of this resurgence is the following:
The Shepparton Aboriginal Keeping Place operates as a museum and a
cultural centre. The Shepparton Aboriginal Keeping Place houses artifacts
which are owned by the local Aboriginal community, these are preserved and
maintained to ensure that they are kept in perpetuity. The displays of the
artifacts and dioramas which depict realistic scenes of traditional lifestyles
from this region offer visitors, both national and international, an insight into
and appreciation of Aboriginal culture. … More importantly the Keeping
Place occupies the position as guardian of the culture.7
A keeping place is an Aboriginal cultural centre which has the primary
purpose of “keeping” the culture preserved (as distinct from the European
term “museum” which typically enshrines the past). Such keeping places also
act as community centres. They are sometimes called living cultural centres.
Dreamtime and dreaming, translations from the Arrernte word altyerre,
were well established in AusE by the 1960s, but as understanding of Aboriginal culture increased, terms such as dreaming track (or path) to describe the
path followed by a dreamtime being through the landscape, entered the
language: “Parngurr is an important sacred site on the Dreaming Track from
Poebourne, through Karunjini, the Mamersley Range, Marble Bar, Nullagine,
Pangu, Well 33 and Kiwi Kurra toward Alice Springs” (North West Telegraph
(Port Hedland). 15 May 1991: 1/6). This dreaming track may also be called
a song line. The dreamtime beings may be called ancestor spirits. In central
Australia the Pitjantjatjara word tjukurpa meaning ‘Dreaming, Law’ is gradually replacing the English term. Smoking ceremonies, in which smoke is used
for ritual purposes, especially after death, are mentioned in the early literature,
and then from the 1970s in works of AbE. In the 1990s, the term moves into
mainstream AusE.
Some terms for Aboriginal art have been in AusE for a long time,
including bark painting (1897), mimi art (1949) and Quinkan (1969), but
these terms have increased in recent years. These include: acrylic art, the term
for a style of painting used by central Australian Aboriginal artists, constructing traditional designs with the aid of Western acrylic paint and canvas;
figurative art, the term for a style of (especially ancient) Aboriginal art in
which human, animal, or mythological figures predominate, as distinct from
Aboriginal art consisting of circles, lines, and other such non-figurative ele-
ments; Gurangara art, a term for a style of ancient Aboriginal art from the
Pilbara region of Western Australia, consisting mostly of detailed and lively
rock engravings, depicting animals and humans, the latter often with sexual
organs carefully depicted; lightning figure, a term used to describe figures
depicted in ancient rock paintings in northern Australia, associated with the
“dry lightning” which appears some weeks before the rains; stencil art, a form
of ancient art on rock, depicting patterns of hands, human figures, animals,
etc., and created by holding the object to be depicted against a rock surface
and spraying mouthfuls of liquid pigment around the object, so creating a
negative image; urban art, the term used to describe a style of Aboriginal art
developed by city-dwellers; wandjina art, a style of art found in the
Kimberley region of Western Australia depicting large figures of the
Wandjinas and their adventures (the Wandjinas control the monsoons). These
terms recognise the diversity of Aboriginal culture in the past and especially in
the present. Cultural tourism is the term used to describe various ways of
making aspects of Aboriginal culture accessible to the wider community —
the display and sale of arts and crafts, dance performances, educational tours
of traditional lands, and such.
Since the publication of such works as Kevin Gilbert’s Living Black
(1977) there have been many works published by Aboriginal writers. Some of
these works are written in AbE, or include words from AbE. Sally Morgan’s
My Place (1987) includes boolyah man and maban for a spiritually powerful
person. Ruby Langford Ginibi’s Real Deadly (1992) includes dubay
‘woman’, bagal ‘an Aboriginal person’, jarjum ‘a baby’, gungabul ‘a policeman’, and gubba ‘a white person’. These works are often written in life story
genre. While the words remain, technically, in AbE (they are usually glossed
in the texts), it is through such writings that more terms are likely to spread
into AusE.
Reclaiming the past
Australians undoubtedly use, or have used, many more Aboriginal words than
the 400 which are listed in Australian Aboriginal Words in English: Their
Origin and Meaning. There are many other words in AusE, mainly terms for
flora and fauna, which are probably of Aboriginal origin. These are the terms
which dictionaries provide with a “probably from an Aboriginal language”
etymology. The paucity or absence of records of some languages makes
precise attribution impossible in many cases. Nevertheless, there has been
some success in reclaiming the past.
Recent work at the Australian National Dictionary Centre has filled in
some of the gaps, with the original languages of some 60 more words identified. The Western Desert language is the source of a number of terms for
mammals, including: dargawarra, the hopping-mouse Notomys alexis of arid
western and central Australia; mingkiri (also menkie) the mouse Pseudomys
hermannsburgensis of sandy spinifex country from the Pilbara through to
south-western Queensland; oorarrie, the fawn-coloured hopping-mouse
Notomys cervinus of gibber plains of central Australia; pallyoora, the rodent
Pseudomys australis, once known from various locations in central and eastern Australia, but now restricted to the gibber plains of the Lake Eyre basin;
pitchi-pitchi and wuhl-wuhl, names for the long-legged marsupial mouse
Antechinomys laniger spenceri; wilkintie, the rare hopping-mouse Notomys
fuscus which inhabits sand-dunes in central Australia; wintarro, the bandicoot Isodon auratus, formerly widespread over arid central Australia, but now
limited to the Kimberley and to a thriving population on Barrow Island off the
Western Australian coast. Other newly identified borrowings of terms for
mammals include the following: mongan, the ringtail possum Pseudochirulus
herbertensis of north-east Queensland, is from Warrgamay (Herbert River,
north Queensland); tcharibeena (also jarabeena), a name for the tree-kangaroo Dendrolagus bennettianus which inhabits a small area north of the
Daintree river in Queensland, is from Kuku Yalanji (Bloomfield River area of
northern Queensland); toolah, a name for the ringtail possum Pseudochirops
archeri of a small area of rainforest in north-east Queensland, is from
Similarly, the original languages of many common plant names have been
identified. These include the following: amulla, the shrub or trailing plant
Myoporum debile, of New South Wales and Queensland, is from Darambal
(Rockhampton region); burgan, the shrub Kunzea peduncularis (formerly
known as Leptospermum phylicoides) of Victoria, New South Wales, and
Queensland, is from Wuywurung; chittick (also chittock), the medium-sized
shrub Lambertia inermis of Western Australia, is from Nyungar; illyarrie, the
small tree Eucalyptus erythrocorys of Western Australia, is from Nhanta;
jerry-jerry, the annual herb Ammannia multiflora of Queensland, is from
Mayi-Yapi and Mayi-Kulan, spoken around the Cloncurry River and the
Norman River, north Queensland; midyim, the shrub Austromyrtus dulcis of
coastal areas between Grafton in NSW and Fraser Island in Queensland, with
small edible berries, is from Yagara; migum, the tree Eucalyptus leucophloia
of the Pilbara area of Western Australia, is from Yindjibarndi; millaa-millaa,
the climbing shrub Eleaganus latifolia of rainforests north from the central
Queensland coast, is from the Ngajan dialect of Dyirbal; wynnum, the small
tree Pandanus pedunculatus of coastal districts of New South Wales and
Queensland, or its northern species Pandanus tectorius of coastal districts north
of Brisbane, with large fruits which resemble a pineapple, is from Yagara.
These are terms which have existed in AusE for some time, but it is only
now that we can definitely say that they are borrowings from Aboriginal
languages. In recent years there has been a move to replace European names of
flora and fauna with Aboriginal names. Braithwaite et al. (1995: 3) lament the
fact that while the first edition of Australian Aboriginal Words in English “lists
almost fifty names for marsupials, it provides no names for rodents”. They note
that many rodents “have been burdened with common names that are
uninspiring, unwieldy or actually downright ugly”, and ask, for example: “How
has it come about that one of Australia’s most attractive rodents, Pseudomys
australis, is handicapped with a name — the Plains Rat — that seems designed
to arouse contempt?” (1995: 1). Braithwaite et al. recommend 59 names from
Aboriginal languages. R. Strahan, in his A Dictionary of Australian Mammal
Names (1981: 181–2), expressed some reservations about the moves to introduce Aboriginal names for mammals into the lexicon of AusE:
As a matter of conscience, belated patriotism, or as a means of replacing long
names by shorter ones, many mid-twentieth-century naturalists have attempted to resurrect forgotten Aboriginal names for currently recognised
species. Some, such as Quokka (Setonix brachyurus), Numbat (Myrmecobius
fasciatus), Dibbler (Antechinus apicalis), Kowari (Dasyuroides byrnei),
Dunnart (Sminthopsis), and possibly Mulgara (Dasycercus cristicauda),
Kultarr (Antechinomys), and Bobuck (Trichosurus caninus) are becoming
generally or locally accepted. Others such as Marl, Mundarda, Noolbenger,
Wurrup, Dalgyte, Wuhl-wuhl, Pitchi-pitchi, and Wilkintie, advocated by
some Western Australian naturalists, have not penetrated into general usage.
Marl, mundarda, noolbenger, and dalgyte were in the first edition of Aboriginal Words in English. In the 1995 edition of his The Mammals of Australia,
Strahan still describes mundarda, noolbenger, dalgyte and wurrup as indigenous names, but allows marl, wuhl-wuhl, pitchi-pitchi, and wilkintie as
common names.
It is now a common practice to use Aboriginal names for new species.
The small marsupial Dasykaluta rosamondae found in areas of woolly spinifex in the Pilbara region of Western Australia, was not described until 1964,
and was then wrongly regarded as an antechinus. The genus name was
suggested in 1982: dasy refers to its membership of the genus Dasyurus
(Greek dasus ‘rough, hairy’ + oura ‘tail’), and kaluta is the name for the
creature in the Nyamal language (Archer, ed. 1982: 435). Kaluta has now
become the standard common name. Australia’s smallest rock wallaby,
Petrogale burbidgei, which is restricted to rugged, inhospitable parts of the
Kimberley region, Western Australia, was identified only very recently, and
initially it was called warabi (from Wunambul, north Kimberley). The new
name, monjon, also from Wunambal was proposed in 1990. More problematic was the process of naming the new genus Ningaui. The genus name was
given in 1975 to a group of very small marsupials, weighing between two and
ten grams. Archer (1975: 243) suggested the name: “The generic name
Ningaui is here given masculine gender. It is an Aboriginal name given to tiny
mythological beings that are hairy, have short feet, and only come out at night
to hunt for food all of which is eaten raw (Roberts and Mountford 1969). The
allusion to these dasyurids involves their very tiny size, hairy and (compare
with the related dasyurid Sminthopsis spp.) short feet, and nocturnal habits”.
Roberts and Mountford (1969: 54) do not identify the language. But it is clear
from context that their story refers to Melville Island, and that the term is from
Tiwi nyingawi ‘short ghost’.
There is a comparable move to use indigenous names for flora, especially
in Western Australia and the Northern Territory (Bennett 1993; Strong 1987;
Abbott 1983). The bush-tucker industry is leading to a knowledge of bushtucker terms in the wider community. In cookbooks the term warrigal greens
is commonly used, with recipes such as “Warrigal Pasta”, “Warrigal Pasta
Salad”, and “Warrigal Greens Soup” (Cherikoff and Isaacs 1989: 96–7).
Muntry (the edible fruit of the shrub Kunzea pomifera) is recorded from the
nineteenth century, but it is now widely used in association with bush tucker:
“Other commercially used foods include rosella flowers, Kakadu plum, lemon
myrtle, aniseed myrtle, muntries, lemon aspen, native currants, riberries,
native peppermint, bush tomato, wild lime, quandong, bush peppercorn, pepper leaf, and warrigal greens”.8 One of the most widely used terms in the bush
foods industry is akudjura. This is a term for the bush tomato, and the name
was chosen from the Alyawarr language of the Northern Territory. Internet
sites provide such recipes as “Akudjura crusted blackened salmon cutlets” and
“Akudjura tapinade and bush tomato oil”.9 A Sydney restaurant offers an
entree of “prawns stuffed with seafood kibble fritto misto, rocket & baby
endive, akudjura dressing”.10 In regional contexts, the use of Aboriginal plant
names to describe bush tucker is even more widespread. A book on bushtucker plants from south-west Western Australia, geared to a popular audience
(Daw et al. 1997), lists such names as tjunguri (Thysanotus patersonii),
nyilla-nyilla Lysiana casuarinae), mooja (Nuytsia floribunda), djubak
(Burnettia nigricans), koola (Podocarpus drouynianus), poolgarla (Banksia
grandis), and cadgeegurrup (Persoonia saccata). It is likely in the future that
more and more indigenous names will replace European names for Australian
fauna — just as the Australian Easter bilby is in the process of replacing the
European Easter bunny.
What’s in a name?
The explorer T.L. Mitchell was sympathetic to the use of Aboriginal place
names, but the sight of a mountain range in Victoria brought forth British
memories and loyalties:
In adding this noble range of mountains to my map, I felt some difficulty in
deciding on a name. To give appellations that may become current in the
mouths of future generations, has often been a perplexing subject with me,
whether they have been required to distinguish new counties, towns, or
villages, or such great natural features of the earth, as mountains and rivers. I
have always gladly adopted aboriginal names, and in the absence of these, I
have endeavoured to find some good reason for the application of others,
considering descriptive names the best, such being in general the character of
those used by the natives of this and other countries. Names of individuals
seem eligible enough, when at all connected with the history of the discovery,
or that of the nation by whom it was made. The capes on the coast, I was then
approaching, were chiefly distinguished with the names of naval heroes; and
as such capes were but subordinate points of the primitive range, I ventured to
connect this summit with the name of the sovereign in whose reign the
extensive, valuable, and interesting region below was first explored; and, I
confess, it was not without some pride, as a Briton, that I … gave the name of
the Grampians, to these extreme summits of the southern hemisphere.11
Mitchell’s Grampians underwent a name change in Victoria in 1991 when its
Aboriginal name, Gariwerd, was adopted. A change of government brought a
reversal, but in much of the tourist promotional literature from the area, both
Gariwerd and Grampians are used. This is becoming more common — in
many places in Australia the European name now exists alongside the Aboriginal name.
Typical of the changes taking place in central Australia are the following
extracts from a tourist guide:
Uluru-Kata Tjuta National Park covers an area of 126,132 hectares. Uluru
(Ayers Rock) and Kata Tjuta (Mt Olga) are amongst the world’s great natural
wonders. … [At Uluru] there are three free ranger guided tours. … The
Kuniya Walk will explain to you about the Kuniya, the ancestral python and
its involvement in the creation of the surrounding landscape features. You
will also be informed on some of the details of the Tjukurpa and what it means
to the Anangu in the context of today’s world. … There are numerous fruits in
the area, which play a role in the Anangu diet. These include kampurarpa, the
desert raisin and tjilka, wild tomatoes.12
Here is the term Tjukurpa being used as standard English (for Anangu, see
below). The Aboriginal names for fruits are explained, but the Aboriginal
words are given syntactic priority. Aboriginal terms — the Kuniya Walk —
now define the landscape. The shift from Ayers Rock (named in 1873 by
William Gosse after the then Premier of South Australia Henry Ayers) to the
Aboriginal name Uluru is well established in AusE. The shift from the Olgas
(named in 1872 by Ernest Giles after Queen Olga of Württemberg) to Kata
Tjuta is still in the process of change, but it is significant that tourist material
of this kind gives primacy to the Aboriginal name. Elsewhere, the Katherine
Gorge National Park (Northern Territory) is also the Nitmiluk National Park,
the Coburg Peninsula National Park (Northern Territory) is also the Gurig
National Park, and the Jervis Bay National Park (NSW) is also the Booderee
National Park.
Non-indigenous Australians are also becoming more familiar with the
names of Aboriginal peoples. As a result of native-title claims, some of the
names of Aboriginal peoples now have Australia-wide currency: Meriam
Mir, Wik, Yorta Yorta, Ngarrendjeri. This is especially true of regional
areas, where the names of Aboriginal peoples will be known in a way they
were not twenty years ago — the Larrakia of the Darwin area, the Wiradhuri
of New South Wales, the Ngunnawul of the Canberra region, the Nyungar of
south-west Western Australia, the Wathaurong of southern Victoria. A significant shift away from the use of the terms Aborigines or Aboriginals has
occurred in regional areas. Many Aborigines understandably dislike the use of
Aborigine or Aboriginal since these terms have been foisted on them and often
carry with them negative cultural baggage. They prefer to use the word for
“person” from a local language. Two of these words, Koori and Murri,
appeared in the first edition of The Macquarie Dictionary (1981). Koori is
from Awakabal (Newcastle region) and neighbouring languages, and is used
in most of New South Wales and Victoria. Murri is from Kamilaroi (eastern
New South Wales) and from many languages in south and central Queensland,
and is used over most of south and central Queensland. These terms entered
mainstream AusE during the 1980s. The Australian National Dictionary
(1988) included more of these terms: Nyoongah is from the Nyungar language of south-west Western Australia, and is used in the Perth area; Wongi is
from the Western Desert group of languages, and is used in the Kalgoorlie
area; Yammagi is from the Watjari language (Murchison River, Western
Australia), and is used in the Murchison River area. Australian Aboriginal
Words in English (1990) added more: Bama is from many north Queensland
languages, and is used in north Queensland; Mulba is from the Panyjima
language (north-west Western Australia) and is used in the Pilbara region;
Nunga is from the Nhangka language (south-west South Australia) and is
used in southern South Australia; Yolngu is from the the Yolngu languages of
north-eastern Arnhem Land, and is used in that area; Yuin is from the
Adnyamathanha language (South Australia), and is used in the Flinders
Ranges area. The second edition of The Australian Concise Oxford Dictionary
(1992) added two more: Anangu is from Pitjantjatjara and is used in the
Western Desert of central Australia; Yura is from the Adnyamathanha language and is used in the Flinders Ranges area of South Australia. In Tasmania
the term Koori was used for a time, but this was a term closely associated with
the mainland. An alternative muttonbird koori appeared briefly, but the
official term is now Palawa.13 This term appeared in the third edition of The
Australian Concise Oxford Dictionary (1997). Finally, the terms saltwater
people and freshwater people are now commonly heard to describe an
Aboriginal people who live by the sea or along inland watercourses.
Who’s sorry now?
The word makarrata is from the Yolngu language of Queensland, and it
refers to an Aboriginal ritual symbolising the restoration of peace after a
dispute. In 1979 the word was chosen by the National Aboriginal Conference
to refer to a proposed treaty between the Australian government and the
Aboriginal nation. The Aboriginal Treaty Committee drafted a treaty, and
from 1981 to 1983 a Senate Standing Committee on Constitutional and Legal
Affairs considered the matter. But the idea of the makarrata languished, and
the term has dropped out of use. Talk of a treaty and reconciliation continues,
but the possibility of reconciliation has been hampered by a conservative
backlash against the advances of the past thirty years. This is most tellingly
illustrated by the introduction into the debate of the term black armband,
especially in the form the black armband view of history. Those who use the
term claim that in recent times there has been too much emphasis placed on
white dispossession of the indigenous people, and insufficient emphasis on,
and pride in, Australian (read “white Australian”) achievements. Even so, the
history of white-black relations in Australia is now part of the historical
record. A recently published encyclopedia of battles in which Australians
have fought (Coulthard-Clark 1998) lists 306 engagements. Forty-three of
them were between Aborigines and Europeans. Here Anzac Cove, Kokoda,
and Long Tan are listed alongside Richmond Hill (1795), One Tree Hill
(1843), and Battle Mountain (1884), sites of engagements between Aborigines and Europeans. This will be a surprise to those Australians who believe
that Australia is the continent on which no war has ever been fought.
The story of the word Moomba is a sorry tale of the 1950s. Sorry was the
most discussed word of 1997, and in 1998 the first National Sorry Day was
held. Sorry is a word which italicises both the changes that have occurred in
Australian society since the 1950s and the continuing tensions between its
indigenous and non-indigenous people. To many in the non-indigenous Australian community, Sorry Day meant a day of saying sorry for what had
happened in the past. Being sorry in this context meant making an apology,
and also expressing profound regret. In the media reports of those expressions
of sorrow we saw many Australians expressing that apology with tears. In
AbE, sorry is closer in meaning to standard English sorrowful, and is usually
associated with grief and mourning. Sorry business is a ceremony associated
with death, a sorry camp is a mourning camp, and a sorry cut is an incision
signifying ritual mourning. Thus Sorry Day was another Day of Mourning.
Yet the two meanings are not distinct; they overlap, even if not entirely
conterminous, not yet entirely reconciled.
As the frontier is pushed back, as the past is recovered and the present
filled out with living cultures, and as the physical and linguistic maps are
rewritten, AusE is registering a profound change of attitude towards its
indigenous peoples, just as those indigenous peoples are asserting their place
in the lexicon of AusE.
Sydney Morning Herald. 24 Dec. 1996, 5/2.
The Australian. 9 April 1998, 2/3–5.
The origin of the term is explained by Peter Wignell in “The influence of Aboriginal
English”, Australian Style 5: 2 (1997), 1–4. The usage You budju is also provided by
The dates here and elsewhere are based on the citation evidence at the Australian
National Dictionary Centre, Australian National University.
The term native title was widely used in New Zealand from 1847 onwards. It was first
used elsewhere by the British Privy Council in Amodu Tijani v Secretary, Southern
Nigeria (1921) 2 A.C. 399, and probably used there as an alternative to Indian title, which
was used in the United States.
The term was introduced by P. Read The Stolen Generations: The Removal of Aboriginal
Children in NSW 1883 to 1969 (1982. Sydney: NSW Ministry of Aboriginal Affairs). It
was only in the 1990s that it gained wide currency.
Shepparton Aboriginal Arts Council. 8 July 1997. URL
“Rural Industry Strategy”. 24 Feb. 1988. URL ruraltex.
“Bush Tucker Supply Australia: list of many bushfood recipes for chefs”. 10 Sept. 1998.
“Edna’s Table Restaurant … Australian native bush cuisine”. 21 April 1998. URL http://
Mitchell, T.L. 2nd edn 1839. Three Expeditions into the Interior of Eastern Australia.
London: T. & W. Boone, vol. 2, 179–80.
“Uluru-Kata Tjutu National Park”. 31 May 1996. URL apa002/
Origins of Palawa. 1994. Hobart: Tasmanian Aboriginal Centre, np. “It had been suggested for some time that the Tasmanian Aboriginal community find a name to call
ourselves instead of Koori which is used of Victorian and New South Wales Aborigines.
Through … the Tasmanian Aboriginal Languages Project now called the ‘Palawa Karni’
Program, meaning Tassie Blackfella’s talk, research has reinforced that Palawa was
recorded by early historians as meaning ‘native’. Fanny Cochrane Smith was recorded
singing in 1903 and in those recordings Fanny refers to her people as Palawa and through
consultation with the Tasmanian Aboriginal community people growing up on the Bass
Strait Islands remembered being called Palawa”.
Abbott, Ian. 1983. Aboriginal Names for Plant Species in South-Western Australia. Forests
Department of Western Australia.
Archer, Michael. 1975. “Ningaui, a new genus of tiny Dasyurids (Marsupilia) and two new
species, N. timealeyi and N. ridei, from arid Western Australia”. Memoirs of the
Queensland Museum 17: 243.
———. ed. 1982. Carnivorous Marsupials. Vol. 2. Sydney: Royal Zoological Society of
New South Wales.
Arthur, Jillian M. 1996. Aboriginal English: A Cultural Study. Melbourne: Oxford University Press.
Bennett, E.M. 1993. Common and Aboriginal Names of Western Australian Plant Species.
Glen Forest: Wildflower Society of Western Australia, Eastern Hills Branch.
Blake, Barry. 1981. Australian Aboriginal Languages. Sydney: Angus & Robertson.
Braithwaite, Richard W., S.R. Morton, A.S. Burbidge and J.H. Calaby. 1995. Australian
Names for Australian Rodents. Canberra: CSIRO.
Dixon, Robert M.W., William S. Ramson and Mandy Thomas. 1990. Australian Aboriginal
Words in English: Their Origin and Meaning. Melbourne: Oxford University Press.
Cherikoff, Vic and Jennifer Isaacs. 1989. The Bush Food Handbook. Sydney: Ti Tree Press.
Coulthard-Clark, Chris. 1998. Where Australians Fought: The Encyclopaedia of Australia’s
Battles. Sydney: Allen & Unwin.
Daw, Brad, Trevor Walley and Greg Keighery. 1997. Bush Tucker: Plants of the SouthWest. Como, Western Australia: Department of Conservation and Land Management.
Gilbert, Kevin. 1977. Living Black. Melbourne: Allen Lane.
Ginibi, Ruby Langford. 1992. Real Deadly. Sydney: Angus & Robertson.
McGregor, Craig. 1966. Profile of Australia. London: Hodder & Stoughton.
Morgan, Sally. 1987. My Place. Fremantle: Fremantle Arts Centre Press.
Ramson, William S., ed. 1988. The Australian National Dictionary. Melbourne: Oxford
University Press.
Roberts, Ainslie and Charles P. Mountford. 1969. Australian Aboriginal Myths in Paintings. Adelaide: Rigby.
Strahan, Ronald. 1981. A Dictionary of Australian Mammal Names. Sydney: Angus &
———. 1995. The Mammals of Australia. Rev. edn. Sydney: Reed Books.
Strong, Bruce W. 1987. Checklist of Preferred Common Names of Plants of the Northern
Territory. Conservation Commission of the Northern Territory, Technical Report 27.
Australian English — an identity crisis
Susan Butler
The Macquarie Dictionary
In recent years it has become apparent that Australian children take it for
granted that what they speak is called “Australian”. Just as some children
speak French, others speak German, they speak “Australian”. In other words,
they accept as their natural dialect what is identified in formal contexts as
Australian English (henceforth, “AusE”) and in informal contexts is referred
to as “Aussie English” or “Strine”.
AusE is dear to the hearts of those of us who are Australian — we know
each other by the sound of the language we speak, by the special words we
use, by the sense of shared experience and a common history that filters
through it. AusE therefore becomes one of the icons of our culture.
Our understanding has a historical depth, in that it includes words that are
part of our collective consciousness, words which are markers of different
periods of our experience in this place.
So our cultural understanding retrieves for us words from our history
(Table 1). We look back to the convict period (canary ‘a convict’, from the
black and yellow clothing they wore), to the pastoral period (paddock ‘an
enclosed area of land of any dimension’), to the Aboriginal borrowings which
followed the pattern of settlement from Port Phillip to Perth (billabong ‘a
water hole’), to the goldrushes (fossick ‘to cast around in soil for specks of
gold’ and later, ‘to search through other things for an item wanted’). The late
1800s were not only a period in which the bush was seen as a key to our newly
created national identity but also a time of rapid city expansion marked by an
explosion of urban slang (wowser ‘a person who takes a strongly illiberal
attitude to life and morals’). World War I brought with it some army slang
(furphy ‘a rumour, usually unfounded’, supposedly from Furphy & Sons, the
suppliers of water carts around which the soldiers gathered and traded these
unreliable rumours). Then followed the Depression (battler ‘a person who
struggles to make ends meet and to live decently against the odds’). World
Table 1. Lexical items in Australian English
new chum
ticket of leave
cocky farmer
police (those who trap offenders)
a cache of stores or goods (from thieves slang)
a newly arrived convict; later, an immigrant to the colony
a document giving a convict permission to travel within a
certain district
the manager of the station; later, any man
uncultivated land outside city limits
bush covered with light vegetation
someone who claims land by right of occupation
a farmer with a small holding (from the idea that he scratches
for a living in the soil just like the cockatoo)
a partner in getting a living in the bush, especially on the
the station manager; later any man
a crane of northern Australia
a West Australian tree
the well-known marsupial species
a kind of acacia with strong roots and hard wood
the kingfisher-type species of bird
the burrowing animal
the small brightly-coloured bird
mullock heap
a miner
a heap of soil left over from a gold mine operation, possibly
containing small pieces of gold
the tin in which water is boiled and tea is made
an eccentric miner or bushman
a policeman (from the name of the Governor at the time,
Joseph La Trobe)
to keep a claim under close surveillance so that no one else
takes possession in the miner’s absence
1890s TO 1914
bush ballad
back o’beyond
beyond the black stump
a ballad which has a bush setting
a long way away
a long way away
a person who lives off others
a person who collects bottles from houses
a street vendor who sells rabbit meat
a scheme or trick
brass razoo
very good
a friend or mate
a rumour, false report
an insignificant amount of money
an improvised place to sleep on a veranda
sustenance payments (the dole)
emu parade
milk run
Lady Blamey
an assembly for picking up litter
soldier’s uniform for fatigue duty
a routine trip
a beer glass
mentally disturbed
a fool
to relax or rest
War II brought a wave of Americanisms and military slang (troppo ‘insane as
a result of spending time in the tropics’). That trickle of Americanisms has
now grown into a flood.
Recently a nervousness has crept into the linguistic self-image of Australians. They ask, have we kept our hold on our own special language or is it
slipping away from us? Are we being taken over by America — linguistically
as well as economically and culturally? Do we speak AusE any more?
Most are aware that language is always changing, and that on the whole
there is no cause for alarm. However, most also feel more comfortable with
change that has happened over a long period of time. The opening lines of
Chaucer’s poem, for instance, are more likely to intrigue and delight than to
Whan that Aprille with his showres soot
The droht of March has perced to the root.
Change in our own lifetime, on the other hand, is equated with decay and
Lexical change and borrowing
Change in all areas of our lives is reflected in the vocabulary we use:
bootscoot (dancing), indie music (music), blood bin (sport), greenfield (environment), cyberspace (computers), caller ID (communications), Higgs particle (science), biopanic (health), home invasion (crime), lats (exercise), bok
choy (food), dreadlocks (hair), nose ring (fashion) and arcade game (computer games) are all items added to the Macquarie Dictionary between the
second (1991) and third (1997) editions.
The strongest external influence on AusE today is American English
(“AmE”). Compare this set of words which are recent borrowings from AmE:
schmooze, schlep, smick, d’oh, dreck, high five, himbo, push polling, wannabe
and zine with the select few borrowed from British English over the same
period: gobsmacked, mad cow disease and spice girl.
It may seem that AusE has travelled from its roots, but the dialect has seen
this kind of influence before. The word which many think of as quintessentially Australian — the bush — is borrowed from AmE via the goldfields. Its
ultimate origin is in Dutch.
Bush is used in the United States, in Australia, New Zealand and South
Africa. Baker, whose earliest record of the word is dated 1803, traces it
through South African English to Dutch, but there is no evidence for this;
Morris, following the OED [Oxford English Dictionary], describes it as a
‘recent, and probably a direct adoption of the Dutch Bosch, in colonies
originally Dutch’. The word occurs early in the United States, as does
bushranger, and it is probable that both words were borrowed into Australian English from American. Governor Hunter uses the word in 1800 without
comment: ‘…about twenty or thirty Natives thereupon immediately came out
of the Bush and saluted the witness…’ (Ramson 1961: 357–358)
As Ramson remarks, the compound bushranger has an earlier U.S. history:
Bushranger is recorded in American English in 1758: ‘Outside of ordinary
meal-times the following Brethren need butter: the night-watchman, the
herdsmen, the Bush-Ranger sometimes, the threshers, and sometimes the
carpenters.’ Used in the United States with the sense of frontiersman or
woodsman, the word is almost certainly a translation of the Dutch boschloper, woods runner. The intermediate forms bossloper and bushloper
were both in use before 1758. Early Australian quotations suggest that the
word was first used in Australia of men who lived in and were familiar with
the bush and that the use of the word to describe a highwayman or robber was
a later development… (Ramson 1961: 359)
American influence, then, on AusE is not new. What is different is that the
influence of that culture is, at the turn of the century, closer to hand than ever
before. Media such as film and television may transmit American culture to
Australia with a delay of months; the Internet now makes us contemporaneous.
The dramatic reduction in transmission time, and the concomitant improvement in the flow of data for the lexicographer’s files, is revealed in the
immediate circulation via the Internet of the following list. The American
Dialect Society holds an annual competition to choose the Word of the Year.
Nominations for 1997 included:
chopsocky Movie industry slang for the currently-popular Hong Kong
hyper-violent action films that enjoyed an explosion of popularity in the US
and Europe in ‘97 and crossed over into American films such as Rumble in the
Bronx and Supercop.
cradle cam, kiddie cam The increasingly popular Web cameras being installed in daycare centres and grade schools so that parents can monitor their
children from their desktops at home and work. Video cameras that are set up
surreptitiously to monitor nannies and babysitters are called nanny cams.
digipet The generic term for electronic toys that simulate a real-world pet,
requiring constant care and feeding. The most widely known digipet is the
Japanese Tamagocchi (or Tamagotchi).
geeksploitation Taking advantage of twenty-something digital workers,
flushed with pioneer enthusiasm and willing to work long hours if bolstered
by junk food, flexible work schedules, and no dress code. Rolling Stone,
MSNBC’s The Site, and Nightline all did stories about geeksploitation in
handhelds Digital devices such as PDAs (Personal Digital Assistants), electronic organisers, and handheld PCs (PCs that run Windows CE). Any
computers that can be held in one hand. Also, palmtop.
spam The many flavours of spam (unwanted ad-related email and BBS
postings) continued to be popular. Hence, spam wars, the constant back and
forth battles between spammers, spamhaters and legislators.
stalkerazzi Paparazzi who will go to just about any lengths to get the shot
they desire. This term continued to gain in popularity throughout the year,
especially after Lady Di’s death.
World Wide Wait What started out as a humorous interpretation of the
World Wide Web became a popular newspaper headline and marketing
slogan on the problem of increasing lag times on Internet connections.
The winner for 1997 was millennium bug.
Ten years ago in the normal course of events Australian lexicographers
would have waited for years for these words to filter through the media and
make their way into the Australian press before they were considered as
candidates for dictionary entry.
The term couch potato once seemed to set the track record for this kind of
process. It was coined in AmE in the late 1970s as an elaborate joke. The
cartoonist Robert Armstrong represented boob tubers — that is, people who
were addicted to the boob tube or idiot box — as vegetable roots. This led to
the formation of “The Couch Potato Club”. This bit of social irony caught on
in America and was picked up by the magazine New Yorker in mid-1987.
From there it spread to the Sun News-Pictorial in December 1987, a mere six
months gap.
This rapid transmission is now not at all unusual. No sooner, for example,
had Macquarie lexicographers first noticed the toy called “Tamagotchi” and
begun to wonder what the generic name might be for this trade-marked toy,
than the Internet provided digipet as the answer. It still has to be established
that digipet has currency in Australia. Some of the words in the above list may
not be current in AusE — geeksploitation and stalkerazzi, for example. But
once, Australian lexicographers would have lived in ignorance of their fleeting existence in AmE. And the Americanisms that survived would still have
taken ten to fifteen years to filter into AusE.
This is not to say, however, that Australia is flooded by AmE and that
every new coinage in that variety necessarily finds its place in AusE.
If, for example, the Barnhart Dictionary Companion for 1997 is compared with the Third Edition of Macquarie, which appeared at the end of
1997, one would find that a filtering process had selected, from the wealth of
possible offerings in AmE, just those items which had some relevance in
So, for example, AusE takes in beach volleyball, boot camp (in a broader
use than the military), caffe latte, combination therapy, drive-by (in various
compounds), infobahn, human capital, lapware and macarena. On the other
hand, AusE has ignored advanced television ‘digital transmission of television
signals which enables viewers to receive very sharp images or several programs at once’; and the verb Anita-Hill ‘to attack (someone) publicly in order
to discredit their qualifications for a job or appointment, especially during
public hearings’, which is too specific to American politics to have made the
transfer to our situation; and attack-fax ‘a facsimile transmission over telephone lines which aggressively, often unfairly, states the position or activity
of an opponent’.1 This last has not yet been borrowed, possibly because it may
represent a cultural difference between Australians and Americans in terms of
what is considered fair treatment of an opponent. (Its apparent exclusion,
however, may not be permanent, since it would appear to be in the same
category as push polling and spin doctor, both prior US imports into the
Australian political vocabulary.)
There are many lexical innovations attested in Barnhart as having currency in AmE which are totally unknown to Australians. But the average
Australian would be surprised by the notion that there are any checks or barriers
erected which prevent AmE from imposing itself on AusE. The notion that
there is a great deal of AmE that Australians do not know is not one that gets any
kind of public attention. On the contrary, it is assumed that because American
television is so pervasive Australians know everything that can be known about
contemporary AmE, and that Australians are being Americanised out of their
AusE heritage. This, however, is very much an urban preoccupation.
The city and the bush
The distinction between the city and the bush has always been an important
one in Australian social history, and it is a distinction which lexicographers
remain well aware of. On the one hand we must not fall into the trap of
thinking of the bush as isolated and backward. Every country pub has its own
satellite dish; it is possible to be as tuned in to the world at Uluru as in Sydney
or Melbourne, whether by cable TV or by the Internet.
On the other hand, there is indeed a marked distinction between the
language of the city and language of the bush that dates back to the interna-
tionalisation of the Australian economy (that is, to the late 1970s). A good deal
of the anxiety that currently circulates amongst Australians about the viability
of their English in the new millennium is in fact part of a familiar pattern of
time lag between the language of the city and the language of the bush.
While it is still possible in the bush to talk about blokes and sheilas, to use
expressions like bonzer and ridgy-didge, it is not possible to talk like that in
the middle of the Sydney CBD. Every now and then the difference surfaces in
newspaper reports that incorporate bush talk, such as the following:
High and dry in the far west [headline] It’s 11 am and 63 degrees Celsius on
the grass in the Girilambone Public School playground. The daily trickle of
water from Bogan Shire’s Wilga tank has long been exhausted by the 25
households and the school in the parched township, 45 kilometres west of
Nyngan in western NSW. The seven students in the composite class of years
3, 5 and 6 log the temperature as Girilambone wilts under the onslaught of hot
winds from the centre of the continent … “By gingeys it’s hot,” said Tyson
Castley before dashing off for a game of soccer in the midday sun. “These
kids are really tough,” principal Terry Cone, said. “They’re used to it.”
(Sydney Morning Herald, 27 November 1997, p.3)
The expression by gingeys is not one that has currency in urban Australia
It is not, however, the case that the city has abandoned AusE in favour of
AmE. There is too much that is unique to an Australian’s experience of the
world and too much that is now entrenched in the national history for that to
happen. Australians need the words to express their particular culture, which
is not the same as and never will be the same as American culture.
We can turn to the example of the bush again, and note how the word has
travelled within AusE and in particular the way in which its connotations have
In colonial times the bush was threatening and dangerous.
1814 The bush is exceedingly thick and bad travelling on account of the sharp
rocks. (Australian National Dictionary [AND] p.112)
Even when the sense of danger receded, the bush was still an impenetrable
1836 Our road lay through the bush. In India, I should have said the jungle,
and in Europe, the forest. (AND p.112)
The perception was that the bush was good for nothing much until it was
cleared and turned into useful and productive land. But then as early as the
1930s a note of nostalgia crept into discussions of the bush, as a natural
treasure that was being destroyed.
1935 The ‘bush’ is the country in its natural state — as it was before men cut
down the trees and disturbed its flora and ousted the kangaroo, wallaby, and
emu and annihilated the birds that keep our trees healthy. (AND p.112)
This became a continuing theme such that a search through a lexicographer’s
linguistic corpus now almost invariably shows that bush is mentioned in the
context of the need to save plants and animals that are becoming extinct, or as
a small patch that is being steadily diminished by urban development, or even
as something which in itself has almost magical qualities of healing. It has
become a middle-class Australian goal to escape to the bush and be imbued
with the qualities that it somehow transmits just by the magic of being there.
Paddy Pallin, Bushwalking around Sydney, 1959. If you have never walked in
the bush you have never really seen it. If you have not tramped along fernytracks by some quiet stream or scrambled over rocky ridges with the wind
(yes, and maybe the rain) in your hair, you have never enjoyed the bush. If
you have not stood on a rocky eminence looking across a hazy blue valley and
with the eye traced, in anticipation, the route you are to take, you do not know
the bush. If you have not rested at dusk and cooked a meal over a scented fire
of gum-sticks and yarned and sung songs around the fire — if you have not
done these things, you have not entered into your heritage as a true Australian.
(OZCORP [The Macquarie Dictionary’s database of AusE])
In a similar way, mate is not a term which AusE can claim to have
invented but one which Australians have, in a real sense, made their own.
A mate in the Australian context is fundamentally ‘a partner’. Life in the
bush in colonial times was hard and dangerous; it was safer to travel and work
in pairs, and many occupations — goldmining in particular — were simply
impossible to carry out as an individual.
1838 …the work, which a man cannot face alone, requiring always the
assistance of ‘neighbours’, or ‘mates’, or ‘partners’ as they are severally
called, even in the minute details.
1845 Two generally travel together, who are called mates; they are partners
and divide all their earnings. (AND p.389)
The mates in bushlore then went to war and the word acquired an extra patina
of emotion and tradition.
1919 The boy had joined his mates in one of the little cemeteries on the
Western front. (AND p.390)
However, by the late 20th century the word mate had very little denotation despite its still powerful connotation.
1972 A mate in Australia is simply that which a bloke must have around him.
Mates do not necessarily want to know you. (AND p.390)
Sydney Morning Herald, 7 Mar, p.45. 1990 [Nick Whitlam and Malcolm
Turnbull] His last words as Whitlam walked out of the door concluded with
that most ominous of political salutations. It ended in ‘mate’ is all Whitlam
can remember. (OZCORP)
The Macquarie Dictionary has an Internet site called “KiD” (Kids Internet
Dictionary), which invites children to contribute words. It is evident that the
American media have influenced the contributions, but it is noteworthy that
alongside d’oh and nerd and wannabe there is grouse:
1996 grouse! Grouse is a word meaning cool, or something that you might
call excellent. (KiD)
So the ballast of history still appears to give the Australian variety of English
some stability, tossed as it is by the currents of AmE.
Non-lexical aspects
There is more to language, of course, than the words we choose to use. And in
those other elements of language we find a similar complex pattern of accommodation and resistance.
The pronunciation of AusE remains reasonably constant. Despite some
shifts in stress patterns, as sandwiched between AmE and British English
Australians have learned to accept momentárily alongside mómentarily,
cígarette with cigarétte, and mágazine as well as magazíne, the accent continues to be distinctive within World Englishes.
Australian spelling is its own unique blend of American and British
English – Australians will accept colour and color, but not center which they
insist is to be spelt centre. The former example illustrates the common
tolerance in Australia of different spelling practices between reading and
writing. Two thirds of the nation’s newspapers use the color spelling and only
one third use colour, but Australians almost universally write colour.
Australians pitch the border between formal and informal language at a
point that seems relaxed and colloquial to the rest of the world. And the
national style still has some of that traditional black humour and laconic
understatement that typified the humour of the bush.
Australians have experienced and are still experiencing a period of great
change, so it is understandable that some cultural traditions are being reassessed. No doubt they will, as they have done in the past, rework their
attitude to the national variety of English in the light of these experiences, but
AusE — its lexicon, its pronunciation, its nuances — will remain an essential
part of what is seen as quintessentially Australian.
Examples and definitions from the Barnhart Dictionary Companion, 10 (1), Summer
Australian National Dictionary. 1988. Melbourne: Oxford University Press.
Barnhart Dictionary Companion, 10 (1), Summer 1997. Springfield: Merriam-Webster
Macquarie Dictionary. 3rd edition, 1997. Sydney: Macquarie Library Pty Ltd.
Ramson, William S. 1961. “Historical Study of the Australian Vocabulary”. Ph.D., University of Sydney.
Corpus evidence on
Australian style and usage
Pam Peters
Macquarie University
Recognition of Australian English (henceforth, “AusE”) as a legitimate and
independent variety is only a few decades old. World War II was a watershed
for pioneering work on the Australian lexicon by S.J. Baker (1946), and on the
Australian accent by A.G. Mitchell and A. Delbridge (1966). But until then the
English of Australia was widely regarded — both Down Under and Up On Top
— as an outpost of British English (“BrE”). This view dies hard. The “Story of
English”, compiled by McCrum, Cran and MacNeil for the BBC in 1985,
presented AusE under the heading “Echoes of an English voice”, not as if one
might expect to find a fresh voice in it, or a distinctive style. Elsewhere, such
Australianisms as were recognised were of the extravagant “stone the crows”
type. They could be used for humour on the Goon Show, but they hardly
suggested that there was ample subtler material to be found and described.
Pioneer corpus work on English usage
The means to systematically describe English usage came into being only
slowly in the 1960s and 1970s, with the development of the first reference
corpora in USA and Britain. The American Brown corpus was assembled
1961–5, in the shadow of Chomskyan approaches to language which downplayed the value of empirical evidence. British grammarians however recognised its importance. Quirk had already begun the Survey of English Usage
(primarily spoken) at University College London; and Leech subsequently
undertook to compile an exact British counterpart to the Brown corpus.
Difficulties in funding the latter meant slow progress and it was eventually
completed off-shore thanks to Norwegian resources and energy, hence its
acronym LOB, standing for Lancaster-Oslo/Bergen.
While the LOB and Brown corpora were intended to support comparative
and linguistic studies of standard English, the COBUILD corpus developed at
the University of Birmingham from the late 70s on was aimed more specifically at BrE lexicography. The Birmingham corpus was considerably larger
than Brown and LOB, starting with seven million words, and growing to 27
million (before being superseded by the so-called Bank of English of the
1990s which contains 200 million). In spite of its greater magnitude, even the
COBUILD corpus contained less than the lexicographic researcher would
wish in terms of common English expressions and idioms, let alone more
technical vocabulary. The hope of supporting every word and definition in the
dictionary headword list with corpus evidence could not be realised (Cobuild
English Language Dictionary 1987: xv)
Corpus developments in Australia
The 1980s also saw the development of computer corpora in Australia. From
the start the objectives were comparative, to examine parallels and divergences between AusE and the northern hemisphere varieties, and so the
corpora constructed matched those in Britain and USA in their size and
constituency. An early venture of this kind was the Melbourne-Surrey Corpus
of newspaper editorials (from the Melbourne Age 1980–81). It was designed
to match a small (100 000 word) British corpus of editorials from the Times,
which had been used to examine the prevalence of notional rather than formal
agreement of a verb with “corporate” nouns like committee (Nixon 1972). BrE
speakers are often said to be readier to use notional agreement than Americans, and the Times editorials used it much more freely than those from The
Age (Ahmad and Corbett 1987: 39). However recent longitudinal analysis of
Times data (Bauer 1984: 63,66) has shown sharply changing profiles on this
in the Times during the 20th century, probably reflecting the penchant of
individual Chief Editors. Hundt (1996: 74–5) was able to show that the
patterns of agreement with corporate nouns vary in different parts of the same
newspaper. The conclusions drawn from small and specialised databases are
necessarily narrow in scope, and provided some of the motivation for compiling a broadly based Australian corpus similar to Brown and LOB.
Begun in 1986, the Australian Corpus of English (“ACE”) was compiled
at Macquarie University by Peters, Collins and Blair. As with LOB in Britain,
the compilation of ACE was rather protracted due to the dearth of research
funds for the building of research databases. ACE was intended to match
Brown and LOB as closely as possible in its structure, so that frequencies
could be directly compared from one to another, against the same generic mix
of data. In the event, some particular genres, especially those of fiction had to
be rethought in terms of what was available in Australia (Green and Peters
1991: 45–9). The shortfall in publications in categories N (Adventure and
Western) and P (Romance) were made up with the addition of two new
categories (S: Historical fiction and W: Women’s fiction): but the overall
balance of fiction and nonfiction of Brown and LOB (1:3) was maintained.
The most important difference between ACE and Brown/LOB is in the
date of the materials: in the former, all from 1986. This makes it 25 years
younger than its northern hemisphere models, and so intercomparisons may
involve either historical or regional parameters, or both (cf. Holmes
1994: 28). It was unavoidable, given that many of the Brown/LOB genres
were not published in Australia in 1961. At the same time it was important to
work with up-to-date publications, in order to create a contemporary Australian reference corpus. Yet because ACE data is at a remove from Brown and
LOB both temporally and regionally, some forms of “triangulation” (reference to other corpora or empirical data locatable in time or space) are needed
to distinguish regional from temporal differences.
The comparability and contemporaneity of computer corpora was a prime
concern in the design of the International Corpus of English (“ICE”), a
cooperative venture in corpus-building in more than a dozen English-speaking
countries where the language functions as either a first or a second language.
Each corpus consists of 1 million words of spoken as well as written language
(approximately 50/50) from a predetermined set of genres, and samples were
collected between 1990 and 1994. The project was led by Greenbaum at
University College London, and the British ICE corpus ICE-GB provided
material for his Oxford English Grammar, published only a few months
before his death in 1996. An Australian ICE corpus has been compiled by
Peters at Macquarie University, and selected spoken samples from it and
several other ICE corpora are to be published on CD-ROM in 1999. Other
fruits of this collaborative corpus venture are to be found in intervarietal
comparisons of English usage in World Englishes 15, and some of those
findings are discussed below.
One other Australian corpus development contributes very strongly to the
finer description of AusE: the aptly named Ozcorp, developed by the publisher of the Macquarie Dictionary since 1990. Like the Cobuild corpus, its
specific purpose is to support the dictionary headword list rather than provide
comparisons with any other database, and it collects a range of publications to
illustrate both current and older Australian usage. In 1998, Ozcorp consisted
of around 24 million words, including 12.5 million fiction (ranging in date-ofpublication from 1867 on); and approximately 11.5 million nonfiction, including newsprint. The third edition of the Dictionary, published in 1997, included
for the first time more than 3180 citations drawn from the corpus to exemplify
particular senses of words, and to show something of the range of authors
whose writing contributes to AusE at large. The less visible role of the corpus
is to supply frequency information on variant forms of head words, in order to
decide their order of precedence in an entry or the weighting of alternatives in
usage notes. Subtle aspects of a word’s connotation also emerge through
Ozcorp data (see Butler, this volume). Exponential growth in computer capacity and in the intricacies of software have facilitated the storage, search and
retrieval of data from large corpora such as Ozcorp. They provide insights into
lower frequency items from open word classes, whereas the smaller parallel
corpora such as ACE offer ample data on higher frequency items associated
with the closed classes of the language. The relativities of usage can be
illuminated by comparisons of (a) complementary items within the same
database, or (b) the same item in parallel databases. Either kind of comparison
helps to show what is marked or unmarked in a given variety of usage: for
example, maybe/perhaps. The ratio between these in ACE, LOB and Brown
shows that the occurrence of maybe is quite marked for British against the
prevailing use of perhaps. In the other Englishes the relativities are reversed.
The regional and stylistic status of conjunctive like
The regional contrasts afforded by data from comparative corpora can be
further refined by reference to generic variation, wherever the frequencies in,
say, fiction and nonfiction and other subgenres of prose can be compared. The
divergent status of the conjunction like in British and American usage can be
demonstrated roughly via their different frequencies (19 to 1012, significant at
the .001 level of probability). In fact, conjunctive like appears only in nonfiction in LOB, whereas it is used in both nonfiction and fiction in the Brown
corpus (Peters 1994: 249) and in ACE (Peters and Delbridge 1997: 310). Its
broader pattern of distribution in the Brown and ACE databases makes it a
relatively unmarked and stylistically neutral item in the newer Englishes. But
in BrE it is stylistically marked and strongly associated with the more freewheeling styles of fiction, where one is not constrained by notions of what is
“standard”. Against Fowler’s defence of the use of conjunctive like (1966:
334–6), this is the more remarkable. Corpus evidence on conjunctive like
shows that in BrE it is still confined to the fiction genres (less formal registers), whereas in American English (“AmE”) and AusE it appears in the
standard register as well. One might therefore say that its stylistic status has
been recalibrated in the newer Englishes. The increased polysemy of conjunctive like is a further sign of its productive role within the standard in USA and
Australia (Peters 1994: 447–8).
Regional divergences in the maintenance of the subjunctive
Stylistic recalibration of the subjunctive can also be seen through its contrasting frequencies in British, American and Australian corpora. The mandative
subjunctive (for example, “demand that he come…”) has shown a strong
decline in 20th century BrE, correlating with grammarians’ comments that it
has become restricted to formal style (Quirk et al. 1985: 1012–3), and with its
low level of occurrence in LOB. However it maintains its strength in AmE and
AusE according to the evidence of Brown and ACE (Peters 1998). In Australia the mandative participates freely in both standard and informal writing,
witness the following from category A (Newspaper reporting) and category W
(Women’s Fiction) in ACE:
A31 6691
W08 1483
…received a pink slip demanding that she pay a fine.
I hated the cold weather which demanded that I be home early
Corpus evidence also shows the widening gulf between usage of the mandative and that of the were subjunctive in AusE. Its use in hypothetical and
conditional clauses in ACE is substantially lower than in either Brown or
LOB, and correlates with the popular observation that “the subjunctive is
dying”. Increasingly the were subjunctive represents a formal choice, and
removes itself from the standard range of style.
The dynamics of style, by which expressions may contract to the formal
register or expand from a base in the informal register, are clearly regionally
determined. Evidence on the subjunctive shows a three-way difference: AmE
maintains both mandative and the were constructions, while current Australian usage confirms the former, and BrE the latter, according to the parallel
corpora. Further evidence on the disuse of the mandative in BrE (and support
for the proposition that this is a regional difference, rather than one due to the
lapse of time between ACE, LOB and Brown) comes from Hundt (1996: 78–
9), who compared the levels of mandatives and should paraphrases in 1990s
data from the Guardian and the Miami Herald. However, counterevidence of
a subjunctive revival was found by Overgaard (1995) in her collection from a
variety of British sources at spaced intervals from 1900 to 1990. Further
research is therefore needed to see whether the mandative subjunctive is
reaffirming its place in standard BrE, perhaps under American influence. It
would provide evidence of the way in which the calibration of style in one
variety can impact on another.
Variation in the use of contractions
Let us now turn to an aspect of usage on which AusE proves to be quite
distinct from both American and BrE: its use of contractions in written
language. Contractions are, in Labov’s terms (1978: 237, 314), one of the
indicators if not markers of informality, being the regular targets of editorial
intervention in both academic and bureaucratic documents. This pattern of
distribution is visible in the relative appearances of it’s and don’t in the
nonfiction genres of the three parallel corpora. (See Table 1.)
In all three corpora, the usage of it’s and don’t is at its lowest in categories
H and J, confirming they are often felt to be inappropriate for the more formal
kinds of prose. They are notably absent also from category D (Religion) in
Brown, presumably because they are thought incongruous with the solemnity
of the subject matter. This detail correlates interestingly with the finding that
the American Brown data shows their commitment to the “substance of
religion” (in much more frequent references to Christ, God, Jesus, i.e. to
Table 1. Variable occurrence of it’s and don’t in the nonfiction genres of ACE, LOB and
Brown , showing the rate of occurrence per sample
Press: reportage
Press: editorial
Press: reviews
Popular lore
Belles lettres
entities that command reverence); whereas the greater frequency in LOB of
terms like bishop, parish, vicar tended rather to project the institutional
manifestation of religion (Leech and Fallon 1992: 42). Table 1 also shows
that Australian usage of those contractions is almost always higher than in the
other two earlier corpora. The trend, and its minor vagaries, called for a more
comprehensive study of contractions, of those like it’s which contract the
primary auxiliary or modal, and those like don’t which contract the negative.
Table 2 sets out the total token frequencies of types belonging to both
groups. The highest values in each set are in bold type.
The overall frequencies for contractions in the Australian corpus are
often (in 24/33 cases) higher than those of the other two. This is most evident
in the contractions of the present tenses of be and have, where if not first, ACE
comes a close second to LOB. The same can be said of the set contracted with
‘d, whether this represents “had” or “would”. In all three sets so far, it is the
contractions involving the second and first person singular on which ACE
frequencies are lower than LOB, a point to be further investigated. Yet it may
seem surprising that the British use of contractions is as high as it is (when the
stereotypical view is that BrE prefers the formal style). Both this, and the
equally surprising fact that the American data so often yields the lowest
scores, may be related to differences in their sampling of fiction (see below).
A further linguistic factor would correlate with Brown’s relatively low frequencies for the contractions in the left-hand column: the lower frequencies
overall in Brown of the verbs be and have (Hofland and Johansson 1982: 36).
This curious fact has not passed unnoticed, and some explanation for the
Table 2. Frequencies of contractions in ACE, LOB and Brown
(Note that the contracted ’s and ’d are ambiguous out of context. The ’s can represent either
“is” or “has”, and ’d either “had” or “would”.)
lowered uses of have and its contractions may be found in the observations
that AmE makes less use of the present perfect (Quirk et al. 1985: 193, 540,
581, 1015–8). Görlach (1987: 53–4) recognises it as one of the syntactic
divergences between AmE and BrE, though not exactly “colonial lag”. Yet
according to Strang (1970: 149–151, 190), the elaboration of the perfect (and
continuous) aspect is a relatively recent feature of English grammar, as is the
full paradigm for passive constructions. Both were developing their paradigms between 16th and 19th centuries, during the very centuries when AmE
was developing independently from British. Unfortunately for this inquiry,
the most detailed studies of the passive (e.g. Granger 1983) make generic
rather than regional comparisons to demonstrate the variability of the construction. Yet if, as seems possible, AmE at large is less given to perfect and
passive constructions, it would help to account for Brown’s lower frequencies
of the verbs be and have, as well as their contractions. The low informational
value of have and be in compound verbs makes them susceptible to weak
stress and contraction in speech, and the contracted forms are more likely to be
transferred into prose by writers (British, Australian) whose variety of English
makes extensive use of them.
Among the set of negative contractions, the ACE frequencies are again
usually (in 10 out of 15 instances) the highest of the three corpora. Here the
most remarkable exceptions are the Brown figures for hadn’t and wouldn’t
(discounting the very small values and differences for hasn’t). The relatively
high scores of Brown for hadn’t and wouldn’t make interesting comparison
with its prevailingly low scores for the contraction ’d (corresponding to either
had or would, as noted above). These would seem to be complementary
phenomena, and provide evidence for the fact that AmE, more strongly than
BrE, seems to prefer to contract the negative rather than the auxiliary/modal,
where the language allows either. Thus he hadn’t done it rather than he’d not
done it. Quirk et al. (1985: 123) note the tendency of Scottish and northern
English speakers to prefer the latter. In the recent Frown corpus (analogous to
Brown but from 1991), Hundt (1996: 44) found an increase over Brown in the
use of all three negative contractions of the verb have; and she reports from
Krug (1994) an increase in all negative contractions, both of which suggest the
consolidation of this mode of contraction in the USA. It is also consistent with
Tottie’s larger findings on negation, that is there is increasing use of the
speech-style not forms of negation, at the expense of those using no (Tottie
1991). They are part of the larger “envelope of variability” (Holmes 1994: 30).
The variable frequencies of the contractions under discussion are influenced not only by the behaviour of the auxiliaries and patterns of negation, but
also by the roles of the pronouns involved. Contractions formed with the first
and second person pronouns and associated with direct address are naturally
less widely distributed in a corpus of prose than those such as it’s with its
various expository roles. The craft of fiction gives scope for contractions of all
kinds as a means of lending authenticity to speech and spoken narrative, and
the relaxation of stylistic constraints on the use of contractions is a convention
of fiction in all varieties of English. Style and grammar checkers associated
with word processing regularly relax the formal rules for those who create
fiction on the computer. The fiction categories of LOB harboured much higher
levels of contraction (of verbs BE, HAVE, will and would, as well as NOT)
than the nonfiction (see Kjellmer 1998). But the sense of what is “standard” in
each variety is far from constant, if usage books in Australia, UK and USA are
anything to go by (Peters and Young 1997); in fact the range of published
opinions on acceptability proved wider in America than Britain. So the
generic distribution of contractions in the three corpora called for closer
Table 3 confirms that there are substantial differences in the distribution
of these contractions, which underpin the divergent totals discussed earlier.
Reading vertically, you would note that in all three corpora the expository
contraction it’s is to be found in much greater numbers and relative frequency
in nonfiction than the person-specific contractions. It’s (whether as it is or it
has) works both cohesively and as an expository device (= prop it) in cleft
constructions and quasi-existential (“atmospheric”) statements. Still, (reading
horizontally) its relative frequency in nonfiction is much higher for ACE than
for either LOB or Brown. The same kind of difference holds for there’s
(usually is, rarely has) which is also widely used for expository purposes in
existential statements. Its occasional role as a deictic presentational device, as
in There’s gratitude for you, is more at home in the context of speech than
writing (Collins 1999: 80), at least in this contracted form, and therefore less
likely to boost its occurrence in the nonfictional genres of the corpus. But the
naturalisation of existential there’s in Australian nonfictional prose is also
evidenced by the larger number of examples in which it was found coupled
with plural items, as in “there’s …” In ACE 5/17 examples of this occurred in
the nonfiction categories, compared with 0/5 in LOB and 2/8 in Brown. In
almost all cases the complementary items were quantitative: there’s lots of;
Table 3. Frequencies of selected contractions in fiction and nonfiction in the three corpora, showing the rate of occurrence per sample
nonfiction fiction
rate per sample
rate per sample
rate per sample
rate per sample
rate per sample
rate per sample
rate per sample
total no. of
there’s plenty of; though the Australian fictional examples also presented
more allusive estimates of quantity, as in there’s the kids’ bedrooms. As
elsewhere in English grammar, quantitative statements seem to neutralise the
propensity for formal agreement in terms of number (Reid 1991: 277–280).
The distribution of won’t and wouldn’t, also non-person-specific, also
extends well into nonfictional writing where ACE is concerned. Yet the rates
for wouldn’t show it to be much less acceptable in nonfiction for British and
American writers, and much more at home in fiction. Again there is a marked
differential between their use in the Australian data and that from the British
and American corpora. Within standard Australian prose both existential and
negative contractions can appear and are stylistically more neutral than in
comparable British and American writing. The acceptability rates of the
person-specific contractions (I’m, you’re, we’re) in nonfiction is also notably
higher in ACE than LOB or Brown. The differential is greatest for we’re,
which can of course be used for solidarity with the reader in various kinds of
persuasive and procedural writing, apart from its role in scripted conversation.
The rates of I’m and you’re in fiction samples, as shown in Table 3, are
also of interest. In both cases the highest rate is to be found in the British data,
and the implication might be that British fiction somehow fosters larger
amounts of direct speech and interactive writing than its counterpart in Australia and USA. Greater amounts of interactive writing are also suggested by the
proportions of different kinds of sentence-final punctuation shown in the
corpus. A mini-study of 36 samples from the fiction and nonfiction in each
corpus produced the following results:
Table 4. Frequencies of sentence-final punctuation marks in a subset of samples from
fiction and nonfiction in the three corpora
fiction nonfiction
fiction nonfiction
24 samples 12 samples 24 samples 12 samples 24 samples 12 samples
Full stop (.)
Q. mark (?)
Excl. mark (!)
Total of
Data from LOB again registers the highest number of both question and
exclamation marks, most strikingly the latter, which appear more than three
times as often as in either ACE or Brown. This again suggests more extensive
use of interactive syntax in the forms of interrogation, imperatives and exclamation, whether or not there are one-to-one correspondences between the
grammatical function and the choice of punctuation mark (Meyer 1987: 21).
But the distributions of punctuation also correlate with the numbers of sentences in each subset of samples (and of course their length). The data
therefore indicates that the American fiction data typically presents longer,
narrative-style sentences than either of the other corpora, though the contrast
with LOB is the most striking. The difference can be explained by differences
in the sampling procedures for fiction in the two corpora. The Brown Manual
(1964; rev, 1979: 1) makes it clear that fiction involving more than 50%
dialogue was avoided, whereas no comparable policy is mentioned in the LOB
Manual (Johanssen 1978). This would certainly explain the surprisingly high
levels of contractions found in LOB fiction, and the stark contrast with their
frequencies in nonfiction. Their acceptability in contemporary British writing
is still an open question, despite Westergren-Axelsson’s findings (1998) that
contractions are a good deal commoner in British newspapers of the 1990s
than those of the 1960s. The democratisation of British press style does not
necessarily reflect a more general shift in the norms of British prose.
Meanwhile the wider distribution of contractions in the Australian ACE
database, and their higher rates of occurrence in many kinds of nonfiction
(Table 1), show that they are not so constrained by the notion of being
informal, and are not excluded from the standard forms of Australian writing.
Other congruent findings on Australian usage
These findings on the acceptance of contractions within standard Australian
prose correlate well with those from a variety of other corpus-based studies.
The use of get in passive constructions has been shown by Collins (1996: 54–
5) to occur more frequently in Australian than in British or American data.
Australian use of the modal can (Collins 1988: 285) is significantly higher
than that of may and might (though these have often been presented as its
“politer”, and more formal equivalents for certain functions). Complementing
this higher use of informal features in standard AusE are the lower levels of
use of more formal features. Corpus research demonstrates that shall is quite
rare in comparison with will (Hundt 1998: 50); fewer makes little showing in
comparison with less (Peters 1996: 60); and AusE prefers not to use the
possessive gerund-participle: “…surprised at its coming so soon”, but rather
“…surprised at it coming so soon” (Peters and Delbridge 1997: 309). All
these corpus-based studies converge to show how items which are elsewhere
excluded from standard prose have been recalibrated within Australian style.
In his elicitation study of divided and debatable usage, Collins (1989: 148–9)
found that young adult Australians consistently chose variants which were
less linguistically conservative, and this too provides triangulation on the
Australians’ accommodation of contractions and other informal devices
within standard prose correlates with their often negative orientation to formality, which is devalued in favour of styles of behaviour which are obviously
egalitarian and inclusive. Collocations such as “dreadful elitism” (Melbourne
Age 16 September 1998) express this attitude, and the often negative connotations of elite/elitism/elitist are demonstrated by Delbridge (this volume) on a
broad base of evidence from Ozcorp. The negative semantic development for
elitism/elitist is not restricted to AusE, yet it is certainly less established in
other varieties. A search of the British National Corpus captured the negative
applications in less than 20% of instances (cf. Delbridge’s 40%.) Most significantly the negative meaning has yet to be recognised in British (or American)
dictionaries, whereas it has been built into the most recent edition of the
Macquarie Dictionary (1997). In Australia it is the key to an important
stylistic value.
Semantic evidence from Ozcorp thus coincides with the results of numerous quantitative studies based on ACE and other corpora, to demonstrate the
recalibration within standard usage of what are elsewhere items of informal
style. Many Australians prefer styles of communication which work directly
(more like face-to-face encounter), and distance themselves from verbal behaviour which seems unnecessarily formal or stuffy. In all this we arguably
have a fresh voice, rather than “echoes of an English voice” — one which
speaks through the wide range of writing in Australian corpora. The Australian voice provides a counterpoint in relation to movements in English prose
elsewhere, its current written style tending to embrace the spoken style rather
than distance itself from it — as have writers in various schools of English
prose over the last five centuries (Gordon 1966). The rapprochement between
writing and speech is a force in the continual vitalisation of written style.
Ahmad, Kurshid and Greville Corbett. 1987. “The Melbourne-Surrey corpus”. ICAME
Journal 11: 39–43.
Baker, Sidney J. 1946 (repr. 1966; 1978). The Australian Language. Sydney: Currawong
Bauer, Laurie. 1994 Watching English Change. London: Longman.
Cobuild English Language Dictionary. 1987. London and Glasgow: Collins.
Collins, Peter. 1988. “The semantics of some modals in contemporary Australian English”.
Australian Journal of Linguistics 8: 261–286.
Collins, Peter. 1989. “Divided and debatable usage”. In Peter Collins and David Blair, eds.
Australian English: the Language of a New Society. Brisbane: University of Queensland
Press, 138–149.
Collins, Peter. 1996. “Get-passives in English”. World Englishes 15:43–56.
Collins, Peter. 1999. “The deictic presentation construction in English”. In Peter Collins
and David Lee, eds. The Clause in English: in Honour of Rodney Huddleston. Amsterdam: Benjamins.
Fowler, Henry W. 1926; 2nd ed., 1966. A Dictionary of Modern English Usage. Oxford,
Francis, W. Nelson and Henry Kucera. 1964; rev. 1977, 1979. Manual of Information to
Accompany a Standard Corpus of Present-day Edited American English. Providence,
Rhode Island: Brown University.
Görlach, Manfred. 1987. “Colonial lag? The alleged conservative character of American
English and other ‘colonial’ varieties”. English World-Wide 8: 41–60.
Gordon, Ian A. 1966. The Movement of English Prose. London: Longman.
Granger, Sylviane. 1983. The be + past participle Construction in Spoken English. Amsterdam: North Holland.
Green, Elizabeth and Pam Peters. 1991. “The Australian corpus project and Australian
English”. ICAME Journal 15: 37–53.
Greenbaum, Sidney. 1996. The Oxford English Grammar. Oxford: Oxford University
Hofland, Knut and Stig Johansson. 1982. Word Frequencies in British and American
English. Bergen: Norwegian Computing Centre for the Humanities.
Holmes, Janet. 1994. “Inferring language change from computer corpora; some methodological problems”. ICAME Journal 18: 27–40.
Hundt, M. 1996. “New Zealand English and its relation to British and American English”.
Inaugural Dissertation zür Erlangung der Doktorwurde, Albert-Ludwigs-Universität zu
Freiburg im Breisgau.
———. 1998. New Zealand English Grammar: Fact or Fiction? A Corpus-based Study in
Morphosyntactic Variation. Amsterdam: Benjamins.
Johansson, Stig, with Geoffrey Leech and Helen Goodluck. 1978. Manual of Information to
Accompany the Lancaster-Oslo/Bergen Corpus of British English. Oslo: University of
Oslo, Department of English.
Kjellmer, Göran. 1998. “On contraction in modern English”. Studia Neophilologica 69:
Labov, William. 1978. Sociolinguistic Patterns. Oxford: Basil Blackwell.
Leech, Geoffrey and Roger Fallon. 1992 “Computer corpora — What do they tell us about
culture?” ICAME Journal 16: 29–50.
Macquarie Dictionary. 1981; 2nd ed. 1991; 3rd ed. 1997. Sydney: Macquarie Library.
McCrum, Robert, William Cran and Robert MacNeil. 1986. The Story of English. London:
Faber and Faber.
Meyer, C. 1987. A Linguistic Study of American Punctuation. New York: Peter Lang.
Mitchell, Alexander G. and Arthur Delbridge. 1966. The Speech of Australian Adolescents.
Sydney: Angus and Robertson.
Nixon, Graham. 1972. “Corporate-concord phenomena in English”. Studia Neophilologica
94: 120–126.
Overgaard, Gerd. 1996. The Mandative Subjunctive in American and British English in the
20th century. (Studia Uppsaliensia 94). Uppsala: Acta Universitatis Uppsaliensis.
Peters, Pam. 1987. “Towards a corpus of Australian English”. ICAME Journal 11: 27–38.
Peters, Pam. 1992. “Corpus evidence on some points of usage”. In J. Aarts et al., eds.
English Language Corpora: Design, Analysis and Exploitation. Amsterdam: Rodopi,
Peters, Pam. 1994. Cambridge Australian English Style Guide. Melbourne: Cambridge
University Press.
Peters, Pam. 1996. “Comparative insights into comparison”. World Englishes 15: 57–68.
Peters, Pam. 1998. “The survival of the subjunctive in Australia”. English World-Wide 19:
Peters, Pam and Arthur Delbridge. 1997. “Fowler’s legacy”. In Edgar Schneider, ed.
Englishes Around the World vol. 2: Carribbean, Africa, Asia, Australasia. Amsterdam:
Benjamins, 301–318.
Peters, Pam and Wendy Young. 1997. “English grammar and the lexicography of usage”.
Journal of English Linguistics 25: 315–331.
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman.
Reid, Wallis H. 1991. Verb and Noun Number in English: a Functional Explanation.
London: Longman.
Strang, Barbara H. 1970. A History of English London: Methuen.
Tottie, Gunnel. 1991. Negation in English Speech and Writing: a Study in Variation.
London: Academic Press.
Westergren-Axelsson, Margareta. 1998. Contraction in British Newspapers in the Late 20th
Century. (Studia Anglistica Uppsaliensia 102). Uppsala: Acta Universitatis Uppsaliensis.
Section B
English in Australia: variation
Torres Strait English1
Anna Shnukal
Aboriginal and Torres Strait Islander Studies Unit,
The University of Queensland
Despite the publication of a number of academic and popular accounts of
Aboriginal English varieties (see Malcolm, this volume), no comparable
studies have been made of the English spoken by the second group of
indigenous Australians, the Torres Strait Islanders.
The 45,000 self-identified Torres Strait Islanders are a minority within a
minority, comprising only 0.2% of the total Australian population (1996
census figures). Originally Melanesian peoples from the southern coast of
Papua New Guinea, they have long intermarried with Aboriginal Australians,
Pacific Islanders, Malays, Filipinos and Europeans to create a unique people
and culture. The outer islands have been losing population since the end of
World War II: home to the majority of Torres Strait-dwelling Islanders is now
Thursday Island, the administrative, political and educational centre of the
Strait, but over three-quarters of the Islander population live in coastal cities
and towns in mainland Queensland, where their children speak English as
their primary language.
Traditionally, Torres Strait Islanders spoke two unrelated languages,
which symbolised the major ethnological and geomorphological division
between east and west. Kala Lagaw Ya, structurally a Pama-Nyungan Aboriginal language, was spoken in the western and central islands; Meriam Mir,
a Papuan language, in the east. Towards the turn of the century these were
joined by a locally-creolised form of Pacific Pidgin English, the language of
the nineteenth century fisheries, now officially known as Yumpla Tok or
Torres Strait Creole (Shnukal 1988). Though analytically distinct from Eng-
lish, the creole was until recently believed to be English by Islanders and
Europeans alike. Formal instruction in English was introduced to the islands
by European schoolteachers, as the Islanders were drawn increasingly under
Queensland administrative control, but, despite an education system whose
language of instruction was nominally English, English was until the 1980s
spoken almost exclusively by Europeans.
Most younger Islanders are bilingual in the creole and English, older male
Islanders are often trilingual in their island language, the creole and English,
and a few high-status older men, usually ex-teachers, priests, storekeepers or
politicians, are quadrilingual. All four languages are functionally distinct,
although the creole and English, which for years were balanced in a classically
diglossic pattern, are tending to overlap in more domains than previously. The
traditional languages are emblematic of older custom and home island group;
the creole is the Islander lingua franca and the “low” variety when compared
with English; English is still perceived as the language of Europeans, the
“high” variety of administration, mass media, religion, the courts and education. It is almost exclusively the language of writing.
Nevertheless, Islanders today have a far greater active and passive knowledge of English than even ten years ago and are exposed to a far wider range of
oral language varieties than in mainstream communities, including different
dialects of the traditional languages, different dialects of the creole, Australian
English (henceforth “AusE”) in both standard and non-standard forms, Torres
Strait English (TSE), Aboriginal English and, through television and videos,
varieties of American and British English. Although in fact Islanders regularly, though largely unconsciously, cope with an extraordinarily rich language mix, they are rarely explicitly made aware of all the different language
varieties around them or taught the systematic differences among them. As it
happens, Islanders are in general highly successful language learners and have
a lively interest in language that manifests itself in puns (often multilingual)
and various kinds of wordplay. They have made an art form of stratagems
which disguise their lack of facility in English.
Torres Strait English
Despite mutual influence and some convergence, Torres Strait English (TSE)
is not the same as Torres Strait Creole (TSC). The two are phonologically,
grammatically and semantically distinct, despite the fact that the lexicon of
TSC is based primarily on English, that it continues to borrow from English
and that speakers of both languages are themselves confused as to the boundary between them. TSE is a variety, or more accurately a spectrum of varieties
of AusE; it poses a number of analytical problems which are similar to those of
Aboriginal English.
Who, for example, constitute the community of speakers of TSE? What
are the community norms? Is TSE the first language of a group of speakers?
To what degree should TSE be considered an interlanguage, as opposed to a
normative variety of AusE? If TSE is recognised as a variety of AusE, should
it be classified as an ethnolect, a non-standard ethnic variety of AusE, spoken
mainly by both homeland and mainland Torres Strait Islanders? Or is it rather
a regional non-standard variety of AusE, centred geographically on Thursday
Island? Since TSE has been and continues to be influenced phonologically,
morphologically, syntactically and lexically by the creole and by the two
traditional Torres Strait languages, where do its boundaries lie, especially the
boundary between TSE and TSC, in both of which can be heard, as an
Aboriginal colleague once put it, “the shadow of the traditional languages”? Is
the TSC east/west dialect difference — itself a continuation of the traditional
language differences — manifested in TSE or is it elided? What should one
make of the hybrid TSE/TSC varieties spoken by the children at Thursday
Island primary and high schools? What are the implications for English
literacy for these children?
Accurate answers require a formal sociolinguistic analysis, not attempted
in this preliminary descriptive account, and any analysis must take into
account considerable variation, both linguistic and social. Almost all speakers
of TSE are Torres Strait Islanders, though not all Torres Strait Islanders are
speakers of TSE. European children and adolescents brought up on Thursday
Island speak it as well as Standard AusE (SAusE), which I take to be a
linguistic abstraction based mainly on written English. Many Torres Strait
Islanders do not speak TSE, or speak it only among friends and relatives. For
older Islanders now resident on the mainland, whose first language is usually a
traditional language or the creole, TSE is clearly a second language.
2.1. Preliminary notes
There exists no scholarly description of TSE. My intention here is to provide an
impressionistic preliminary sketch, comparing features of TSE as they differ
from SAusE and to offer some general observations. I have not formally
studied TSE: the following observations are based on many years of listening,
speaking and recording in homeland and mainland Torres Strait communities
and their schools, but particularly on a 1996 analysis of the written English of
Thursday Island State High School students from Years 8 to 11, which provides
the bulk of the examples. The students’ grammar and spelling are unchanged.
Other examples come from written work from Islander Grade 5 and 6 students
at the Black Community School in Townsville in 1981; sermons, community
church, store and canteen noticeboards; administrative and personal correspondence; and letters and notices in the local newspaper; that is, towards the more
formal end of the spectrum.2 While the processes of simplification, regularisation and transfer found in TSE are typical of interlanguage and some of the
examples must have been produced by interlanguage speakers, I have chosen
only examples which illustrate apparently systematised features, shared by
many, if not all, speakers.
TSE is characterised by a number of variable phonological, grammatical,
and lexical features, some of which also occur in other non-standard AusE
varieties, including Aboriginal Englishes, others of which are unique to it.
Islanders, by virtue of their position as an indigenous minority in Australian
society, are more often exposed to non-standard (and therefore stigmatised)
AusE grammatical forms, such as past tense done, give and seen; to nonstandard agreement you was/they was; to the non-standard plural demonstrative article them boys rather than those boys; and these non-standard forms
tend towards categorical status in TSE.
2.2. Phonetics and phonology
There is considerable phonetic (and possibly some phonemic) variability
among speakers, depending mainly on first language, length and intensity of
exposure to SAusE and TSC, and preparedness to learn English. Few TSE
speakers employ the full range of SAusE consonant phonemes, although the
substitution of homorganic stops for fricatives is less common than a decade
ago: people “now are more sophisticated than before; they laugh at pronunciations like “shellfish” for selfish, whereas a decade ago they wouldn’t have
known”. Nevertheless, English fricatives and affricates still cause confusion,
particularly when more than one occur in the same word; for example, [ʃ] and
[z] in shoes, [s] and [z] in schools, [d] and [z] in judges, and older speakers
commonly hypercorrect initial aspirate [h]. Speaker voice quality, the product
of both phonatory and resonant settings, resembles the Torres Strait languages
more than it does other varieties of AusE. In the words of Sharpe (1970: 116),
who offers the most comprehensive observations to date on the importance of
voice quality to the study of English in Australia, Aboriginal people “seem
very conscious of voice quality”. This is certainly true also of Islanders.
Characteristic of spoken TSE are the following variable sound features,
some of which are represented in written English:
six-monophthong system: /i, e, a, o, u, ə/. The long/short distinction between
AusE monophthongs is generally lost in TSE. TSE-speakers may not
distinguish, for example, between sit/seat and heart/hut and tend to
pronounce all vowels as long monophthongs (with the exception of
“true” diphthongs). This may be the result of transference from the
creole, which has five monophthongal vowel phonemes and is in the
process of adding schwa.
replacement of fricatives [f/v] and [θ/ð] by their corresponding homorganic
stops, [p/b] and [t/d], respectively.
replacement of fricatives [ʃ/] and affricates [tʃ/d] by the two fricatives [s/z],
the only fricatives found in the island languages.
absence of initial aspirate [h], e.g. The actor died in america by [of] an heart
hypercorrection of initial aspirate [h] in pronunciation, e.g., [hakst] ‘asked’;
[hol] ‘old’.
absence of linking -n- between the indefinite article and word-initial vowel,
e.g., by dinghy it take you half a[n] hour or [an] hour; a[n] ordinary
house; a[n] uncle.
simplification of word-final consonant clusters, e.g. because you might get
loss [lost] by traveling in a bad wheather. This may be the result of
transference from the island languages, exposure to the simplification
of such clusters characteristic of informal AusE, or both.
2.3. Morphology
The morphology of TSE is both less complex and more regular than that of
SAusE. The nominal and verbal inflectional suffixes of SAusE tend to be
omitted in TSE, as they are in rapid spoken AusE, especially when a conso-
nant cluster results, though they can be retrieved by appeal to a target.
Variable morphological features of TSE are:
absence of noun plural suffix -s, especially when plurality is indicated by
grammatical or real-world context, e.g., We know all our land
mark[s], everything; If the tourist[s] are visiting the Thursday Island;
we kill animal[s] up here in the Torre Straight but for good reason;
people kill animal[s] for experiment; inside the basket there are thick
woods and big coconut[s]; one of main reason[s] I chose to create this
magazine; they are one of the main thing[s] that make the world a
wonderful place; to play marble[s]; or people just put them in plastic
garbage bag[s]; Yesterday my grandson ask me all sorts of
question[s]. He even ask if witch[es] existed.; all these question[s];
The two island[s] outside of Murray is called Dauar and Waier; The
two bloke[s] kept on carrying on until when Jeremy said …; kick in
between two post[s]; plus two small shop[s]; about 5 drum[s] of
petrol; nine finger[s]; there are lots of song[s] written by you; not
many town[s].
overgeneralisation of the plural suffix (hypercorrection of irregular English
plurals), e.g., young childrens; The childrens read his books all over
the world; Some of the fishermens goes to fish there.
omission of the nominal possessive suffix -s, e.g., Sometime I go to my
cousin[’s] place to play marble; my dad[’s] great granpa[’s] name
Pablo Ahmat; Mr Farriner[’s] bakehouse; Georgina[’s] mum pass
away last week…
transposition of 3rd person singular verb suffix to singular subject, e.g., Its
mean …; Its happen …
no count/mass distinction, e.g., Just one bread eh? [loaf of bread]; to make
jackets and other clothings; too many destruction; equipments; informations; have a tea; I would like to know more about … how I can
help to stop these poor treatments from happening; inside the basket
there are thick woods [pieces of wood] and big coconut[s].
no pluralia tantum distinction, e.g., The umpire of the game where’s [wears]
black t’shirt, black sox, black short and black shoe.
non-standard use of the definite article, either by inserting it when SAusE
speakers would not, e.g., if we have to speak about the self-government; Luke … start to go to the English school; If the tourist are
visiting the Thursday Island …; or omitting it when SAusE would
insert it, e.g., one of [the] main reason I chose to create this magazine
is …; They follow [the] living God; When we start make a move to
contact Jesus, [the] Holy Spirit helps us to be still. [The] Holy Spirit
helps us …; I told [the] Archbishop to answer all these question.
non-standard use of the indefinite article, by inserting it before pluralia
tantum nouns, e.g., he were wearing a joggers; Luke was wearing a
blue jeans, or other plural nouns, e.g. The novel is about a witch who
… act like an ordinary ladies; Then they went to Hollywood and
became a movie stars; by inserting it where SAusE would normally
omit it, e.g., because you might get loss [lost] by traveling in a bad
wheather; or by omitting it when required by SAusE, e.g., We had
[an] exam to complete; by dinghy it take you half a hour or [an] hour;
Boigu Island is also [a] very small island; We still got [a] chance.
confusion between masculine and feminine gender in the third person singular
personal pronoun, e.g., The novel was about witches and an old lady
with his grandson. The old lady kept on telling him grandson about
witches; The novel was about a little boy and her grandmother; He’s
name was Susannah; She done his [her] work very well. This may be
a transference error from TSC, where the singular 3rd person possessive pronoun em is unmarked for gender.
absence of third person singular present tense verb suffix -s, e.g., Jesus
enjoy[s] any food; My brother belong[s] to the boy scouts.
use of non-standard past tense verb forms characteristic of other non-standard,
including Aboriginal, English varieties, e.g., we also seen [saw] a
picture; I done [did] it all in half an hour; They give [gave] up their
overgeneralised past tense forms, e.g., he had spilt his head open and bleeded
to death; Jim Henson camed up with a most fantastic and adventurous
movie; Dian Fossey keeped a diary; She slipped, fell and hited her
head on a desk. The latter form occurred in a number of student essays
despite the written instructions to recast as narrative: “You slipped,
fell and hit your head on a desk.”
use of the uninflected form of the verb in narrative to indicate past meaning,
especially when the context clarifies the relative time, e.g., he dive[d]
for pearl or crayfish hunting; Georgina mum pass[ed] away last week
on the 17–3–95; At new years day all the man drest up with coconut
lives [leaves] and carry [carried] baskets; Only ones [once] I get
[got] hit; Once my Dad ask[ed] if he could look after our dog; He
look[ed] at me; My family move[d] there because my grandpa
move[d] there; Last time I hide [hid] my key in the dining room and I
can’t [couldn’t] find it; My little brother took the key and throw
[threw] it someway in the garden; Yesterday my grandson ask[ed] me
all sorts of question; He was born in Norway and shift[ed] to England; Bruno was my favourite in the movie he always make [made]
me laugh because he always eat [ate] foods; Nothing else happen[ed]
for the rest of the afternoon; John get [got] here as soon as he can
[could]. Uninflected forms of verbs (and nouns) tend to be the default
option in TSE, as can be seen also in the following examples, where
the progressive suffix -ing is omitted, e.g., I remember one time we
were play[ing] scidels [TSC skidel from English skittles] at the big
filled [field]; I am eagerly wait[ing] on your reply.
While the uninflected verb form is preferred, some TSE speakers variably use
two preverbal nonstandard tense markers, been (often pronounced [bi]
and spelled “be”) ‘past tense’ and go ‘future tense’, e.g., What you
been done for us? ‘What did you do for us?’; So that’s the resolution
be come out from that workshop there ‘So that’s the resolution that
came out of that last workshop’; God He go come then ‘God will come
then’; Who actually go monitor them outcomes? ‘Who will actually
monitor those outcomes?’ This usage reproduces the TSC pattern,
which expresses past and future tense through the analytic pre-verbal
markers bi(n) and go, respectively, whereas English uses primarily
verb inflection, vowel alternation, or modal verb. The first example
here shows the over-generalisation of past tense formation, the
speaker using both preverbal past tense marker and nonstandard past
tense form done.
future tense after time conjunctions when, before and as soon as with future
meaning, e.g., We will get in touch with you later, as soon as we will
finish what you needed; When he will take me there, I …
lack of agreement between subject and tensed verb, e.g., The cannon is
outside the Council Chamber and lie on the side of the shed; one
person have to go throw the dice; our island look like a dugong; you
have to do as it ask you to; it take you 15 minutes; Every body have to
be in the house when they came out; When it rains, the Boigu airport
sometimes get big logs and tree in the ways; he were wearing a
joggers; When the boat reaches the wharf then everyone comes up
and walk on to the wharf; I really like to follow people that doesn’t do
the wrong thing, but always do the right thing; Usually when tourists
comes on TI they get to see everything; It all was about the witches
was trying to turn every children into a mouse; The umpires also is
control of the game; The girls was on the other basketball court; So
today some periods was very fun and some periods wasn’t enjoyable;
Animals doesn’t like to be left alone; People who goes on trips.
non-standard demonstrative adjectives, e.g., Them comments can go inside
you; have meetings with all them communities; them issues, them key
issues; the people there standing by with them Apostles.
non-standard comparative adjective constructions, e.g., It’s is more bigger;
more better; because I am a little bit short than the others; we come
more closer to that festival, the Coming of the Light.
absence of the adverbial suffix -ly, producing an invariant adjectival/adverbial
form, e.g. quick as both adjective and adverb. Many non-standard
English varieties also regularise this morphological distinction, which
is not categorical.
substitution of negative form no for “not”, possibly transference from TSC,
e.g., but no condemning anything, We no belong to sit down.
2.4. Syntax
TSE syntax is also less complex than SAusE, though whether it is less
complex than other purely spoken varieties of non-standard English would
need a detailed comparative analysis. TSE has borrowed syntactic constructions from TSC, e.g. predicate marking, the belong-construction and verbchaining, which do not occur in other AusE varieties. Prepositional usage,
where it differs from SAusE, also appears to be a TSC transfer.
Some variable syntactic features of TSE are:
absence of BE, whether copula, base form or progressive auxiliary, e.g.,
because it [is] peaceful; Thursday Island [is] right up the tip of
Australia and it is really small; you have to do as it ask you to but if it
[is] correct then do what it ask you to; my dad great granpa name [is]
Pablo Ahmat; God [is] always more; we [are] happy to helping you
out; Thank you for coming and will [we’ll be] sure to see you next
week; As from today canteen new trading hours will [be] only in the
afternoon Monday – Saturday; We should take care of our animals
and not [be] killing or torturing them; So really is that all you [are]
going to be doing; I [was] closing up and went to turn off the
a special case of copula absence, which occurs in existential clauses beginning
with there is/was/are/were, e.g., There [there are] two teams in the
game. In rapid AusE speech, there are/were are similarly elided and
the copula disappears.
absence of subject relative pronouns required in SAusE but not in all nonstandard varieties of English, e.g., It all was about the witches [who]
was trying to turn every children into a mouse; So that’s the resolution [that/which] be come out from that workshop there.
pronoun copying in relative clauses, e.g., This is a national event which we
find it very important; … or someone that you want him to be saved.
use of two past tense markers, analytic preverbal been and postverbal inflection -ed or its equivalent, e.g., when Jesus Christ been ascended to
Heaven; What you been done for us?
verb tense mixing, e.g., I leave him with my mother when I went to do the
shopping; Suddenly he noticed that strange people carrying cameras
and running through the streets, he decide that he would check it out;
That was very funny because I can’t picture are [a] big woman like her
running up and down the court; when they arrived on Thursday Island
Willie Nelson will take them on the bus to show them around T.I.;
Every body have to be in the house when they came out; If you want to
go there it took half an hour on the plane.
modal verb strings patterned on TSC constructions, e.g., he had must of
slipped; Can’t you tell the boss I’ll can get soon finished; I’ll tell him
you’ll can make it. The last two sentences are also examples of lexical
double negatives, e.g., he couldn’t see nothing accept [except] a big green
frog sitting on a hill; You don’t want no money anywhere?, which are
characteristic of many non-standard English varieties.
Absence of data is not a good basis for argument: however, an analysis of
Thursday Island students’ written English and many years of observation
suggest that the syntactic inventory of many TSE speakers may lack three
SAusE rules, passivisation, it-clefting and backshifting, the first two being
more typical of, though not exclusive to, written English. The passive construction in TSE is rarely used in speech and when it occurs in Islanders’
written English, differs from SAusE in that the base form of the verb + from
are used instead of the past participle + by: a young man was scabbing through
new garbage that was recently throw out from the neighbours. Focusing in
TSE is generally through insertion of focus markers, although a variant of Left
Dislocation exists and will be discussed below. The English rule which
“backshifts” the verb in reported speech is rare: in the second period Mrs
Hallewell told us we can’t go cause we were silly and making to [too] much
noise; Then she went on and told them what time they’ll come; Once my Dad
ask if he could look after our dog while we’re on holiday.
While the grammar of many TSE speakers may lack syntactic rules found
in SAusE, it may nevertheless contain rules, perhaps unique to TSE, borrowed
from or influenced by TSC. Predicate-marking occurs variably with singular
third person subjects, e.g., God He still exist — God always more; God He go
come then; but Moses he already tell thempla. The belong-construction
produces sentences like: We no belong to sit down, we belong to fight back,
‘We mustn’t take this lying down, we have to fight back’, which is a calque of
the modal construction used in TSC to express obligation: subject + blo (from
belong) + verb, e.g., Ai blo go nau ‘I have to leave right away’. This
construction is rather like the obsolescent bound to construction, as in I’m
bound to say that … TSE also has a rule of verb chaining after the verbs go,
come and try, rather than the conjoined construction of other AusE varieties,
e.g., They can go visit it the “green hill”; one person have to go throw the
dice; Then try see if the car and the light bulb works. This may be transference
from TSC, reinforced by exposure to American English.
Other syntactic rules are similar to, but not identical with, SAusE and
appear to be English-creole hybrids or simplifications of English rules, sometimes combined with creole elements. Thus, the TSE interrogative rule also
inserts an initial interrogative word but lacks either subject-verb inversion or
do-insertion, e.g., How big you want it to be?; How many plants you want us
to plant?; What time of the day you would like to start?; What you done to
these people in Edward River?; Why we doubt?; Why we don’t come together
much?. The main focusing rule in TSE shifts the focused phrase to the left of
the sentence, as does left dislocation, but inserts the focus marker now after the
left-dislocated phrase: From that seminar now we sit down and discuss
everything. Examples of other constructions which differ slightly from SAusE
are the use of for to introduce purpose clauses, e.g., We need a cultural centre
in major centres for [in order to] practise our culture and to educate the kids;
we can go to the playground for [in order to] play football; people kill animal
for [in order to] experiment (although the target here might be for experiments); the insertion of when after clause-initial time expressions, e.g.,
Everytime when holiday comes …; The last time when I chucked out my
rubbish …; and the use of instead as a subordinating conjunction, e.g., so that
instead you and your relations have to carry your bags home, you put your
bags at the back of the truck and you just have to walk home when the truck
carries your bags home.
A feature of TSE syntax is multifunctionality, the use of the same lexical
item in a wide range of grammatical functions. Whether the incidence of
multifunctionality in TSE is greater than in other contemporary varieties of
English, however, remains an empirical question. Moreover, given that TSC
also exhibits the typical creole tendency towards multifunctionality, the following sentences could be analysed as examples of either multifunctionality
or creole transference: They got argue [argument, disagreement] there over on
TI; Drinkers please do not asked for serve [service] in the morning, with the
verbs argue and serve used as nouns; We fright [frightened] to go forward,
with the noun fright used as adjective; He schooled [went to school] at Cairns,
with the noun school used as intransitive verb; This is not downing you
[putting you down], with the adverb down used as transitive verb but not, as is
common in AusE, in the sporting context.
2.5. Lexicon
The non-standard morphological and syntactic constructions discussed above
are less marked for outsiders than the distinctive and constantly evolving
lexicon of TSE borrowed from the traditional Torres Strait languages and the
creole. Adolescent speakers also borrow from Aboriginal English, especially
the varieties found in the Queensland coastal cities of Cairns, Townsville and
Brisbane, with their large Islander populations. Like their peers throughout
Australia, they also borrow from Black American English, which has great
Traditional language words are borrowed where no English equivalents
exist for culturally significant phenomena (which include ancestors, culture
heroes, placenames and the diversity of the natural world), for rhetorical
effect, because the English word does not have the same cultural resonance, or
because the item belongs to “shame” vocabulary and the English word is felt
to be too explicit. The traditional language stratum, which consists almost
entirely of nouns (including proper nouns), some verbs (used as euphemisms)
and interjections, fulfils a number of functions: it provides detail and accuracy, adds a prestigious element of tradition, signals the speaker’s knowledge
of custom, underlines group allegiance and permits euphemism. Thus, it may
be compared with the stylistically “elevated” and prestigious lexicon which
English borrowed from Latin and Greek and which serves analogous functions. TSE-speakers do not normally make use of this stratum of SAusE,
which attracts derision: the speaker is accused of being “flash”, that is,
arrogant. In fact, Islanders delight in “breaking down” English vocabulary to
its Anglo-Saxon elements and frequently heckle local politicians who use
Graeco-Latin vocabulary.
Some examples of common nouns are close kin terms of address and reference: aka ‘grandmother’, ama ‘mother, mother’s sister’, ata ‘grandparent’, athei ‘grandfather’, awa ‘mother’s brother’, marigeth ‘in-law
responsible for funeral ceremony’ (now often pluralised as marigeths)
and popa ‘grandparent’, used in everyday conversation and printed in
the death and In Memoriam notices in the local Thursday Island paper;
culturally exclusive words like alag ‘leaders of the annual harvest
festival’, bethei ‘driftwood’, gathawar (lit. ‘stranded turtle’) ‘nohoper’, sopsop ‘vegetable stew’, wongai ‘island date’; and euphemisms like kaka and kuma ‘faeces’, kebi meta (lit. ‘small house’)
‘toilet’, pipi ‘urine’, wakei ‘thighs’.
Names of ancestors, kinfolk, culture heroes and places clearly have no English
equivalents and are transferred unaltered to TSE. Erub (Darnley)
Islanders, for example, including those whose families have lived for
two generations on the mainland, still commonly refer in story, song
and conversation to the mythical beings who created much of their
home island landscape and to its villages, bays, points, plantations,
springs, hills, fishtraps and reefs, all of which are individually named.
Of Islanders still living on their home islands, Nietschmann (1989,
82–83) writes: “Discussions of the remembered past and the mythical
past, songs, legends and everyday conversations are filled with references to places.” This is just as true for Islanders living on Thursday
Island and the mainland: knowledge of these names is part of one’s
claim to home island identity.
Among the few traditional language verbs commonly used in TSE are euphemisms from baby talk: kaka and kuma ‘to defecate’, pipi ‘to urinate’.
Interjections in TSE include greetings, farewells and expressions of sympathy: au ‘don’t you agree?’, eso ‘thank you’, gar ‘from my heart’, kasa
‘for no purpose’, maiem ‘welcome’, matha ‘just’, sa ‘well?’, wa ‘yes’,
wadh ‘definitely’, wagar ‘yes, I sympathise’, yagar ‘sorry’, yawo
‘goodbye’. A friend’s email to me in April 1998 read in part: I will
catch up with you next time, garr. Yawo for now and big esso for
everything. Happy Easter! May you rest peacefully, Kapu Ama (lit.
‘good mother’) ended an In Memoriam notice in Torres News, February 1999. Between close friends Islanders can begin English conversations and emails with: Sa, which way? ‘Well, tell me, how are you?’
A second lexical stratum of TSE consists of nouns, pronouns, verbs, adjectives/adverbs and prepositions borrowed from the creole. It can be difficult to
distinguish English-derived creole words from standard English vocabulary.
One test is morphological, e.g., They got argue [argument, disagreement]
there over on TI; another test is in the mismatch of semantic range. Speakers
may borrow from the creole when no English equivalent exists, for rhetorical
effect, for cultural resonance and sometimes because the speaker is unaware
that the word choice is semantically or syntactically marked for mainstream
Some examples of common nouns are: kaikai ‘food, feast’, kapmauri ‘sand
oven’ (both of these as well as nana also used as verbs), kumala
‘sweet potato’, lavalava ‘male sarong’, leg ‘foot’, manggru ‘mangrove’, manyota ‘cassava’, namas ‘raw fish’, nana ‘food’, pakalolo
‘Fijian porridge’, pwakablad ‘pigblood’, susu ‘breast’, waster ‘layabout’ (probably from the nautical term waister ‘unskilled seaman’,
reinforced by the colloquial AusE waster ‘wastrel’).
A subset of the above examples consists of kin terms of respectful address,
body parts and some other nouns. Unlike the above, most of these
appear to be English words but are not semantically equivalent: auntie
‘female of generation above’, bala (from brother) ‘male sibling, close
friend, any male relative of the same or younger generation whose
name is tabooed’, boy ‘son, nephew’, cousin-brother ‘male cousin’,
cousin-sister ‘female cousin’, girl ‘daughter, niece’, sissy (from sister) ‘female sibling, close friend, any female relative of the same or
younger generation whose name is tabooed’, tawi ‘brother-in-law’,
uncle ‘male of generation above’; hand ‘from base of fingers to
elbow’, leg ‘from base of toes to knee’; and domestic words like bun
‘scone’, domboy ‘dumpling’, grease ‘fat, oil’, scone ‘fritter’, sugarbag ‘honeycomb’.
TSC second person pronouns may be used when the speaker wishes or feels
obliged to express explicitly distinctions which not found in English,
i.e., to clarify what is felt to be the ambiguity of English you: youme
‘you (sg.) and I’, youpla ‘you (pl.) but not I’, youmpla ‘you (pl.) and
I’, e.g., Youmpla can’t do everything one time. Priests and pastors
often use these pronouns for rhetorical and empathetic effect, e.g.,
Time now to prepare youmeself.
Unambiguous TSC verb borrowings include kaikai, kapmauri, nana, e.g., You
wanna kaikai ice, uh?. Examples of probable borrowings from TSC
are: when I didn’t born [when I wasn’t born]; were [where] animal
came [became] big; The world is coming [is becoming] very old; This
is not downing you [putting you down]; He got up from bed [got out
of bed], She grew [raised] the two boys; One of my friends said to
[told] another friend about my hideout; Max doesn’t school [go to
school] up here; I was born there and bought up there and my family
stay [lived] there for a long time.
Examples of adjectives and adjectival phrases are: other kind ‘unwell’, deadly
‘terrific’, halfsense ‘stupid’. It may be the case that words such as
deadly, moll ‘amoral female’ and waster, which largely disappeared
from the active vocabulary of AusE a hundred years ago and reappeared relatively recently in adolescent slang, were reborrowed from
indigenous varieties.
Adverbs and adverbial phrases are mainly spatial: e.g. He can talk language
both sides [both traditional languages]; They moved to one side [to the
other side]; I now read from the Bible, page 15, New Testament side
[of the New Testament]; longside ‘next door’; anyway ‘anywhere’,
someway ‘somewhere’, or spatio-temporal and temporal: behind
‘later’, bye and bye ‘eventually’, first ‘immediately’, e.g., Come here
first [straight away]; Youmpla can’t do everything one time [at the
same time], but they include other marked usage such as: First period
we had maths which wasn’t great, then we had typing which wasn’t
great too [either].
Examples of marked prepositional usage are: At [on] new years day; the story
happened at [in] England; at [in] Norway; he died at [on] November
1990 by [at] the age of seventy-four; The actor died in america by [of]
an heart attack; She had him down [in] Cairns; I don’t know why
people tiss [tease] me for [about] him any way; which will be held at
Cairns for only two days [in only two days time]; The animals didn’t
do any harm for [to] the people that are killing them; Our child was
out there, I was scared from [of] him; we have to be very careful from
[of] all kind of dangers; Watch out from [for] witches folks; a young
man was scabbing through new garbage that was recently throw[n]
out from [by] the neighbours; These are the four areas inside [in] that
paper; I get worried of [about] this; I only hade brothers on [with]
me; cruelty on [to] animals; if you want to go on [by] plane it take you
15 minutes or so; they also … work on [at] ordinary jobs; The cannon
is outside the Council Chamber and lie on the side of [beside] the
shed; those alag came and throw big sticks to [at] us; They can go visit
it the ‘green hill’ where [with] the rainforest and the cannons;
Thursday Island is right on top of Australia, the little island where
[with] other small islands around it; We always travel to Yam island
with [by] dinghy. TSC prepositions are based on English forms,
though with different semantic ranges, and most of the non-standard
prepositional usages above are calques of TSC constructions. Thus,
the writers have chosen at or on to express the basic TSC preposition
of location, lo (from along), which according to context can be
translated as ‘along, at, by, in, on, through’; for with time phrases to
specify length of future time; from to express the action of moving
away from objects of fear or concern, modelled on TSC prait prom
‘frightened of’; and where or on as calques of TSC we (from where) or
ene (possibly from in the), respectively, both meaning ‘with’.
Continuity of verbal tradition
TSE-speakers continue to observe aspects of Torres Strait verbal tradition to a
greater or lesser extent depending on age and circumstance. Islanders still
practise name taboo, the avoidance of affinal personal names — or even of
phonological strings homophonous with those names. For example, many
Islanders would feel shame if they happened to utter the word “road” in the
presence of an in-law named “Rod”, because the two words are homophonous
in TSE. Similarly, Islanders may refer publicly to their in-laws as “Mrs X, Mr
Y”, a practice which preserves the cultural force of the TSC avoidance terms,
oman ‘female in-law of same generation’ or tawi ‘male in-law of same
generation’ (themselves a continuation of traditional language name taboo).
Name exchange has disappeared with bureaucratisation but not the verbal
strategy of including close friends in the immediate family and showing
respectful politeness by addressing them as sissy ‘female sibling’ or my sister,
bala ‘male sibling’ or my brother, auntie ‘female of above generation’, uncle
‘male of above generation’, aka ‘elderly female’ and athei ‘elderly male’.
Traditional language lexical items are used as euphemisms for body parts and
bodily functions. Words like rat, which have power to summon up unpleasant
phenomena, are likewise avoided but, whereas TSC-speakers substitute a
traditional language word, TSE-speakers are more likely to spell out the word
as “r-a-t”. Many speakers also follow the island language pattern in their use
of English yes to assent to the truth of a statement or question, whether
affirmative or negative. In using yes to signify assent to a preceding negative,
they contravene the English rule which requires no in such cases.
Proper names are another important site of verbal continuity. Surnames
became common only this century and most were derived from the name of a
patrilineal ancestor; children are given, in addition to an English first name,
the name of a grandparent or other senior relative and an English nickname,
useful when the “real” name is tabooed. In the past, the nickname might have
been, e.g., Karom ‘Clumsy’ (lit. ‘goanna’); now it is “Porgy-boy” for a plump
George. Names of culture heroes and significant locations are similarly transmitted through traditional stories and performance, whether in English or
island languages.
In spoken narrative, rhetorical markers appear to have been transmitted
from traditional language to creole to English: one time initiates a narrative;
finish now closes it; the same tracks a participant in the narrative and is
equivalent to the legal phrase, the aforesaid; all right marks the end of the oral
equivalent of a paragraph; now focuses on the preceding left-dislocated word
or phrase; and said or say introduces a quotation, e.g., So the teacher told me
said: “Try to hold it straight.”
Like the island languages, TSE privileges the spatial dimension over the
temporal, which is reverse of the predominant Western European pattern
reflected in expressions such as “There’s a time and a place for everything” or
“to arrange a time and place for a meeting”. Locational words like here, there,
front and behind imply a secondary temporal sense for mainstream English
speakers: now, then, earlier and later, respectively. However, Islanders tend
to use the spatial terms where mainstream English speakers would either omit
them or use their temporal equivalents — overtly temporal terms are marked
for Islanders, as one can see in Ai blo go nau ‘I have to leave right away’.
Thus, the sentence, So that’s the resolution be come out from that workshop
there, though it signifies both the time and place of the workshop, would in
Standard English be more commonly rendered as So that’s the resolution that
came out of that last workshop and I’ll come behind would be I’ll come later.
Island languages require a spatial expression with verbs of location and
movement and, when calqued into TSE, the sentences may be marked for
mainstream speakers, e.g., The people there standing by with them Apostles
… According to the anthropologist Nietschmann (1989: 82), “[t]he cultural
history of Torres Strait Islanders has a strong spatial dimension. Events in the
historical and mythical past occurred at places, not simply at specific dates.
Exactly where something is said to have happened is more important than
when it happened. In travelling about their islands, waters and reefs, Torres
Strait people pass through their history, which is linked to land and sea
environments. The concept that history occurred at places, not at times, is
emphasised by many indigenous peoples, and certainly by Melanesians.”
Conclusion: implications for literacy
A brief impressionistic description of TSE cannot do justice to this variety. Its
sound inventory, grammar and lexicon contain features which occur in other
non-standard AusE speech varieties as well as features which are unique to it
and appear to be transferred from the island languages. TSE morphology is the
result of simplification and overgeneralisation processes, typical of creoles
and interlanguages but also of the history of English, which have reduced
some of the irregularities found in SAusE, such as nominal and verbal inflectional suffixes, the nominal count/mass distinction and pluralia tantum, and
the distinction between strong and weak verbs.
TSE speakers manifest a preference for multifunctionality and lexical
invariance. Their syntactic inventory also appears to lack certain stylisticallymarked rules of SAusE more characteristic of writing than speech, while
containing rules, probably unique to it and apparent transfers from TSC, such
as the belong-construction and verb-chaining. TSE, like the island languages,
privileges the spatial over the temporal and its speakers rely for rhetorical and
narrative effect on the use of verb repetition to represent the passage of time,
context-bound deictic markers and prosodic and paralinguistic cues of intonation, voice quality, sound-level shifts and relative speed of utterance, as well
as a host of nonverbal features, such as facial expression and hand and body
TSE lexicon lacks the Graeco-Latin stratum of SAusE but contains two
stylistically marked strata borrowed from the traditional languages and the
creole, respectively. It is these borrowings which most obviously distinguish
TSE from other varieties of English.
These general observations have implications for written English, since
TSE-speakers naturally reproduce their grammars in writing. I have already
noted that Islanders are exposed to a variety of Englishes, that their acquisition
of the local variety of English is influenced more by non-standard oral
varieties than by the written standard, and that as language learners they have
strategies which allow them to appear more competent in English than is
actually the case. In fact, if all communication were conducted in spoken
English, only the non-standard lexical items would be particularly remarked
upon: non-standard grammar becomes apparent in written texts.
Most trivially, the above examples from students’ essays show that their
spelling is affected by the absence in their speech of initial h and the distinction between long and short monophthongs and substitution of stops for
homorganic fricatives. More significantly, word-final consonant cluster simplification affects morphology and syntax, particularly in the written representation of constructions with past participles and conditional verbs, the base or
uninflected form being preferred to the suffixed form. Examples occur in the
representation of noun phrases with past participles, e.g., Guest, distinguish
guests, ladies and gentlemen; and the first publish book was ‘Kiss Kiss’; in
passives, e.g., This village is call Sigabadu; it is look after by the Noahs
family; It should be stop strait away; The equipment need in a touch football
games are two things; a young man was scabbing through new garbage that
was recently throw out from the neighbours; they were kill for meat; His head
was bash and he died at November 1990; It was first publish in the year 1983;
and in conditional clauses, e.g., I[’d] rather write my assignment out not type;
Now I[’d] like to welcome Miss Blanchard; I like my English class but I[’d]
appreciate it if we watch video more often.
An analysis of the written corrections on high school student essays and
discussions with teachers found that by far the majority of the non-standard
features objected to by teachers were either morphological, namely absent or
overgeneralised suffixes, which affected the syntactic structure of sentences,
or lexical, such as the use of hand for arm and leg for foot, which are based on
island language semantic distinctions. Given no explicit teaching or modelling
of SAusE, students either overgeneralise or use as default the base, uninflected
form of nouns, verbs, adjectives and adverbs. Literacy is also affected by
broader factors: code-switching between English and the creole, where the
boundaries are in any case unclear; discourse sequencing and signposting,
which present particular difficulties for members of a predominantly oral
tradition who are asked to write context-free prose for readers who do not
share the same knowledge base; and the cultural privileging of the spatial over
the temporal, which, when the two conflict, confounds English narrative
I particularly wish to acknowledge the help given to me by staff and students of the
Thursday Island State High School, which gave me permission to reproduce part of my
1996 consultancy report on the major linguistic differences between English and Torres
Strait Creole. Cathy Hallewell, head of the High School English Department, arranged for
me to view student essays, which provided most of the examples, and I benefited from
discussions with Slim Apiuta, Dion Barnett, Margaret Beetham, Lorraine Boyland, Joy
Brady, Janine Butcher, Robyn Coleston, Paul Crosisca, Christine Cutts, Agnes Cygan,
Patsy David, John Favas, Ellie Gaffney, Annie Gamia, Steve Garrett, Cathy Hallewell,
Jenni Jensen, Judy Ketchell, Charlene Lee, Lizzie Lui, Theresa Maloney, Michelle
Martin, Kathleen Mene, Carolyn Modee, Harriet Naawi, Selon Namok, Azzie Noah,
Lillian Noah, Marcus Pedro, Albert Reuben, Matt O’Riley, John Singe, Christine
Stratigos and Denise Turnbull. I am also grateful to Koiki Mabo (now deceased) for his
introduction to the Black Community School in Townsville, and to Tom Lowah (also
deceased) for his introduction to the Cairns Islander community. Funding for the research
was provided by the Australian Institute of Aboriginal and Torres Strait Islander Studies,
the Australian Research Council and the Queensland Department of Education, Peninsula
The reader is also referred to Willie Thaiday’s autobiography, Under the Act, for
numerous examples of written Torres Strait English.
Nietschmann, Bernard. 1989. “Traditional sea territories, resources and rights in Torres
Strait”. In John Cordell, ed. A Sea of Small Boats. Cambridge, Mass: Cultural Survival,
Inc., 60–93.
Sharpe, Margaret C. 1970. “Voice quality: a suggested framework for description and some
observations”. In Stephen A. Wurm and Donald C. Laycock, eds. Pacific Linguistic
Studies in Honour of Arthur Capell. Canberra, ACT: Pacific Linguistics, C-13, 115–34.
Shnukal, Anna. 1988. Broken: an Introduction to the Creole Language of Torres Strait.
Canberra, ACT: Pacific Linguistics, C-107.
Thaiday, Willie. 1981. Under the Act. Townsville, Q: N.Q. Black Publishing Co.
Aboriginal English:
Adopted code of a surviving culture
Ian G. Malcolm
Edith Cowan University
The term “Aboriginal English” (henceforth (“AbE”) will be used in this
chapter to refer to a range of varieties of English spoken widely, and sometimes written, by Indigenous Australians, which have developed independently of, but alongside, Australian English (henceforth “AusE”) since 1788.
These varieties, while maintaining certain regional markers and exhibiting a
spectrum of stylistic variation, which may emphasise their relatedness to
pidgins/creoles on the one hand and AusE on the other, have developed
sufficient commonality both of form and of function to enable them to be
comprehended, collectively, as a dialect. While the term “Aboriginal English”
is used in this paper and in most of the relevant literature, the use of indigenous
terms in naming the dialect is preferred by some writers. For example,
Enemburu (1989), writing in Victoria about AbE across Australia uses the
term “Koori English” on the grounds that his paper is written in Koori country.
In keeping with the circumstances which have brought it into existence,
AbE is both English and an Aboriginal language. Sandefur in his survey
(1983) includes it among “modern Australian Aboriginal languages” and
other scholars, including Dixon (1980), Eades (1983) and Harkins (1994)
have demonstrated the linguistic and sociolinguistic continuity of AbE with
communicative forms and patterns of Aboriginal people which pre-date their
adoption of English.
Aboriginal English may also be compared with other “New Englishes”
spoken in parts of Asia, Africa and the Pacific, which derive from a history of
colonisation involving English. In this regard, it may be seen as an indigenised
variety of English, in the sense of “an originally foreign language which has
become adapted to the local culture and undergone linguistic changes which
reflect the influence of local languages” (Siegel 1997: 119).
The literature on AbE comprises a number of formal and informal sources.
There is a discontinuous research tradition reflected in a number of theses or
dissertations relating to studies of groups of speakers in Queensland (Readdy
1961; Dutton 1964, 1965; Alexander 1965, 1968; Eades 1983), Western
Australia (Malcolm 1979; Muecke 1981), New South Wales (Hitchen 1992),
Victoria (Fesl 1977), the Northern Territory (Harkins 1988) and South Australia (Wilson 1996). The educational relevance of AbE has motivated a number
of linguistic descriptions of the dialect in (or in relation to) the educational
domain, for example, in Queensland (Department of Education, Queensland,
Van Leer Project 1970), the Northern Territory (Sharpe 1976, 1977; Walker
1982), Western Australia (Kaldor and Malcolm 1979, 1982; Malcolm 1996;
Malcolm et al. 1999) and South Australia (Mühlhäusler and Rose 1996; Sleep
1996). In addition, there have been studies which have focused upon particular
aspects of the semantics of AbE (for example Harkins 1990, 1994), its lexicon
(Arthur 1996), its discourse features (Sansom 1980; Muecke 1983; Malcolm
1994), its varieties (Douglas 1976) and its origins (Malcolm and Koscielecki
Some research has focused on the particular educational issues associated
with the use of varieties of AbE (for example Flint 1968; McKenry 1995; Kale
1995; Gibbs 1998), and a number of research-based resource materials are
available for teachers (for example Hudson 1992; McRae 1994; Slattery 1994;
Eades 1995; Malcolm 1995; Berry and Hudson 1997).
There is also a non-linguistic literature on AbE. The distinctive use of
English by Aboriginal speakers came early to the attention of colonial commentators and was often the subject of critical comment (as discussed, for
example, in Troy 1990 and Malcolm and Koscielecki 1997). It has also been
represented in creative literature by non-Aboriginal and, more recently, by
Aboriginal writers.
This chapter will be concerned principally with AbE as spoken in the
southern parts of Australia and with the socio-historical circumstances of the
progressive indigenisation of the dialect.
Reconstruction of some social aspects of initial contact
2.1. The concept of language ecology
Mühlhäusler (1996) has employed the theoretical notion of language ecology
as a frame within which to locate his recent research on AbE in the lives
of teenagers in desert communities. His language ecological model of the
Pitjantjatjara people sees the key concepts of Dreaming, Law and Language as
fundamentally linked to the people’s wider ecology, and that ecology as
ensuring the maintenance of their well-being while enabling their adaptation
to changing circumstances.
It is suggested by Mühlhäusler that English has an important place in
contemporary language ecology for Aboriginal people, in view of the fact that
it is for the great majority of them the first language. However the place of
various English-based forms of communication in that ecology has not yet
been definitively established by research (Mühlhäusler 1996: 157).
Clearly, English was at first a foreign intrusion into the Aboriginal
language ecology. We shall attempt to trace some of the factors associated
with the progressive adaptation of that ecology to include it.
2.2. The pre-existing linguistic ecology
At the time of the establishment of New South Wales the Australian continent
was occupied by upwards of 300,000 Aboriginal people (perhaps considerably more, according to Reynolds 1996: 20). Geographical isolation had, up
to this point, provided a buffer which had minimised contact with the outside
world, saved the local languages and cultures from the kind of assaults
associated with foreign aggression and conquest experienced in many other
parts of the world and favoured the development of a distinctive view of the
The social organisation which prevailed in Indigenous Australia was
localised, with people associating primarily with others who shared with them
a common territory and language (Rumsey 1993: 192). There were perhaps
500 reasonably distinct “dialectal or linguistic units” (Berndt and Berndt
1964: 28), comprising, on the average, 500 to 600 people. The linguistic
variation across the continent was, then, very considerable.
Although the linguistic/social units (or “tribes”) were relatively self-
contained, multilingualism would have been common (Dixon 1980: 69;
Elwell 1982; Rumsey 1993: 105; Mühlhäusler 1996), favoured by the seminomadic lifestyle and, in some cases, by exogamous marriage patterns. A
speaker’s identification with a particular language was strongly associated
with the land where that language belonged. Even today, as Rumsey
(1993: 199) has noted, land-related “language ownership” is important to
Aboriginal people whether or not the language is still spoken.
The pre-contact culture was one in which knowledge was orally transmitted but subject to careful controls. Not everybody was free to use all the
languages or varieties which they knew were used within their group or to pass
on knowledge which had been committed to them. This was, as Walsh
(1994: 225) has called it, a “knowledge economy”. Language — with its levels
of variation — was important in its capacity of preserving restricted knowledge
within the appropriate group, as well as for communicating knowledge.
The other key element affecting the pre-contact linguistic ecology was,
according to Berndt (1979: 13) the motivating force of religion: “The Dreaming exemplified these two facets of humanity, the spiritual and the physical.
The interplay between these permeates the whole of Aboriginal religion —
and social living in general” (1979: 15). As we have noted, Mühlhäusler
(1996) saw Dreaming, Law and Language as central to the linguistic ecology
of the Central Australian groups he was investigating. Muecke, similarly,
relates the Dreaming together with language, and sees the relationship as
observable in the way in which Kimberley narratives “instruct by providing
categories for the understanding of Aboriginal history, law and spiritual life”
(1981: 270).
The Australian linguistic world into which English was so unexpectedly
planted, then, was one in which perspectives extended little beyond the local
group, in which language, land and identity had long been fused in a spiritually validated union. It was a self-contained world where there was an understood social and linguistic consensus, and where a wider world might only
enter people’s awareness through transitory involvement in ceremony or
trading. But, as Berndt and Berndt (1964: 38) observed:
people who rely heavily on their own resources and their traditional background, whose view of the world is turned inward rather than outward, tend to
suffer considerably when they come into contact with aliens whose whole
way of life is very different — in this case, with Europeans.
2.3. Some circumstances attending early contact
When the British took up residence in the vicinity of Sydney in 1788 they
came in significant numbers considering the relatively low indigenous population. The Eora people, numbering some 1,500 would have been almost
equalled by the 1,300 arriving in the first fleet (Horton 1994: 342). The
isolation which had been so influential in determining the evolution of the
local culture was eliminated overnight and, as new arrivals continued to come,
the majority status of the Aborigines was rapidly reversed. The imbalance was
aggravated by the devastating effect among the Aboriginal people of fatal
diseases. By 1840 there were 38,005 convicts in New South Wales and none
remained of the original Aboriginal population of 1500 in Port Jackson
(Malcolm and Koscielecki 1997: 8, 17). The growth of the immigrant population accelerated as, from the mid-19th Century, the convicts were outnumbered by gold seekers, but over the same period, where contact occurred, the
Aboriginal population declined significantly. Indigenous Australians, then,
quickly became a linguistic and cultural minority.
When communication between the Aborigines and the colonists took
place, it was apparent that the norms for interaction and interpretation were not
shared. At first the Aborigines, who used sign language as a functional skill
while hunting, employed it to overcome the language barrier (as no doubt they
had with speakers of unfamiliar Australian languages), and this enabled an
elementary exchange of messages to take place. However, conversation by
signs and gestures, even if extended, as it sometimes was (Troy 1993: 33), is
open to a great deal of misinterpretation and sometimes led to confusions which
took some time to clarify (Malcolm and Koscielecki 1977: 48). Moreover,
understanding the implications of the contact experience and the expansionist
ambitions of the British entailed a cognitive leap which would hardly have been
credible at first contact for a people who had no prior experience of land
conquest (Berndt and Berndt 1964: 39). It followed from this that the settlers
were pleasantly surprised at the amicable nature of the initial response the
Aborigines made to them. They were, in the words of Tench, a senior officer of
the first fleet, at first “timorous”, but “shewed no signs of resentment at the
Governor’s going ashore” and were “highly entertained” by the newcomers
(Troy 1993: 33). The cautious curiosity at what was happening was a typical
reaction of the Indigenous people in different parts of the country to their initial
contacts with the British. However they had, as yet, no semantic categories in
which to place these intruders and no prior communicative experience which
could equip them to interpret or respond to what they were doing.
Time was, of course, to change this. It was not long before the Aboriginal
people began to have their possessions stolen and (despite the apparently
peaceable intentions of Governor Phillip) to suffer physical attack from the
colonists (Troy 1993: 35). They also became aware that, unlike other visitors
to their shores, these new arrivals intended to stay and to occupy their land.
The Aborigines, then, divined the assertion of power which lay behind what
the British were doing and resisted it, albeit inevitably ineffectually.
Co-occurrent with the loss of their isolation the Aborigines suffered a
threat to their traditional group, rather than individual, orientation. Phillip had
come with orders “to open an intercourse with the natives, and to conciliate
their affections, enjoining all our subjects to live in amity and kindness with
them” (Troy 1993: 34). The Aborigines, for the most part, wanted to be left
alone. Despite the initial curiosity some of them exhibited at the arrival of the
newcomers, they had their own lives to lead and their own social system to
maintain. So Phillip was faced with a dilemma. The kind of intercourse which
he was supposed to open with the natives projected upon them the role of
grateful imbibers of the wonders of British culture and civilisation. It took no
account of the maintenance or even recognition of their existing culture, or of
the terms upon which they might communicate with outsiders. It did not
foresee that the Aborigines might not, as individuals, have any interest in
engaging in conversations with the visitors. Phillip was not prepared to wait to
learn why the Aborigines were behaving as they were. Under pressure from
his superiors, he took action to sever individuals from the group so that they
would be forced to remain for an extended period in British company and to
learn the English language and culture. Thus, Arabanoo was captured in
Manly Cove and detained among the colonists with a rope attached to a
handcuff on his wrist, so that he might learn the English language and ways
(Troy 1993: 36). His detention ended six months later with his death from
smallpox, and later others, most notably Bennelong — a gifted communicator
— were captured to take his place.
The insensitivity of the colonists to the trauma which occurred when
individual Aborigines were separated from their group showed in other ways
as well. The desire to expose them to civilisation led to some (including
Bennelong) being sent to England. Many were attached to Europeans as
assistants in various colonial occupations, including exploration. From 1815 a
boarding school was opened in Parramatta, the first of many initiatives designed to take Aboriginal children away from the influence of their own
people so that they could more readily learn to communicate and live the
European way. (See further Malcolm and Koscielecki 1997).
The coming of the colonists also irreparably damaged the fundamental
link between the people, their land and their language. First they were progressively dispossessed of their land, then of their languages. The Eora people,
around Sydney, were the first to be affected but inevitably the need, or desire,
for agricultural, pastoral and mining land stretched wider afield and affected
all areas of settlement. Not only did settlers move out into Aboriginal lands,
but Aboriginal people were drawn away from their lands to the settlements.
Significant population shifts took place all over the country, weakening the
fundamental ties which bound people to a traditional area and its language
(Kaldor and Malcolm 1979: 409).
Walsh (1993: 2) has observed that Australian languages began to decline
soon after the arrival of the Europeans. Although isolated individuals among
the newcomers showed a respect for the local languages, and attempted to
record their features and to learn them, the commonest viewpoint was that
they were inferior to English and were an obstacle to the people’s learning of
English. The speech of Aborigines using their languages was disparagingly
referred to as “yabber” or “gibberish” (Malcolm and Koscielecki 1997: 20,
42) and people were under pressure from official and unofficial sources to
stop using them. English was reinforced through both education and employment practices. In time, many people came to be ashamed to use their languages in public (Donaldson 1985) and the continuity of intergenerational
transmission was broken. In some areas population movements occasioned
tensions among Indigenous groups (Elwell 1982: 85; Harris 1993: 145).
Some stronger indigenous languages took over functions of wider communication as others were weakened (Walsh 1993: 8). With the weakening of
home language transmission, the way was open in some places for existing
English-based pidgins to be transformed into mother tongues, or creoles.
The break in continuity with the language and with the land caused a
fundamental lifestyle change for the Aboriginal people. Having previously
depended on the land, and its associated language, for survival they now
became dependent on the settlers and their language. The early colonists came
to see them as almost aggressively mendicant as they transferred to them the
patterns of expectation of sharing and mutual support which they had tradi-
tionally held within their own groups (Troy 1990: 125; Malcolm and
Koscielecki 1997: 42, 52).
As we have already noted the Aboriginal culture had, prior to contact,
ascribed identity to people through the use of language which constructed
them as owned by, and owning, their traditional lands. After contact another
language, the language associated with the power which had displaced them,
was being used to ascribe a different identity to them. From the earliest times
(even well before 1788), the Aborigines had been constructed by Europeans
as miserable and deficient — a form of description which became increasingly
current in the 19th century as it was associated with what Hollinsworth
(1998: 89) has described as “scientific racism”, that is, the social Darwinist
assumption that Aborigines are on the lowest rung of the human evolutionary
Many Aborigines found themselves caught between two cultures: one
gave them respect and identity, and the other robbed them of these, even
though it could feed them. Consequently many Aborigines assumed a pattern
of “culture shifting”, spending extended periods among European people but
unpredictably moving back, sometimes permanently, to their own people
(Malcolm and Koscielecki 1997: 50, 51). This pattern of adaptation to a
“bicultural milieu” (Dixon 1980: 74), in my view, bears an essential relationship to the fact and the nature of the maintenance and use of AbE.
The spiritual interpretation of their existence and environment was thus
an essential element of the linguistic and cultural ecology of Aboriginal
people before contact. Consistent with this are the many reports that, in early
encounters with white people, Aboriginal people fitted them into their world
as spirits, perhaps spirits of their own departed relatives (Malcolm and
Koscielecki 1997: 7, 84). In this way we see that, just as the white people
were constructing the Aborigines according to the received wisdom of their
day, the Aborigines were doing the same to them. In time, however, as their
world changed irrevocably, so too would their world view, and the linguistic
and cultural ecology which accompanied it.
Reconstruction of some linguistic aspects of early contact
3.1. Varieties of English
The 18th Century English which was first heard by Aboriginal people would
have exhibited a range of varieties. According to Horvath (1985: 35), from
the early years of settlement there were three major social groups: the ruling
élite, who were, for the most part, English and Anglican; the convict group,
who were predominantly working class, male and often Irish Catholics; and
the free immigrants, including wage labourers, clerks and tradesmen with
money and servants. According to Horvath the variety which had the greatest
influence on AusE in general was Cockney, which would have carried prestige on the basis of its association with the large number of London city
convicts and given expression to a typically Australian ethos. The relevance of
this ethos to the formation of AbE is less apparent. Those Aboriginal people
who were exposed to English in the early years of settlement, while having
some access to the convicts, were not likely to have had intensive involvement
with them (Troy 1993: 41). They would also have had involvement with the
sailors, some of whom, according to the early records, they readily mimicked
and thereby learnt English swearing (Troy 1990: 16), and some of whom
could be expected to have attempted to communicate with them in Nautical
Jargon (Romaine 1988: 99) or Pacific Jargon English (Mühlhäusler
1986: 146; Troy 1990: 8). Others, who came into the employ of the settlers,
would have had exposure to a variety of regional as well as prestige dialects.
On the whole, it seems clear that the settlement situation was not such as to
expose Aboriginal people to one consistent or dominant variety of English. A
number of varieties were current which, by a process of levelling, would lead
to AusE. Other factors, however, would enter into the way in which AbE was
to develop.
3.2. Aboriginal language varieties
As we have observed, the pre-existing linguistic ecology of Australia was
multilingual and groups as small as 500-600 identified with a common variety.
The settlers in the early colony, although it took them some years to become
aware of it (Troy 1993: 43), were inhabiting territory which was associated
with a number of different languages. Some of these languages soon contrib-
uted changes to the lexicon of English as used in the colony. Dixon (1980: 9)
has shown how the words gin, dingo, woomera and gunyah resulted from
attempts to represent in English phonology the Dharuk words /diyin/ ‘woman’,
/din-go/ ‘tame dog’, /wamara/ ‘throwing stick’ and /gunyi/ ‘dwelling’. According to Dixon (1980: 11), the recognition, several years after settlement, that
there were multiple Aboriginal languages on the continent was the first
important linguistic discovery about Australia. The second, noted in 1841 by
Governor George Grey in his account of his explorations in Western Australia,
was that there was a typological similarity among many of the languages
spoken along the coastal areas of Western Australia, New South Wales and
South Australia. The fact that the colonisers’ language, English, was planted in
a context where there were a number of typologically related indigenous
languages being spoken is of significance in the chain of development which
eventually led to AbE (Lefebvre 1997, 1998; Siegel 1999).
3.3. Perspectives on the development of AbE
There is general agreement that the development of AbE was preceded by the
pidginisation and creolisation of English during the early years after colonisation. (See, for example, Dixon 1980: 69–77; Kaldor and Malcolm 1979;
Mühlhäusler and Rose 1996: 203). The course of development which was
followed was probably different in different areas. In some areas (e.g. New
South Wales) pidgins and probably creoles existed where only English is now
spoken. In others (e.g. the Kimberley, the Northern Territory and Cape York
Peninsula) varieties of AbE co-exist with creoles. In other locations (e.g.
South Western Australia) there may never have been sustained use of a pidgin
or creole by a locally-based community. There are also alternative explanations of the respective roles of the Aboriginal substrate languages and the
English superstrate in the process leading towards the development of the
dialect. Did children play a key role in the development of the initial creole out
of the pidgin input received from their elders? Or did speakers, or former
speakers, of Aboriginal languages relexify their languages as adults, drawing
on the resources of the various forms of English which were within their
hearing? Was the progression towards AbE a matter of movement along a
post-creole continuum, or was it a process of dialect levelling in the context of
the ever-widening circle of Aboriginal communication in English? It is probably unwise to see substratist and superstratist explanations as necessarily in
conflict with one another (McMahon 1994: 280). AbE shares some features
with pidgin- and creole-related varieties worldwide as well as bearing convincing Aboriginal identifying marks in form and semantics.
3.4. From a mixed jargon to NSW Pidgin English
The development of pidgins and creoles is a dynamic and often discontinuous
process, subject both to global linguistic principles and to local external
factors (Mühlhäusler 1986: 249-250). Bearing in mind these qualifications, it
has been suggested by Mülhlhäusler (1986: 134–250) and Romaine (1988:
117–156) that there is usually a progression through a jargon phase of great
individual variation to a stabilisation phase where a pidgin emerges with a
growing measure of grammatical and lexical regularity to an expanded phase
where the pidgin acquires the power to provide expression in a range of
domains and with a range of pragmatic functions. At any stage, if the variety
becomes the first language of a speech community, creolisation may occur. It
is also possible, under the ongoing influence of the superstrate, for depidginisation or decreolisation to occur.
As we have observed, Aboriginal people from the time of contact were
exposed, intermittently, to a range of varieties of English including social and
regional varieties brought by the colonists as well as contact varieties from
other parts of the Pacific. It has also been suggested that they were exposed to
the kind of “foreigner talk” which native speakers of English often employ
with non-native speakers (Dixon 1980: 71), and to the “broken English”
(Dutton 1983: 91) which they picked up from one another.
In her account of Aboriginal contact with English from 1788–1845, Troy
(1990) suggests that these various inputs led, shortly after contact, to the first
stage of pidgin/creole development, shown in the emergence of a variety she
calls NSW Jargon. This would have been a highly unstable variety, incorporating many transfers from Aboriginal languages, and it would have been
typically stigmatised by native English speakers. Yet, it would have been
maintained for its communicative value.
It is further suggested by Troy (1990) that NSW Jargon soon began to be
used by Aborigines to communicate with one another. In some cases the
Jargon would have performed the functions of a lingua franca, but it also had
other significant functions to perform. It provided its speakers with a resource
for conceptualising the new cultural elements to which they had been exposed
and for reflecting together on the contact experience.
According to this view, the stabilisation which led to the jargon becoming
NSW Pidgin English came about though the use of this variety not with first
language speakers of English but with other Aboriginal people. As Troy
(1990: 9) puts it:
Colonists attempted to teach Aborigines English and to incorporate them as
‘useful’ and ‘civilised’ members of colonial society. However, most Aborigines were not interested in adopting the colonists’ culture and did not have the
degree of access to English that would promote it as a first or second
Once the usefulness of NSW Pidgin English for their own purposes was
realised by the Aboriginal people, it is suggested by Troy (1990: 5) that it
came to be widely used “to convey information about the colonists, inter and
intra group, along their trade routes and information networks”. Accustomed
to having a repertoire of varieties appropriate to different groups, the speakers
of NSW Pidgin English began to adopt it as an auxiliary variety with a
particular social group reference.
3.5. From NSW Pidgin English to AbE
The transformation from a mixed jargon to NSW Pidgin English was well
underway within the first 50 years of settlement. This localised variety was to
be the agent of further linguistic change which would have wide reaching
As NSW Pidgin began to be used as a medium of communication along
traditional trade routes, its influence would have begun to permeate a large
part of Aboriginal Australia. Troy (1990: 2), indeed, suggests that
traditional trade routes and information networks covertly and overtly connected Aborigines all over Australia…it is likely that any developing contact
language would have travelled along these routes as part of the information
about the colonists.
According to Dutton (1983: 90), as what is now Queensland (but was then
part of New South Wales) was settled between 1823 and 1859, NSW Pidgin
English was probably used either in whole or in part as a lingua franca. The
settlement of Queensland involved movement northwards both by inland and
by coastal routes (Dutton 1983: 92) and one might expect that the pidgin was
transported by both of these means. The transmission of the pidgin might have
been favoured, according to Dutton, by the establishment of a settlement of
convicts who, for the first ten years, would have had little European contact,
and by the reported sexual exploitation of Aboriginal women by both soldiers
and convicts (Dutton 1983: 96–97). The Pidgin English used in Queensland
was clearly based on NSW Pidgin and included lexical items from languages
from the Sydney area. It also, however, incorporated items from local languages. By 1870, according to Dutton (1983: 108) it had become the common
lingua franca of much of Queensland.
The form of English used in Queensland, based on NSW Pidgin, continued its transmission along northern and eastern communication routes, giving
rise, along with other influences, in time, to Cape York Creole and Palm
Island AbE in the north and, further east, to Kriol (Dutton 1983: 109; Harris
1986; Troy 1990: 2; Siegel 1997: 135). The transmission continued further
east from the Northern Territory to bring Kriol to the Kimberley during the
mid-20th Century (Harris 1993: 149).
Although detailed documentation is lacking, it is plausible that the same
kind of dissemination occurred southwards to Victoria (originally a part of
New South Wales) and eastwards to South Australia. There, according to
Mühlhäusler and Rose (1996: 203), Nunga English, based on NSW Pidgin,
was used widely in the south in the 1840s–1890s, and in the northern parts
cattle station English, again derived from NSW Pidgin, was used from the
1920s. New South Wales established an outpost at King George’s Sound,
Western Australia, in 1826, and, prior to the settlement in Perth in 1829, NSW
Pidgin could have been transported to Western Australia by sea routes, if not
by inland means (Malcolm and Koscieleki 1997).
It seems likely, then, that the influence of NSW Pidgin on the development of AbE, while observable Australia-wide, has been exercised through
different processes and with different effects according to the conditions
obtaining in the different locations. In order to trace and understand these
processes it is useful to invoke Mühlhäusler’s (1986) distinction between the
developmental and restructuring continua which underlie the movement towards and away from the basilect. According to this distinction, it is possible
to observe a developmental continuum whereby a contact variety progressively changes from a jargon through a stabilised and expanded pidgin to a
creole. At the same time, there may be a movement in an opposing direction
whereby the pidgin or the creole may progressively change in the direction of
the lexifier, or dominant language, in circumstances where its speakers are in
ongoing contact with that language.
The process of development, which leads towards the creation of an
expanded pidgin or a creole is more likely to take place where there has been
drastic social change resulting in the interruption of natural processes of
language transmission, where the dominant and lexifier language is not widely
spoken, and where there are a number of different substrate languages spoken
among the community. Siegel (1997) has shown, drawing on data from Harris
(1986) that this is what happened early in the 20th century when various
displaced Aboriginal groups from the eastern part of the Northern Territory
converged on Roper River (Ngukurr), where the Anglican church had established a mission. According to Harris (1993), many of the Aboriginal adults,
who were multilingual, could communicate with one another in their respective languages. However the children could not, and it was this generation of
children, who had a variety of pidgin (related to NSW Pidgin) in common,
who created the creole (Harris 1993: 149). Siegel (1997: 136) argues that it
was both the nativisation of the pidgin by the children and the widespread
shifting to it by the adult speakers which enabled the creole to emerge and
become quickly established.
Restructuring of a pidgin or creole may occur where the situation of the
speakers of this variety changes so that they are increasingly exposed to the
dominant or lexifier language in a way which causes them to modify the
structure and vocabulary of the pidgin or creole to make it more like the
lexifier language. It is possible that this has occurred in parts of southern
Australia where NSW Pidgin once stabilised and even perhaps creolised but
has now given way to a variety of AbE.
What seems clear when AbE varieties are compared across the continent
(as in, for example, Kaldor and Malcolm 1982; Malcolm and Koscielecki
1997), is that over the period of their ongoing development, whether with
direct contact with pidgin or creole varieties or not, the varieties have retained,
or developed, many features in common. This may be accounted for in part by
the inheritance from NSW Pidgin, but there also would appear to have been a
process of convergence, in that, for example, contemporary varieties of AbE
in Sydney and Perth share many features in common which are not related to
the developmental continuum (Malcolm and Koscielecki 1997). This could
be interpreted as evidence that a process of koinéisation (Trudgill 1986;
Mühlhäusler 1986: 11–12; Lefebvre 1998: 8; Siegel 1997) is in progress,
whereby the culturally integrated, increasingly mobile, and nationally-ori-
ented speech community of Aboriginal Australia is expressing its perceived
commonality in a reduction or attrition of variants, although retaining some
degree of stylistic and regional variation.
Aboriginal English today
4.1. Forms of contemporary AbE
AbE today (see further Kaldor and Malcolm 1991) is a living and changing
linguistic system with sensitivity to the many stylistic, situational and regional
demands made upon it by its diverse, yet unified, speech community.
The features of the dialect most commonly referred to by non-Aboriginal
observers relate to its phonology. Some varieties of AbE (sometimes referred
to by speakers of the dialect as “heavy”) are characterised by significant
phonological transfer from Aboriginal languages or from creole, affecting, in
particular, the pronunciation of sibilant, interdental and labiodental consonants, which may not be articulated at all or may be supplanted by transferred
variants. The voiced/voiceless distinction of consonants may not be consistently recognised. In more restructured varieties (see Section 3.5), consonant
articulation will be much closer to that of AusE speakers, although variable
pronunciation of initial /h/ (even in contexts where it is absent in AusE) is
common, as is the substitution of labio-dental for interdental stops. Consonant
clusters are often simplified. Vowels tend not to be neutralised when in
unstressed positions, and initial unstressed vowels, as in words like along are
often not pronounced. AbE has stress and intonation patterns which contrast
with those of AusE. Stress is more uniformly on the first syllable and an even
high final tone is more common. Vowel lengthening, accompanied by high
key, is a means of emphasis. Liaisons are not necessarily made between
contiguous vowels and patterns of consonant elision are distinctive (for example we’as, ‘we was’; I’s, ‘I was’; we ’ent, ‘we went’).
The grammatical, like the phonological system, of AbE is more variable
than that of AusE and, in less restructured varieties, may share many features
with creoles, as well as with substrate languages (Dixon 1980; Harkins 1994).
The most dominant difference from AusE which characterises most varieties
of AbE is a reduced dependence on the copula for linking subjects to their
complements and on the auxiliary for the expression of verb tense and aspect.
Typically, speakers of AbE select from a repertoire of alternants (including a
zero allomorph) in the expression of the past tense, depending on the verb
stem and on the stylistic level intended. It has been suggested (Malcolm 1996)
that, since the repertoire usually includes standard English alternants, users of
AbE are often to a greater or lesser extent bidialectal, even from the first year
of schooling. The standard English paradigm of subject-verb agreement tends
to be regularised (that is, the third person singular of the verb does not carry
-s). The future is formed with gonna and questions are signalled by intonation
rather than by reversal of subject and verb. The formation of confirmationseeking questions with distinctive tags is also common. Within the noun
phrase, the marking of plurality and possession are optional and may be
achieved by means other than inflection. Personal pronouns in less restructured varieties retain dual vs plural and inclusive vs exclusive distinctions as in
creoles, and even in more restructured varieties may not mark gender in the
third person singular. The resumptive pronoun commonly occurs (That man
he …). Definite and indefinite articles are variable in occurrence and include
the distinctive forms dat and one. Adjectives and adverbs may occur in postclausal extensions (…yellow one; … quick-way) and carry distinctive suffixes.
Prepositions may be variable in occurrence and various locative forms in/at/
on may be used interchangeably or, in less restructured varieties, replaced
with longa.
AbE has a distinctive lexicon incorporating a majority of English-based
lexemes, though many with shifts of meaning, and also items transferred from
Aboriginal languages sometimes via pidgin/creole. The semantic categorisations and fields underlying AbE contrast markedly in many respects with
those of AusE. This has been demonstrated with respect to speakers where
Aboriginal languages are currently spoken in the same community (Harkins
1994) and where they are not (Malcolm et al. 1999). There is evidence (for
example Dixon 1980: 77) that some distinctive usages of words in AbE are
attributable to relexification of concepts from Aboriginal languages. It is
possible that relexification played an important part in the formation of AbE
from the earliest stages and that the maintenance of the dialect for communication in Aboriginal rather than European Australian contexts has favoured the
preservation of these meanings.
Discourse in AbE follows pragmatic rules which show a continuity with
Aboriginal languages rather than with AusE. There are distinct rules of
speaking which relate closely to the traditional Aboriginal concept (mentioned
in Section 2.2) of the knowledge economy and control the release of information (Eades 1983). Certain forms of taboo speech in wider Australian society
may have been taken over to perform particular dispute resolution functions in
Aboriginal society (Sansom 1980; Langton 1988). Aboriginal speakers, using
English in non-Aboriginal contexts, may fail to conform to their interlocutors’
expectations as to such matters as eye contact, turn taking and the showing of
attention (Walsh 1994). These and many other features of pragmatic contrast
may have significant consequences in educational (Malcolm 1979) and legal
(Eades 1992; Koch 1985; Cooke 1998) contexts. AbE is also the medium of
oral art, especially in the form of narrative, which incorporates features
common to oral art forms in other primary oral cultures as well as sustaining a
construction of the world which is consistent with traditional Aboriginal
values (Muecke 1981; Malcolm 1994).
4.2. Functions of contemporary AbE
It has been suggested in this chapter (Section 3.4), following the views of
Troy on NSW Pidgin, that AbE arose out of the desire and need of Aboriginal
people to communicate with one another in a way which helped them to come
to terms with their changed situation brought about by European settlement. It
was a medium which enabled the development and expression of a shared
Aboriginal post-contact discourse. The fact of the maintenance of AbE suggests that this process is still taking place. Aboriginal people are still coming
to terms with what Dixon (1980) called their “bicultural milieu” and the use of
a bidialectal English repertoire enables a group-specific discourse to be maintained alongside the discourse required for participation in the wider community which is still seen, to a greater or lesser extent, as the discourse of the
Aboriginal English also functions to maintain a meaning system which
links its speakers to their remote and less remote history. It invokes a shared
past which relates its speakers distinctively to the present. This is particularly
significant in the context of the extinction, or imminent extinction, of most
Australian languages. The loss of the traditional languages does not mean the
loss of aboriginality, since Aboriginal people have made English a bearer of
Aboriginal culture (Walsh 1993: 12). AbE is a symbol of cultural maintenance: it is the adopted code of a surviving culture.
Finally, AbE is developing new functions as its speakers are emerging
from the long period in which it has been stigmatised and suppressed. It was
observed by Morgan (1993: 3) that there has been an “extraordinary flowering of Aboriginal writing in English, almost all of it since the 1970s.” It is now
normal for Aboriginal authored texts to be selected for study in the final years
of high school and some of the most popular texts incorporate portions of
dialogue in AbE. Aboriginal writers, according to Ariss (1988) have been
engaged in the construction of an Aboriginal discourse, which indeed functions as a “counterdiscourse” (p. 132) in the wider culture. It draws on a
tradition of oracy and it foregrounds “personal experience, the particularistic,
articulated through its identification with a community” (Ariss 1988: 134).
The dispossession of their land and their languages brought about by colonisation has profoundly affected the language ecology of Aboriginal Australians.
Yet English, in large part an agent of this dispossession, has been integrated
into a post-colonial Aboriginal language ecology as it has gone through a
progressive process of indigenisation which has enabled it to serve certain
functions of Aboriginal cultural maintenance. AbE and AusE carry contrasting historical and cultural associations, and their separateness has been consistently maintained by Aboriginal speakers. Since the reasons for this still
remain, it is likely that Aboriginal English will continue to be maintained as a
distinct dialect, although many of its speakers will maintain bicultural competence through the maintenance of Australian English as a part of their repertoire.
Alexander, Diane H. 1965. Yarrabah Aboriginal English. Unpublished B.A. (Hons) thesis.
Brisbane: University of Queensland.
———. 1968. Woorabinda Aboriginal Australian English. Unpublished M.A. thesis.
Brisbane: University of Queensland.
Ariss, Robert. 1988. “Writing black: the construction of an Aboriginal discourse”. In J.R.
Beckett, ed. Past. Present: The Construction of Aboriginality. Canberra: Aboriginal
Studies Press, 131–45.
Arthur, Jay M. 1996. Aboriginal English: A Cultural Study. Melbourne: Oxford University
Berndt, Ronald M. 1979. “Traditional Aboriginal life in Western Australia: as it was and
is”. In Berndt, Ronald M. and Catherine H. Berndt, eds. Aborigines of the West: Their
Past and Their Present. Nedlands: University of Western Australia Press, 3–27.
——— and Catherine H. Berndt. 1964. The World of the First Australians. Sydney: Ure
Berry, Rosalind and Joyce Hudson. 1997. Making the Jump. Melbourne: Language Australia.
Cooke, Michael. 1998. Anglo/Yolngu Communication in the Criminal Justice System.
Unpublished Ph.D. thesis. Armidale: University of New England.
Department of Education, Queensland, Van Leer Project. 1970. Research Report on Some
Aspects of the Language Development of Pre-School Children. Brisbane: Department of
Dixon, Robert M.W. 1980. The Languages of Australia. Cambridge: Cambridge University
Donaldson, Tamsin. 1985. “From speaking Ngiyampaa to speaking English”. Aboriginal
History 9: 126–147.
Douglas, Wilfrid H. 1976. The Aboriginal Languages of the South-west of Australia. 2nd
edition. Canberra: Australian Institute of Aboriginal Studies.
Dutton, Thomas E. 1964. Some Phonological Aspects of Palm Island Aboriginal English.
Unpublished B.A. (Hons) thesis. Brisbane: University of Queensland.
———. 1965. The Informal English Speech of Palm Island Aboriginal Children, North
Queensland. Unpublished M.A. thesis. Brisbane: University of Queensland.
———. 1983. “The origin and spread of Aboriginal Pidgin English in Queensland: A
Preliminary Account”. Aboriginal History. 7: 90–122.
Eades, Diana. 1983. English as an Aboriginal Language. Unpublished Ph.D. thesis. St.
Lucia: University of Queensland.
———. 1992. Aboriginal English and the Law: Communicating With Aboriginal English
Speaking Clients: A Handbook for Legal Practitioners. Brisbane: Queensland Law
———. 1995. Aboriginal English. Aboriginal Literacy Resource Kit Series. Sydney: Board
of Studies.
Eagleson, Robert D, Susan Kaldor and Ian G. Malcolm. 1982. English and the Aboriginal
Child. Canberra: Curriculum Development Centre.
Elwell, Vanessa M.R. 1982. “Some social factors affecting multilingualism among Aboriginal Australians”. International Journal of the Sociology of Language 36: 83–103.
Enemburu, Irruluma Guruluwini. 1989. Koori English. Melbourne: State Board of Education and Victorian Aboriginal Education Association Inc.
Fesl, Eve D. 1977. Melbourne Aboriginal English. Unpublished B.A. (Hons) thesis.
Melbourne: Monash University.
Flint, Elwyn H. 1968. “Aboriginal English: linguistic description as an aid to teaching”.
English in Australia. 6: 3–21.
Gibbs, Genevieve. 1998. Educating Readers: Aboriginal English in Aboriginal Literature.
Unpublished M.Ed. dissertation. Nedlands: University of Western Australia.
Harkins, Jean. 1988. English as a ‘Two-way’ Language in Alice Springs. M.A. thesis.
Canberra: Australian National University. [Published, revised, as Harkins 1994].
———. 1990. “Shame and shyness in the Aboriginal classroom: a case for ‘practical
semantics’”. Australian Journal of Linguistics 10: 229–306.
———. 1994. Bridging Two Worlds: Aboriginal English and Cross-cultural Understanding. St. Lucia: University of Queensland Press.
Harris, John W. 1986. Northern Territory Pidgins and the Origin of Kriol. (PLS, C-89).
Canberra: Department of Linguistics, Research School of Pacific Studies, Australian
National University.
———. 1993. “Losing and gaining a language: the story of Kriol in the Northern Territory”. In Michael Walsh and Colin Yallop, eds. Language and Culture in Aboriginal
Australia. Canberra: Aboriginal Studies Press, 145–154.
Hitchen, Michael M. 1992. Talkin up: Aboriginal English in Moree. Unpublished M.Litt.
thesis. Armidale: University of New England.
Hollinsworth, David. 1998. Race and Racism in Australia. 2nd Ed. Katoomba: Social
Science Press.
Horton, David, ed. 1994. The Encyclopaedia of Aboriginal Australia, Volume 1. Canberra:
Aboriginal Studies Press.
Horvath, Barbara. 1985. Variation in Sydney English: The Sociolects of Sydney. Cambridge: Cambridge University Press.
Hudson, Joyce. 1992. “Fostering English language in Kimberley schools: an in-service
course for teachers.” In Jeff Siegel, ed. Pidgins, Creoles and Nonstandard Dialects in
Education. Melbourne: Applied Linguistics Association of Australia, 124–125.
Kaldor, Susan and Ian G. Malcolm. 1979. “The language of the school and the language of
the Western Australian Aboriginal schoolchild — implications for education”. In
Ronald M. Berndt and Catherine H. Berndt, eds. Aborigines of the West: Their Past and
Their Present. Nedlands: University of Western Australia Press, 406–437.
———. 1982. “Aboriginal English in country and remote areas.” In Eagleson, Kaldor and
Malcolm, 75–112.
———. 1991. “Aboriginal English — an overview.” In Suzanne Romaine, ed. Language in
Australia. Cambridge: Cambridge University Press, 67–83.
Kale, Joan. 1995. ”AI GAD DHA PAWA”. Practices of Orality and Literacy of an Urban
Torres Strait Islander Child. PhD thesis. James Cook University of North Queensland.
Koch, Harold. 1985. “Non-standard English in an Aboriginal land claim”. In John B. Pride,
ed. Cross-cultural Encounters: Communication and Mis-communication. Melbourne:
River Seine, 176–195.
Langton, Marcia. 1988. “Medicine Square”. In I. Keen, ed. Being Black: Aboriginal
Cultures in ‘Settled’ Australia. Canberra: Aboriginal Studies Press, 201–225.
Lefebvre, Claire. 1997. “Relexification in creole genesis: the case of demonstrative terms in
Haitian Creole”. Journal of Pidgin and Creole Languages. 12: 181–201.
———. 1998. “The field of pidgin and creole linguistics at the turn of the millennium: The
problem of the genesis and development of PCs”. Paper delivered to the SLA Symposium on Pidgin and Creole Linguistics in the 21st Century, New York.
Malcolm, Ian G. 1979. Classroom Communication and the Aboriginal Child: A Sociolinguistic Investigation in Western Australian Primary Schools. Unpublished Ph.D. thesis.
Nedlands: University of Western Australia
———. 1994. “Discourse and discourse strategies in Australian Aboriginal English.”
World Englishes 13: 289–306.
———. 1995. Language and Communication Enhancement for Two Way Education:
Report. Mount Lawley: Centre for Applied Language Research, Edith Cowan University.
———. 1996. “Observations on variability in the verb phrase in Aboriginal English”.
Australian Journal of Linguistics 16: 145–165.
——— and Marek M. Koscielecki. 1997. Aboriginality and English: Report to the Australian Research Council November 1997. Perth: Centre for Applied Language Research,
Edith Cowan University.
———, Yvonne Haig, Patricia Königsberg, Judith Rochecouste, Glenys Collard, Alison
Hill and Rosemary Cahill. 1999. Towards More User-Friendly Education for Speakers
of Aboriginal English. Mount Lawley: Centre for Applied Language and Literacy
Research, Edith Cowan University.
McKenry, Rosemary. 1995. Academic Success for Speakers of Koorie English: The Need
for Teacher Intervention. Unpublished M.Ed. thesis. Melbourne: University of
McMahon, April S. 1994. Understanding Language Change. Cambridge: Cambridge
University Press.
McRae, David. 1994. Langwij Comes to School. Carlton, Vic.: Curriculum Corporation.
Morgan, Marcyliena, ed. 1994. The Social Construction of Identity in Creole Situations.
Los Angeles: Center for Afro American Studies, U.C.L.A.
Morgan, Wendy. 1993. “Aboriginal texts for Australian classrooms”. English in Australia
104: 3–10.
Muecke, Stephen. 1981. Australian Aboriginal Narratives in English: A Study in Discourse
Analysis. Unpublished Ph.D. thesis. Nedlands: University of Western Australia.
———. 1983. “Discourse, history, fiction: language and Aboriginal history.” Australian
Journal of Cultural Studies. 1: 71–89.
Mühlhäusler, Peter. 1986. Pidgin and Creole Linguistics. Oxford: Basil Blackwell.
———. 1996. “English language and literacy within remote and isolated desert communities”. In Jean Clayton (comp.) Desert Schools Volume 2: Research Report. Canberra:
Department of Employment, Education, Training and Youth Affairs, 151–177.
——— and Rose, David. 1996. “The linguistic ecology of Anangu Pitjantjatjara communities”. In Jean Clayton (comp.). Desert Schools Volume 2: Research Report. Canberra:
Department of Employment, Education, Training and Youth Affairs, 179–213.
Readdy, Carol A. 1961. South Queensland Aboriginal English. B.A. Honours thesis.
Brisbane: University of Queensland.
Reynolds, Henry. 1996. Aboriginal Sovereignty: Reflections on Race, State and Nation. St.
Leonards, N.S.W.: Allen and Unwin.
Romaine, Suzanne. 1988. Pidgin and Creole Languages. London: Longman.
Rumsey, Alan. 1993. “Language and territoriality in Aboriginal Australia.” In Michael
Walsh and Colin Yallop, eds. Language and Culture in Aboriginal Australia. Canberra:
Aboriginal Studies Press, 191–206.
Sandefur, John. 1983. “Modern Australian Aboriginal languages: the present state of
knowledge”. English World-Wide 4: 43–68.
Sansom, Basil. 1980. The Camp at Wallaby Cross: Aboriginal Fringe Dwellers in Darwin.
Canberra: Australian Institute of Aboriginal Studies.
Sharpe, Margaret. C. 1976. The English of Alice Springs Aboriginal Children. Report to
Teachers Part 2. Alice Springs: Traeger Park School.
———. 1977. “Alice Springs Aboriginal English.” In Ed Brumby and Eric Vaszolyi, eds.
Language Problems and Aboriginal Education. Mount Lawley: Mount Lawley College
of Advanced Education, 45–50.
Siegel, Jeff. 1997. “Mixing, leveling and pidgin/creole development”. In Arthur K. Spears
and Donald Winford, eds. The Structure of Pidgins and Creoles. Amsterdam, Philadelphia: Benjamins, 111–149.
———. 1999. “Transfer constraints and substrate influence in Melanesian Pidgin”. Journal
of Pidgin and Creole Languages. 14/1:1–44.
Slattery, Sally (comp.). 1994 English Language Acquisition for Secondary Students.
Adelaide: Aboriginal Education Unit, Department of Education and Children’s Services, South Australia.
Sleep, Bronwyn M. 1996. Aboriginal Language Research of Aboriginal Students and their
Families of Crossways Lutheran School, Ceduna, South Australia. Ceduna: Crossways
Lutheran School.
Troy, Jakelin. 1990. Australian Aboriginal Contact With the English Language in New
South Wales: 1788–1845. (PLS B-103). Canberra: Department of Linguistics, Research
School of Pacific Studies, Australian National University.
———. 1993. “Language contact in early colonial New South Wales 1788 to 1791”. In
Michael Walsh and Colin Yallop, eds. Language and Culture in Aboriginal Australia.
Canberra: Aboriginal Studies Press, 33–50.
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Basil Blackwell.
Walker, Richard. 1982. The English Spoken by Aboriginal Entrants to Traeger Park
School. Canberra: Curriculum Development Centre.
Walsh, Michael. 1993. “Languages and their status in Aboriginal Australia”. In Michael
Walsh and Colin Yallop, eds. Language and Culture in Aboriginal Australia. Canberra:
Aboriginal Studies Press, 1–13.
———. 1994. “Interactional styles in the courtroom: an example from northern Australia”.
In John Gibbons, ed. Language and the Law. London: Longman.
Wilson, Greg. 1996. ‘Only Nungas Talk Nunga English’: A Preliminary Description of
Aboriginal Children’s English at Alberton, South Australia. Unpublished M.Litt Ling
thesis. Armidale: University of New England.
Ethnic varieties of Australian English1
Michael Clyne, Edina Eisikovits and Laura Tollfree
Monash University
In this chapter, we will focus our attention on ethnicity as a factor in variation
and change in Australian English (henceforth “AusE”), in particular on those
varieties that mark the ethnicity of second and later generation Australians,
which we will refer to as “ethnolects”. We will begin with a brief account of
the demography of community languages in Australia, and then focus on three
ethnolects, examining their linguistic features and the conditions for their use.
The chapter closes with a discussion of the methodological problems involved
in the study of ethnolects.
Research on ethnic varieties of AusE
It was not until fairly recently that linguists began to pay attention to the
description of ethnic varieties of AusE. One of the first studies to deliberately
include a sample of migrants and their children was Horvath’s (1985) investigation of Sydney English. She noted that migrants who acquire English in
adolescence and especially in adulthood generally retain traces of a non-native
English accent. More specifically, Horvath studied Greek and Italian migrants
and identified a variety used by first generation speakers (which she called
“Ethnic Broad”), which was based on an approximation of Broad Australian
vowels to Greek and Italian vowels. With second generation Greek-Australian
and Italian-Australian speakers she noted a tendency towards General Australian (more cultivated in the case of Greek-Australian speakers).
While Horvath’s study identifies phonological differences between first
generation and second generation migrant speakers of English in Australia,
one might confidently predict that they will be further differentiated in terms
of the extent to which their use of English is characterised by a range of
additional linguistic features: grammatical, pragmatic, discoursal, and lexical.
Demography of community languages
According to the 1996 Census, 14.6% of the Australian population use a
language other than English in the home (while for Sydney and Melbourne the
percentages are 26.4 and 25.4 respectively). Many others employ a language
other than English, not in their homes but in other places, such as the homes of
their parents and other relatives, religious institutions or community groups.
Greek is the community language most used in Melbourne homes by
school and pre-school age children (120,470 speakers). In contemporary
Australia, speakers of community languages tend to live in urban areas at least
to the same extent as the rest of the population.
A large proportion of Melbourne’s Jewish community are descendants of
Eastern European immigrants whose first language was Yiddish. Although
there has been considerable loss of Yiddish in recent decades — the 1991
Survey of the Jewish community (Goldlust 1993) revealed that only 8.8% of
Jews born in Australia spoke any Yiddish at home — for many Jewish
Melbournians, Yiddish has become a symbol of their culture and history, a
link with the religious and cultural traditions of their forebears.
In the 19th century and for part of the 20th century, there were sizeable
enclaves using a language other than English, notably the German settlements
in parts of South Australia, western Victoria, southern Queensland, and the
southern Riverina.
The term “ethnolect” is used here to denote a variety of a language that marks
speakers from groups which originally had another first language. Thus we
may recognise, for instance, a Greek ethnolect of AusE (while at the same
time excluding from this definition the Greek spoken in Australia). The
phenomenon has been described in various other ways. For example Clyne
(1968) refers to the “stabilised transference” of lexical items, features and
structures from the substratum language or variety (that is, the language
originally spoken by the community). Clyne (1981) also speaks of second
generation “foreigner talk” (that is, a register employed by the second generation to first generation users whose English may be characterised as “foreign”
or “non-mainstream”, often their parents or other older relatives). Ethnolects
of AusE which are yet to be examined in detail by linguists include Scottish,
Irish, Northern English, Sri Lankan Burger and American.
4.1 Choice or necessity
For the second generation the use of the community language is partly
determined by the parents’ competence in English. Where parents have a high
level of competence in English, the need to use the community language in the
home is not very great and the second generation Australian will choose to use
or not to use the language for reasons of solidarity, identity, and self-perceived
competence in the language. In many Australian homes at least two languages
are employed, the choice being determined by topic, function and interlocutor.
There may be bilingual communication, with parents speaking one language and the children responding in and speaking to each other in another, or
the whole family may adopt a “mixed code”. Even where parents and children
use English, the community language may be the medium of communication
with grandparents.2
A condition conducive to the development of ethnolects of AusE is a shift
to English together with a desire to express solidarity with the first generation
whose English is heavily marked as “non-native”.
4.2 In-group or out-group use
From the above it follows that the English of some second (and third) generation Australians will show variation according to whether the interlocutor(s)
and/or the situation are perceived as more “mainstream” or more “ethnic”. The
speakers might speak mainstream AusE to Anglo-Australians or to strangers
or to those whose affiliation is not clear, and the ethnolect to their parents and/
or grandparents and to their parents’ friends from the same ethnic group. In
some cases, the ethnolect is employed to their same-age friends from the same
group. This expresses their dual identity (as both Australians and as persons of
a particular ethnicity), and biculturality (mainstream Australian and ethnic
Australian), just as does the use of English and a community language and
switching between them. In some families, the ethnolect will be employed to
the parents, sometimes with switching to mainstream English, while grandparents are addressed in the community language. In other families, there is
switching between the ethnolect of English and the community language. In
still others, communication is in mainstream AusE with parents, and in the
ethnolect with grandparents.
Thus an ethnolect, like a community language, offers a means of expressing linguistic identity, of demonstrating solidarity with one’s ethnic group.
Importantly, it provides a means for those who may no longer be fluent in their
ethnic language to continue to express their identification with, and sense of
belonging to, their ethnic group. Clearly, this will vary from individual to
individual, and from group to group. Those individuals (and groups) whose
ethnic identity is important to them will show more ethnic marking in their
language than those who have chosen to assimilate within the dominant group.
Thus, we might expect a greater use of the ethnolect in groups with strong
ethnic identities (for example, Greek-Australians), than in those with high
levels of assimilation (for example, Dutch-Australians).
Of course, an individual’s feelings of ethnic identity are not static, but are
affected by the context of the interaction: who is being spoken to, and even
what is being spoken about (see Bourhis and Giles 1977). In particular, the
context of in-group communication between speakers who share ethnic identity is conducive to the use of ethnolect, while inter-group communication is
more likely to have the opposite effect.
4.3 Linguistic features of ethnolects
In the present study four types of features are dealt with: lexical (words
transferred from the community language), semantic (meanings transferred
from the community language to corresponding words in English), phonological (phonemes and realisations of phonemes transferred from the community
language to English, or phonemes distributed as in the community language,
giving rise to a “non-native accent”), and syntactic (grammatical rules transferred from the community language).
There are two further categories that are not taken up here: prosodic
(intonation patterns and lexical stress transferred from the community lan-
guage to English) and pragmatic (speech acts performed in English as they
would be performed in the community language). The reason for this is that,
while some exploratory work has been done, too little is known about these at
Some examples
5.1 Assorted families
In an early study by Clyne (1981) — referred to in Section 4 above —
ethnolects were considered to be a type of “foreigner talk”, a register used by
native speakers of English to non-native speakers, in this case by second
generation Australians to their parents and grandparents. The research was
based on recordings of spontaneous conversations in four families, two of
German background and one each of Greek and Hungarian background. In
one family, the younger generation switched between the ethnolect and mainstream AusE, in another between the ethnolect and German.
The ethnolect was phonologically marked in all the families. For example, AusE diphthongs were replaced by the nearest monophthongs (e.g. [o],
[e]), by the nearest diphthongs (e.g. [au], [ai]), or both strategies were used for
different diphthongs. In the families of German origin, final consonants such
as /z/, /v/, /b/, /d/, /g/ were devoiced.
In three families there was also syntactic marking, including the transference of adverb positions from the community language, along with such
typical features of foreigner talk as pronoun deletion (e.g. No love ‘I don’t like
it’), and auxiliary deletion (e.g. How make baklava? ‘How do you make
In one family there was lexical transference from the community language, including conjunctions (und ‘and’, oder ‘or’) and discourse markers
(e.g. na, ne) and the overuse of already in the sense of German schon.
Some of the ethnolects studied by Clyne share syntactic and phonological
features with the foreigner talk that is sometimes used to communicate with
recently arrived migrants, or with others whose command of English is
limited. The phonological features are those prevalent among bilinguals and
bidialectals; for instance, people using both AusE and a northern English
regional variety (Clyne 1977). Some of the speakers on Clyne’s tapes indi-
cated that they used an ethnolect to clients or patients of non-English-speaking
backgrounds. While the speakers themselves explain this in terms of a desire
to help migrants, it is likely that they are also motivated by considerations of
solidarity and identity.
5.2 Greek Australian English
As part of the Monash Australian English project, data were collected from 55
Year 10 students from ten schools. Each student participated in four conversations: two with a research assistant (one of which included the reading of some
words and sentences with sociolinguistic variables), and two self-recorded
(one with same-age friends and the other with family members). The conversations were based on topical issues, but included questions relating to networks and identity.
All the informants had received most of their education in Victoria. One
of the schools chosen was a Greek Orthodox day school, which seven of the
informants attended, four boys and three girls).4 The ethnolects used by these
students were phonologically marked. The features were present more in some
informants (especially two males) than in others and least of all in one of the
females. The males clearly demonstrated a strong ethnolinguistic identity.
One of the females was a focal member of a social network, of which she and
three of the male informants were members, and she was aware of switching
between a more mainstream and a more ethnic variety.
The ethnolect was strongest in the family setting, and least discernible in
the conversation with the research assistant (even though she was of Greek
descent). Two of the speakers in the sample, one male and one female, showed
very marked variation between the family conversation and the others, and at
the beginning of the family tape the male informant exhibited stronger ethnolectal features than his parents did.
The most conspicuous phonological features of the ethnolect, all of which
reflect features of Greek phonology, are given below:
variation in /u/ between [:] and [ɯ], as in pools, boots (3 male speakers,
3 female speakers), with [µ] or slightly rounded [ɯ] being preferred
consistently by 2 males and one female. (General AusE [ü]);
/ə/ advanced to [ε] or almost to [ε] in closed syllables, as in definitely,
expected, houses (3 males, 2 females)(General AusE [ə]);
an open finish in /ə/ as [ia] or [ɐ], as in here (all 7) (General AusE [ə],
[ɐ], []);
voicing of voiceless stops and affricates, e.g. intervocalic /p/ in properly,
/s/ in baseball, /t/ in Tuesday (3 males, 2 females);
heavy aspiration of /k/, as in cold, soccer (3 males, 1 female);
full voicing of consonant + consonant sequences where voicing usually
fades out in English, as in Tuesday (3 males, 3 females);
stopping of fricatives and affricates, as in the, them (2 males, 1 female).
This ethnolect is changing consistently under the influence of mainstream
AusE with respect to /ə/ and /ɔ/, tapping of /t/, final /k/ glottalisation, velarised
and vocalised /l/. The diphthongs of the Greek-Australian informants in /i/
and /aυ/ tend towards Broad Australian realisations.
A female informant, a second generation Greek-Australian from a state
secondary school in a lower middle class suburb, displayed some of these
features, but only in the family conversation and in the conversation with her
same-age friend, a girl with a different ethnic background and a more conspicuous non-native accent. For example, she used [ia] in words such as year
and here with the family and friend (but [i] in the conversations with strangers), [a] for [ə] in soccer and rounders, rounded [ɯ] for /u/ in bedrooms, more
rounded [a] in Carlton, and long monophthongal [o] in: goin’. In the family
tape, and to a lesser extent in the conversation with the friend, there was also
syntactic marking, based on either Greek or foreigner talk, as in:
Can I have money? (‘Can I have some money?’)
How you know? (‘How do you know?’)
We’ll go movies. (‘We’ll go to the movies’)
It should be mentioned that the family tape has numerous instances of switching between ethnolectal AusE and Greek (in cases where the content is
perceived as a private matter between the young person and either of her
parents, three initiated by the mother, one by the father, and one by the
daughter.) Thus mainstream AusE, the ethnolect, and Greek each make a
contribution in the expression of this informant’s multiple identity.
5.3 Yiddish Australian English
The data presented in this section derive from several sources. One was a set
of interviews conducted by Edina Eisikovits in one of the seven Jewish
schools in Melbourne. The school in question serves a Yiddish-speaking
Chassidic community and is situated in a locality of Orthodox and ultraOrthodox concentration. In New York, Fishman (1991) has found that strong
Yiddish maintenance correlates more with religious orthodoxy than with proYiddish language ideology. One of the main groups which continues to
maintain Yiddish in Melbourne is the Ultra-Orthodox sect of religious Jews
for whom Yiddish is the language of daily interaction within the family and
within the community and in particular, the language of learning holy texts.
Hence, Yiddish is the medium of instruction and learning for part of the day in
this group’s community school.
The findings reported here are based on interviews with a sample of
informants (four male and four female students from Years 8–9) from this
community school, as well as interviews with a sample from a larger Jewish
school where there was no formal exposure to Yiddish and a far less overt
identification with Jewish religion and traditions. Because of methodological
concerns with the data collected through interviews in schools (see Section 8), further data were collected through ethnographic fieldwork observations of 12 male and female informants, aged between 12 and 50 years. All
were descendants of Yiddish-speaking parents, but none used Yiddish in their
homes, and all displayed high levels of ethnic identification (for example, in
taking on positions of community and religious leadership).
Ethnolectal features were strongly in evidence in the data collected. It
was the boys at the Ultra-Orthodox community school who displayed the
strongest use of phonological features, in particular, the tendency to add a
voiced stop and [ə] to a velar nasal at the end of a words such as thing [θŋə],
and the use of uvular [] in words such as trouble [tabəl]. Both of these
features appear to be transfers of features of Yiddish pronunciation.
Grammatical features were more widespread. These included the integration of Yiddish verbs into English by dropping the (e)n of the Yiddish
infinitive, for example, schrei [ʃa] (from schreien), as in:
Ten people start to schrei (‘shout; make a fuss’) and everyone has to listen.
Such verbs may then be modified by the addition of English suffixes to form
nouns and adjectives as in:
How did you enjoy the davening (‘prayers’) today?
The Yiddish ethnolect allows ellipsis of a direct object where Standard AusE
requires a pro-form, a structure which has its source in Yiddish (as in Ich hob
‘I have’; Ich will nicht hoben ‘I don’t want to have’). For example:
You have to pay for the food now or you won’t get. (cf. Standard AusE: “You
have to pay for the food now or you won’t get any.”)
There was also some use of the structure bei + NP (from the Yiddish bei mir),
with by meaning either ‘at’ or ‘according to’ and the NP consisting of either a
pronoun or a proper noun For example:
See you then, bei Reizel’s.
There was much evidence of lexical transfers from Yiddish in the everyday
speech of Australian-born descendants of Yiddish speakers, whether or not
Yiddish had been maintained within their families. Examples include greetings such as Wie geht es? (‘How’s it going?’) and Sei gesund (‘Be well’), and
frequently used words such as bubbe [bυbə] (‘grandmother’), nush [naʃ]
(‘sweets’) and kinderlach [kindəlax] (‘children’). While these items are
clearly Germanic in origin, many ethnolectal expressions incorporate both
Hebrew and Germanic sources, for example, grosse metziya (‘big bargain’),
patsch in ponim (‘slap in the face’), a ganze megilla (‘long story; big fuss’).
The lexical tranfers were used in a variety of networks: in single sex and
in cross-gender interactions, as well as in interactions between and within
generations. They were clearly not instances of code switching in that they
occurred even among speakers who had no proficiency in Yiddish and without
any expectation that the interlocutor had proficiency in Yiddish. They occurred predominantly in in-group interactions between speakers who were
clearly identifiable as in-group members. At times, they occurred in discourse
recalling the past, in which they functioned to reinforce a sense of belonging
to a group with a shared history and traditions, common beliefs and practices.
An illustration of this is provided by the following interchange between a male
and a female informant.
My mother used to cook the way the way they did in the alte heim (‘old
country’). You know, lokshen and yoach (‘chicken noodle soup’),
kneidlach (‘matzo balls’) …
Schmalz (‘chicken fat’) and grieven (‘fried chicken skin’) …
5.4 Ethnicity, religion or region
Part of the AusE study deals with three generations of families in the Western
District of Victoria. In the Tarrington-Tabor-Warrayure area, near Hamilton,
about 300 km west of Melbourne, we have recordings of the German and
English of most of the remaining bilinguals in the district in the late 1960s and
early 1970s. We were interested to ascertain how far the presence of the
distinctive features (phonetic, lexical/semantic, syntactic) continued in the
local variety of the later generations. At the time of writing, the Koroit data
was in the process of being analysed so that we shall devote our attention here
to the Tarrington-Tabor-Warrayure data.
Like other German-speaking enclaves, such as those in the Barossa
Valley, the Adelaide Hills, Murray Flats, the Wimmera, Southern Riverina,
and Southern Queensland, those in the Western District use German as the
language of home and farm, community and church, and until 1916, as a
medium of instruction in the local (bilingual) school. The First World War
brought about a gradual shift to English, first in the more public domains
(especially the church, which ceased German services altogether in 1939),
then in the more private ones.
The ethnolect of English of the bilingual generation (those taped in the
60s and 70s, a few of whom are still alive, aged 90 plus) was affected by the
substratum language (the earlier first language), German. Among the distinctive phonetic features were:
Final devoicing of /d/, /z/, /b/, /g/, /ŋg/ and /d/, e.g. in is, bad, bag, railing,
/e/ realised as [ε] (lower than in AusE of the time), e.g. in breakfast.
[ø] for //, e.g. in girls, and [ε] for /æ/ e.g. in answer.
Monophthongs [e], [] and [o], e.g. in lady, seem, suppose.
Short [υ] in [lυθərən] Lutheran.
Short [a] in [pasta] pastor.
Few reduced vowels, e.g. [inɔsεnt] innocent.
Overgeneralisation of the accented form of the definite article in unaccented
position, e.g. You ask the [ði] questions.
Free variation between [ð] and [d] in initial position, e.g. [dεn] for [ðεn] then.
Distinctive semantic features include:
Yet for still (Ger. noch), e.g. We have some German books yet.
Overuse of already (Ger. schon), e.g. They brought them on Friday already.
Different meaning ‘several’ (Ger. verschieden has both meanings), e.g. This
has happened different times.5
The use of bring with, come with, take with, based on Ger. mitbringen,
mitkommen, mitnehmen (‘bring along’, ‘come along’, ‘take along’).
Distinctive syntactic features include:
The relative “to X” construction, e.g. He’s a brother to Colin Linke (Standard
AusE “a brother of C.L.”, Colloquial Ger. Er ist dem C.L. sein Bruder.).
The addition of the definite article with the names of languages, e.g. We
always spoke the German at home.
Comments made at the time of the recordings provide considerable evidence
that people from German enclaves were recognised as such by their English
and in some districts nicknamed the “yet alreadys”.
The grandparent generation (70–80 year olds) — the first in our recent
sample — whose home language as children was German but who rarely
speak it today, except for ‘devilment’ still show some German features in their
English. The devoicing of /z/, e.g. in times, is very common. Less common are
the following phonetic features:
[e] for /æ/ and /ε/, e.g. in have, yes.
Variation between [əυ] and [o], no, only.
[ø] for //, e.g. in girls.
The following sentences illustrate various semantic features:
There’s a Commonwealth Bank agency at the Post Office yet.
We’ve made different friends.
We stopped the German when my father died, before then already.
Well I just brought them with.
The following usages are more idiosyncratic:
Heard the news last week that our bank closes (Ger. simple present for Eng.
progressive or future)
They offered to borrow it to me (Ger. leihen can mean either ‘borrow’ or
By the middle generation (approximate age 50s), the speakers could be
characterised as either Broad or General speakers.6 However some features of
the old ethnolect persist, including:
Short [] in Lutheran, short [a] and final [ə] in Pastor.
The relative “to X” construction, e.g. I am a son to Andrew and Mary,
sometimes in variation with “I am X’s son”.
We’re looking for Dolph Schulz yet. (yet meaning ‘still’)
They brought some German shepherds with. (with for along)
These features no longer occur in the youngest generation (now children or
adolescents).The one exception in our data is:
I am a neighbour to Jill and Kate.
Those that attend Lutheran schools tend more towards Cultivated Australian,
those at state schools more towards General Australian.
The variety whose distinctiveness we have described started off as the
ethnolect of a bilingual settlement with phonetic, semantic and syntactic
features. Within thirty years, the variables seem to have been reduced to final
devoicing and a few markers related to church, family or neighbourhood.
They can now be seen as markers of denomination (Lutheran) or region (this
area of the Western District) rather than ethnicity (German). Most of the
families now see German as representative of their ancestry rather than their
ethnicity. By the youngest generation, there are no more distinctive features.
What this suggests is that even strong geographical concentration and membership of a distinctive religious group may not be sufficient for the permanent
maintenance of an ethnolect after the generations which spoke the community
language have passed away.
Multicultural varieties
Warren (1999) discusses a stylised variety employed by some young Australians of second generation Greek, Italian, Turkish and ‘Yugoslav’ background. She uses as her data interviews with Australian actors of the second
generation who have made Australians from Mediterranean cultures the focus
of their shows, as well as interviews with young second-generation schoolchildren. Her claim is that a type of ethnolect, which some young people of the
second-generation have designated “wogspeak” (wog being a derogatory term
which they have reclaimed), enables some “ethnic” youth to differentiate
themselves from both their parents’ values and those of the “Anglo host
culture”. According to Warren, the actors provide anecdotal evidence of this
“pan-ethnic” variety of AusE which, they argue, acts as a badge of identity.
From interviews with young second-generation schoolchildren, Warren identifies phonetic features such as [a] in final syllables of words such as ‘pleasure’, the replacement of ‘th’ by [d], and the avoidance of reduced vowels, as
well as some grammatical features such as double negatives. The variety is
used by some members of the second generation to fulfil particular functions,
and draws on both the migrant varieties of the first generation and their nonstandard English.
It has been found in another study (see Section 5.1) that some second
generation Australians employ their particular ethnolect to people of other
ethnic backgrounds. This convergence provides a basis for a stylised variety
for dramatic purposes. The incidence of “pan-ethnic” varieties, however,
would depend on networks in which second generation Australians of different backgrounds require a common variety to express their shared identities.7
Influence on national language
The ethnolects of AusE are a rich source of new vocabulary for the mainstream varieties as the eating and drinking habits have become a tangible
indicator of multiculturalism. It is therefore not surprising that many new
lexical items in AusE are in the food domain, including cappuccino, goulash,
schnitzel, souvlaki, cevapcici, humus, pita, and chop suey. There are also
compromise forms such as liver sausage, apple strudel.
There is evidence of earlier ethnolects influencing AusE beyond the food
domain. Some examples include bonzer (from Spanish bonzana) and shicker
(from Yiddish drunken). It will be interesting to see how widespread ethnolectal influences on mainstream AusE are in the future.
Methodology: individual variation and problems of quantification
The study of ethnolects requires considerable modification of the methodology
traditionally employed in sociolinguistic investigations. While all such studies
need to pay attention to the Observer’s Paradox (Labov 1970) and to seek ways
to overcome it, this is especially true for ethnolectal research. This is because
ethnolects are primarily an in-group phenomenon and hence are unlikely to be
used in out-group interaction (a feature of much formal interviewing).
The first requirement for eliciting ethnolectal data is that the interviewer
be an in-group member of the informant’s ethnic group. A speaker is obviously not going to use markers of group solidarity with an interlocutor who is
not part of his/her group. However, even when the interviewer and informant
are part of the same group, formal interviewing techniques are unreliable as a
means of eliciting data representative of the range and uses of the ethnolect.
The interviewer’s ethnicity may not always be immediately apparent to an
informant, as our data occasionally suggests. Individuals may share a common
ethnic background but have different degrees of identification with it, or they
may feel more or less secure in making this identification known to a stranger,
even one with a similar background.
Thus, even when the interviewer and informant had a shared ethnicity,
this was not always sufficient to overcome the formal constraints of the
interview context. This was especially true for the male informants from the
Ultra-Orthodox Jewish school who, because of the strict separation of males
and females in this community, especially during adolescence, were unused to
interacting with females and hence found it extremely difficult to engage in
informal conversation.
Moreover, the interview situation itself may elicit a more formal mainstream style of speech, rather than the informal ethnolectal variety associated
with intimates with shared ethnicity. This is especially true when the interviewer and informant share other aspects of their identities, such as gender and
age, as well as a shared Australian ethnicity.
Spontaneous conversation between self-selected groups of informants is
therefore a more appropriate source of ethnolectal data than traditional interviewing. Such an approach allows us to obtain a first-hand picture of a
community’s use of ethnolect in its everyday context. In addition, it allows us
to relate language behaviour to community values and attitudes. Our work
with second and third generation descendants of Yiddish speakers in
Melbourne suggests there may be several networks within this group whose
different values and attitudes are reflected in varied ethnolectal usage.
Two separate approaches were employed in our work to elicit such data.
Firstly, informants selected (and interviewed) through schools were asked to
provide self-selected recordings of informal family interactions and selfselected peer interactions. The data derived from this approach was variable in
quality. Some informants used the suggested guidelines provided by the
researcher as a formal interviewing schedule, thereby undermining the goal of
eliciting spontaneous interaction.
The second approach was ethnographic: an in-group member observed
the spontaneous interaction of group members in a number of contexts. While
this provided an excellent source of naturalistic data, it was time-consuming
(and of course depended on the availability of such an in-group observer). It
was also at times difficult to tape record data, and hand-recorded data was
clearly less accurate and less amenable to certain kinds of analysis (for
example, phonological analysis).
We are indebted to the Monash Research Fund for financial assistance for the project
Dimensions of Australian English. Our thanks are also due to Peter Collins, Mark
Newbrook, Jane Curtain, Sandra Kipp and Maria Papazoglou for their input and assistance.
There is a rich international and Australian literature on this. See Clyne (1991) for
Australian references, and Auer (1998) for a recent treatment of the indexical function of
alternation between languages.
Studies by Wierzbicka (1985, 1997) and Clyne (1979, 1994) suggest that pragmatic rules,
including those dictating the selection of forms of address and the performance of certain
speech acts, are vehicles of deep cultural values. Clyne (1994) finds that where English is
employed as a lingua franca in Australian workplaces there is substantial variation in the
turn-taking behaviour of first generation Australians from different parts of the world:
Central Europeans and South and West Asians tend to take longer turns and increase
speed and volume to maintain turns, while South-East Asians tend to take shorter turns
and maintain them by slowing down.
Melbourne has two Greek Orthodox schools and one Greek community day school.
However, the vast majority of Greek-Australian youth attend state, Catholic or mainstream independent (Anglican or Uniting Church) schools.
A similar usage has coincidentally become common among young people in Melbourne.
For more on Broad and General, see Cox and Palethorpe, and Tollfree, in this volume.
For examples from multicultural settings in other countries, see Kotsinas (1992), and
Rampton (1995, 1998).
Auer, Peter, ed. 1998. Code-switching in Conversations. London, Routledge.
Bourhis, Richard and Howard Giles. 1977. “The language of intergroup distinctiveness”. In
Howard Giles, ed. Language, Ethnicity and Intergroup Relations. London: Academic
Press, 119–135.
Clyne, Michael. 1968. “Deutscher Idiolekt und deutscher Dialekt in einer zweisprachigen
Siedlung in Australien”. Wirkendes Wort 18: 84–95.
———. 1977. “Multilingualism and pidginization in Australian industry”. Ethnic Studies
1: 40–55.
———. 1979. “Communicative competences in contact”. ITL Review of Applied Linguistics 43: 17–38.
———. 1981. “2nd generation’ foreigner talk in Australia”. International Journal of the
Sociology of Language 28: 69–80.
———. 1991. Community Languages: The Australian Experience. Cambridge: Cambridge
University Press.
———. 1994. Inter-Cultural Communication at Work. Cambridge: Cambridge University
Fishman, Joshua. 1991. Reversing Language Shift. Clevedon: Multilingual Matters.
Goldlust, John. 1973. 1971 Survey of Melbourne Jewish Community. Melbourne: Jewish
Welfare Society Inc.
Horvath, Barbara. 1985. Variation in Sydney English: The Sociolects of Sydney. Cambridge: Cambridge University Press.
Kotsinas, Ulla-Britt. 1981. “Immigrant adolescents’ Swedish in multicultural areas”. In
Cecilia Palmgren, Karin Lövgren and Göran Bolin (eds.), Ethnicity in Youth Culture.
Stockholm: Youth Culture at Stockholm University, 43–62.
Labov, William. 1970. “The study of language with social context”. Studium Generale 23:
Rampton, Ben. 1995. Crossing: Language and Ethnicity among Adolescents. London:
———. 1998. “Language crossing and the redefinition of reality”. In Auer, ed. 1998, 290–
Warren, Jane. 1999. “‘Wogspeak’: transformations of Australian English”. Journal of
Australian Studies. 62: 86–94.
Wells, John. 1982. Accents of English 1: An Introduction. Cambridge: Cambridge University Press.
Wierzbicka, Anna. 1985. “The double life of a bilingual”. In Roland Sussex and Jerzy
Zubrzycki, eds. Polish People and Culture in Australia. Canberra: Department of
Demography (Australian National University), 187–223.
———. 1997. Understanding Cultures Through Their Keywords. New York: Oxford
University Press.
Australian English
and recent migrant groups
Scott Fabius Kiesling
University of Sydney
Migration has long been seen as an important factor affecting language
change. Two main consequences of migration have been proposed. The first
was shown by Labov in his (1963) Martha’s Vineyard study. He suggested
that the sharply increasing number of summer people owning and occupying
property on the island was encroaching on the traditional lifestyle and identity
of the year-round islanders, particularly the fishermen. In response to this
influx of outsiders, the islanders centralised the diphthongal nuclei in /a/ as in
PRICE and /aυ/ as in MOUTH to symbolise their connection with the island.
Migration motivated the “native” group to a linguistic change away from the
norms of the immigrating group. Labov has posited that a similar process has
happened in Philadelphia, which saw a steady migration of African Americans from the south in the early and middle parts of the century. Labov
suggests that this migration started the many changes he has documented in
the European American population in the last 20 years (see Labov 1994).
The other possibility for the effect of migration on language is homogenisation (see Chambers 1995: 58ff, Kerswill and Williams 1992), in which
disparate groups migrate to the same place to form a new community, with no
one previously established group determining the original norms. The second
generation of these migrants levels the differences in the dialects of their
parents. This process has been proposed for the formation of Australian and
North American Englishes. Kerswill and Williams (1992) investigate variation in the “new town” of Milton Keynes in southern England, which was
settled by speakers from different British English dialect areas. They find that
in acquiring a new dialect, the children of the migrants “split the difference” of
these dialects (including RP), exhibiting far less diversity than their parents.
This paper discusses the effects of migration on language variation and
change in Sydney, focusing on how variation in two vowels is embedded in an
area of “multiethnic” western Sydney. Migration has changed the face of the
city in the past forty years, and continues to change it today (even though the
largest migrant population is from the UK). In fact, Australia relies on migrants to keep its population and labour force growing. But in the last thirty
years or so a marked change in migration has occurred; as recently as 1972,
there was in place a policy known as the “White Australia” policy, in which
migration from Asia and the Middle East was all but prohibited (although
many Asian migrants did arrive during and before this policy was implemented, particularly in times when hard labour was needed, as in the building
of railroads). During this time, non-Anglo-Celtic migrants came mostly from
the Mediterranean countries and Eastern Europe, especially Greece, Italy, and
the Balkans. After the White Australia policy ended, one of the first major
waves of migration from Asia was made up of refugees from the Vietnam war;
the Lebanese war in 1982 also prompted a large number of migrants from that
country. There are now significant migrant populations from all parts of the
world, but migration from Asia and the Middle East has had the most striking
impact on Australian society in the last twenty years.
In the 1980s the government had an explicit policy of “multiculturalism”,
which, in contrast to the assimilationist pre-1970 era, tried to affirm and
celebrate the new cultures of Australia. However, there does seem to have
been some underlying anti-immigrant feeling, as evidenced by the meteoric
rise of Pauline Hanson after her maiden speech in Parliament in 1996, in
which she suggested, among other things, that Australia’s culture was being
lost because of Asian migrants. Migration remains one of the most sensitive
topics in Australian politics.
Because the migration situation in Australia is socially significant it is
one in which we might expect linguistic change to flourish. Previous studies
would lead us to predict several possible outcomes of this change. One
possibility is simply that homogenisation will occur as second-generation
migrants learn Australian English in the same way as any other Australians. A
second possibility is the growth of a binary linguistic differentiation between
Anglos and non-Anglos. Finally, a more complex situation may develop as
some migrant groups homogenise (but perhaps form a number of different
varieties in the process), and Anglos become differentiated from these “migrant” varieties. I will argue that a variant of the latter is taking place. My data
suggest that there is no single variety typical of non-Anglo speakers, groups of
whom are differentiated linguistically. Greek Australians in particular seem to
be differentiating themselves from the other groups. The data is based on
speech from sociolinguistic interviews with 12 second generation, young (14–
29 year old), working-class Anglo, Greek, Italian, Lebanese and Vietnamese
men and women. The variables analysed were the diphthongs (ow) as in
GOAT and (ay) as in PRICE. Both of these variables are involved in a
complex change that involves shifts in nucleus, glide, and trajectory. Most
differentiation was found to occur with the women of Greek, Italian, and
Lebanese backgrounds; there is thus an interaction between sex and ethnicity,
with women using these variables to differentiate themselves more than the
men. This fact suggests that there is more change to come, given the important
role that women generally play in sound change.
Migrant groups in Auburn/Fairfield/Bankstown
The interviews were conducted in a multiethnic region of Sydney, the band of
suburbs bounded by Auburn in the east and Fairfield in the West, dipping into
the Bankstown area. Migration to the Auburn/Fairfield area in the past decade
or so has been extensive. The 1996 census reports that 53.5% of the population of Fairfield was born overseas, and 51.5% of that in Auburn. In addition,
in Fairfield, 64.1% spoke languages other than English at home, and 62.6% in
Auburn. The majority of the overseas-born population in these suburbs are
thus from non-English speaking countries.
I have focused on native speakers of Australian English (henceforth
“AusE”) from five ethnic groups: Anglo, Greek, Italian, Lebanese, and Vietnamese. All of the non-Anglo speakers are second-generation Australians
(first native-born generation). I have deliberately chosen migrant groups
which became established at different times. For instance the first major wave
of Greek and Italian migration occurred much earlier than that of the Vietnamese and Lebanese.
As many informants indicated in their interviews, Auburn and the surrounding suburbs were working-class Anglo-dominated communities until
the 1970s. Population changes began in the late 70s and early 80s with an
influx of Lebanese, Turkish, and Vietnamese migrants. Many Anglos in the
area, as well as members of the more “established” ethnic groups (the Italians
and Greeks), have asserted during the interviews that living standards and the
general image of the place have taken a general downturn since the recent
wave of migration. Some talk of the newer migrants turning Auburn into a
“Persian bazaar” full of junk shops, food shops, and kebab shops. They also
claim that local schools that once used to have solid academic standards are
turning into poor quality schools, where the focus is on teaching the immigrant
children English. There is a strong sense amongst many Anglos that their
children are less well catered for, and that schools are geared towards the
“new” migrants. The newer groups, however, have a more positive outlook
and generally see the area as a good one in which to live. It is clear, then, that
immigration is an issue for the people who live in these contiguous suburbs.
This makes them an ideal place to investigate how the second generation of
migrants learn AusE, and to test whether they use variation as an ethnic
identity marker, and to explore whether members of the local Anglo population are using linguistic resources to differentiate themselves from the migrant
Variation and change in Sydney English
Horvath (1985), in her survey of English variation in Sydney, found a pattern
that suggested a general change from varieties characterised as Broad (vernacular) and Cultivated (prestigious) towards the General (in between) variety, following Mitchell (1946) and Mitchell and Delbridge (1965). This
change thus reflects both a “natural” change (from Cultivated to General) and
a reversal of that change (from Broad to General). The shift from the General
to Broad variety may be considered natural because it follows Labov’s (1994)
Principle IIa (in which the nuclei of upgliding diphthongs fall), whereas the
reversal runs counter to this principle. This kind of reversal is not unheard of,
but has not been found in the context of a corresponding change following
Labovian principles.
These opposed trends seem to suggest that some kind of homogenising
process is at work (moving everyone to the general variety), even though
groups are coming from different language backgrounds, and there is an
established vernacular continuum in place. But Horvath also found that, in
addition to sex and class, the most important social factor affecting interspeaker variation in Sydney was ethnicity. She interviewed members of three
ethnic groups (Anglo, Greek, and Italian), and found that the direction of
change was different for Anglos than for Greeks and Italians, with Anglos not
moving away from Broad as much as the Greeks and Italians, who were
leading the reversal to the General variety.
Figure 1 shows this bulging of the middle two sociolects for the Greek
and Italian teenagers, with less of a bulge for Anglo teenagers, and a much
more even distribution across sociolects 1 to 3 for the Anglo adults (there were
no data for adult native-speaker Greek and Italian Australians). There is strong
evidence from Horvath’s data that some kind of ethnicity-influenced change is
occurring in AusE, although whether this change involves homogenisation or
differentiation, and whether it is a change in compliances with or in conflict
with Labovian principles, is unclear from her data. One notable finding by
Horvath, which may suggest that it is Greek speakers in particular who are
creating the significant effects for ethnicity, is that while the Anglo and Italian
teenagers are mostly in sociolect 2, the Greek teenagers are mostly in sociolect
% of population
Anglo A dults
Anglo Teens
Italian Teens
Greek Teens
Ethnic/age group
Sociolect 1 (Broad )
Sociolect 2 (General)
Sociolect 3 (General)
Sociolect 4 (Cultivated)
Figure 1. Speakers from different ethnic age groups classified according to sociolect
(adapted from Horvath 1985: 93)
The sample comprised speakers who live in a fairly small area of Sydney as
described above. Two research assistants from the area were hired to perform
interviews, and their networks and extended networks were used to recruit
informants. One researcher was Greek, the other Vietnamese; both were
The sample was stratified for five ethnicities, including two groups that
have established migrant communities (Greek and Italian), two groups that
have had large migrations only in the past twenty years (Lebanese and
Vietnamese), and speakers of Anglo or Irish background (labelled simply
“Anglo”). The sample was also stratified in a crude way for age (over 30 and
under 30), class (lower working class, upper working class, middle class), and
sex. The data reported here is based on an analysis of the younger working
class speakers for all ethnicities and both sexes (although not Vietnamese
males, who have proved problematic to gain interviews with). The current
data are from a single speaker in most sex/ethnicity categories; however, there
are three Greek males and two Greek females.
Each informant was asked to count to twenty at the start of the interview,
some demographic information was requested, and then the interview progressed following a modified Labovian module format (see Labov 1984).
Samples of the variables under discussion were analysed through spectrographic analysis; vowels in the count at the beginning of the tape were also
measured for use as reference points for normalisation.
The variables
The vowels under consideration are both variable on a number of dimensions
in AusE. Each has a “Broad-General-Cultivated” (hereafter “BGC”) differentiation, with Cultivated being closer to the British RP variant. The upgliding
diphthong /oυ/ in AusE is generally characterised by a lower target than in
other English varieties; in other words, the broader the variant, the lower the
nucleus. In Bernard’s (1970) and Cox’s (1996, and in this volume) analysis of
Australian vowels in H_D frames, the glide target is slightly fronted, and the
trajectory of the diphthong moves first directly upward and then forward.
Anecdotally, this fronting trajectory has become more pronounced, and has
been remarked on by visitors to the country who are not phonetically trained
(in addition to informally by a number of linguists). There are therefore a
number of dimensions on which this vowel may vary: direction of trajectory in
addition to the height and backness of both nucleus and glide. With this fact in
mind, a third measurement in addition to the nucleus and glide targets was
taken as a rough measurement of trajectory. Harrington, Cox and Evans
(1997), however, find that in open syllable contexts the trajectory rises and
then bends back, and suggest that the fronting is due to coarticulation effects
from the final coronal in the H_D frame.
(ay), along with the vowel /e/ as in MATE, is a marker of Australian
identity and, within Australia, class. Like many English varieties, including
Martha’s Vineyard (see Labov 1963), Okracoke (Schilling-Estes 1997) and
London (Wells 1982: 308), the vernacular variant has a higher and more
retracted nucleus target than the standard variant, sometimes with rounding as
well. However, (ay) also shows signs of recent change. This change seems to
be affecting more prestigious speakers such as newsreaders, and also middleand upper-class women. This newer variant shows what sounds like a monophthongisation similar to that in the Southern shift pattern described by Labov
(1994). However, the resulting monophthong is impressionistically higher
than in the American south. This difference may be due to the fact that what is
happening is simply a falling of the glide target without a similar backing —
Harrington, Cox and Evans (1997) found a glide target in the /ε/, rather than
//, range, suggesting that this lowering is indeed under way.
6.1 (ow)
A first MANOVA run on normalised measurements for F1 and F2 of nucleus
and glide, and ethnicity and sex, showed a significant effect for the interaction
between sex and ethnicity for all measures. However, on inspection it became
clear that one Greek woman could be responsible for most of this effect.
(Vietnamese speakers were not included in MANOVAs for both sexes because there were no data for male Vietnamese.) I will discuss her more fully
below, but initially I removed her from the model because of her clear outlier
status. The next run showed significant results for interactions of sex and
Figure 2. (ow) nucleus F1, mean log normalised values by sex and ethnicity
ethnicity on the height dimension of all three measures (nucleus, glide, and
trajectory). The most surprising “non-effect” is for F2 of the glide, since the
newer variant seems impressionistically to have a fronted glide. This result
suggests that if this change is taking place, then it is not originating in the
working class (the speakers in the current sample).
As Figure 2 shows, for the nucleus, men were generally higher (lower F1
values) than women, except for the Anglo man. For women, there is more
difference for ethnicity, with the Anglo woman exhibiting the lowest nucleus
and the Greek and Italian women the highest.
Figure 3 shows the height of the glide. Here the most striking feature is
the Italian split between the man and the woman, and to a lesser extent the
Greek woman. There is relative uniformity for the Lebanese and Anglo
women. For the men, there is more uniformity, with the Anglo highest,
followed by the Greeks, and then a gap to Lebanese and Italian.
Figure 4 shows the height of the trajectory point, which shows
essentially the same pattern. The fact that it is the same pattern indicates that in
terms of height the trajectories for speakers are quite similar.
Figure 3. (ow) glide F1, mean log normalised values by sex and ethnicity
Figure 4. (ow) trajectory F1, mean log normalised values by sex and ethnicity
Figure 5.(ow) nucleus F1, mean log normalised values by sex and ethnicity, including
“outlier” Greek female
These results all have factored out one of the Greek females. However,
outliers are often outliers for a good reason; what happens when she is
included in the data? Figure 5 shows the results for the nucleus F1. Here, it is
clear that this speaker is simply exaggerating the differences present in previous analyses.
There are thus several sociolinguistic patterns to explain. One is the
relatively lower (more vernacular) heights for the men. Another is the interaction between sex and ethnicity, in which the Greek and Italian women show
more standard values than the Lebanese and Anglo. We can view this interaction in two ways. First, we might say that women are more differentiated by
ethnicity. Alternately, we could look at the behaviour within ethnic groups,
and note that for Greek and Italian speakers, gender makes more of a difference than for Lebanese and Anglos (and a distinct kind of difference when
compared to Anglos). How we look at the pattern will depend on the results
for (ay).
6.2 (ay)
In the MANOVA run for this variable, both F1 and F2 values for the nucleus
were significant (p < .01), as was F1 of the glide (no trajectory value was
coded). The interaction between sex and ethnicity is clear from Figures 6–8, in
which differences for the Greeks are always the most pronounced.
Figure 6. (ay) nucleus F1, mean log normalised values by sex and ethnicity, including
“outlier” Greek female
Figure 7. (ay) nucleus F2, mean log normalised values by sex and ethnicity, including
“outlier” Greek female
Figure 8. (ay) glide F1, mean log normalised values by sex and ethnicity, including
“outlier” Greek female
For this variable the same Greek female appears to be an “outlier”.
Without her, there was still a significant (p < .01) overall interaction for sex
and ethnicity. Figures 9–11 show these values and the values for the Vietnamese woman, who is similar to the Anglo and Lebanese women for the nucleus,
but then closer to the Greek and Italian women for the glide and trajectory,
indicating that her trajectory is unique among the speakers.
So the “outlier” does indeed seem to be exaggerating tendencies already
exhibited by the other Greek woman, and by the Italian woman. The general
pattern that emerges is similar to (ow) in that there is an interaction between
sex and ethnicity, with this interaction showing a greater differentiation between women than between men. Greek women again are the among the most
differentiated from their male counterparts, but here it is the Italian women
who pattern with the Greek women rather than the Lebanese women. Again,
the men do not greatly differentiate themselves. Perhaps most significant is the
fact that the interaction in this case has the Greek and Italian women showing
a more vernacular variant than the men.
Figure 9. (ay) nucleus F1, mean log normalised values by sex and ethnicity, excluding
“outlier” Greek female
Figure 10. (ay) nucleus F2, mean log normalised values by sex and ethnicity, excluding
“outlier” Greek female
Figure 11. (ay) glide F1, mean log normalised values by sex and ethnicity, excluding
“outlier” Greek female
7.1 Summary of patterns
The following clear patterns have been found:
There is a consistently significant interaction between sex and ethnicity
for both variables.
Greek women are the most differentiated group in this interaction.
For (ow), Lebanese women are differentiated along with Greek women
from their male counterparts towards the standard.
For (ay), Italian women are differentiated along with Greek women from
their male counterparts toward the vernacular.
Men make much less differentiation by ethnicity than women.
Anglos make very little differentiation by sex.
The clearest conclusion from these data is that there is certainly not a process
of homogenisation going on for the women, although the pattern for men is less
clear. The question remains, however, whether the women are differentiating
themselves in contrast to other women or in contrast to men. In addition, why
are they behaving differently with respect to the vernacular in the two variables?
7.2 Possible “confounding” factors
Before I propose any explanations in terms of the sex/ethnicity categories, let
me first consider some possible “confounding” factors (that is, any factors that
skew the comparability of speakers in the different groups, including: internal
constraints on variability, interviewer effects, style differences). In separate
MANOVA runs, there was a significant effect (p < .05) for the following
environment. Because of the low numbers, a full model integrating all effects
was not possible. However, speakers were coded so as to have a balance of
following environments; the numbers for each speaker were low enough to
inspect manually, and there do not seem to have been any significant differences in following environment for speakers who were significantly different
from most of the other speakers. The effect of the interviewer and style are
similar concerns, in that speakers may not have been taking similar footings in
the interview. This was clearly not a problem for the tokens of (ay) taken from
the counting at the beginning of the interview. As for the rest of the interview,
every attempt was made to find as relaxed and unconscious a part of the
interview as possible; more specific footings would need not only close
discourse analysis of the interviews, but also more tokens than can be extracted from the interviews (especially if internal factors are to be controlled
for as well, since a specific footing is often unique and fleeting and will not
contain a variable in all linguistic environments). Interviewer influence does
not appear to have been significant, because there was no consistent difference, for example, between the Greek speakers who were interviewed by each
interviewer. However, the fact that both interviewers were women could
explain, at least in part, why differences among ethnicities showed up more
for women than for men. Perhaps with a woman interviewer the men focused
more on being “masculine” through their language and less on constructing a
specific ethnic identity, while the women felt less constrained to fit into an
ideal “feminine” identity. But we should not think of this fact as “confounding” the data; rather, it is one of the possible explanations that may be
Thus, there is little evidence that these factors are “confounding” the data,
which suggests that we should look to social factors to explain the differences,
particularly the social lives of the different ethnic groups.
7.3 Explanations
I will consider four interrelated sources of explanation: speaker networks,
ethnolinguistic vitality, gender differences in migrant families, and Anglo
symbolic separation.
Networks. A network explanation, following Milroy (1980), might suggest
that those speakers with denser and more multiplex networks would be more
likely to exhibit vernacular variants. Unfortunately, no formal network measure was undertaken with these speakers, so I cannot comment on this type of
explanation. Alternatively, following Eckert (1989), it could be suggested that
speakers who behave similarly with respect to variation are likely to be in the
same networks. In the study, however, no identifiable correlation was found
between the speakers’ friendship groups and their linguistic behaviour. One of
the Greek women lived in a different part of Fairfield from other speakers, but
went to school with many of them. One of the Greek men had an Anglo parent,
which might account for his similarities to Anglo men (a mild tendency found
with a number of the Greek men). While networks may have played some part
in the results, a more significant factor is the different attitudes in migrant
families towards girls’ and boys’ socialising.
Gender differences in migrant families. Strong gender differences are reflected in many of the stories told in the interviews by women of migrant
background about their lack of freedom growing up. It is very common for
daughters in migrant families to be restricted in their socialising, not allowed
out in the evenings or on weekends, or — if they are — accompanied by
chaperones. This topic was in fact a fruitful one for the interviewers to exploit,
as they could claim common ground with the women on this point. Some of
the restrictions, it seems, were prompted not so much by fear for the girl’s
safety or “chastity”, but rather by fear that a daughter might meet a potential
partner from outside the ethnic group. This situation leads us back to a
network explanation, to the extent that the migrant women’s networks are
more likely to have been restricted to school friends, and this would make their
networks denser and more multiplex. These networks might then provide an
explanation for the fact that the migrant women tended toward the vernacular
for (ay), and for the fact that there was more difference for sex between nonAnglos and Anglos.
Ethnolinguistic vitality. “Ethnolinguistic vitality” is the term used by Giles et
al. (1977) for the ability of migrant groups to maintain their ethnic language. It
has in fact been cited as a factor for the ability of the Greeks, and particularly
Greek women, in Australia to preserve their language longer than other
migrant groups. As Tsolidis (1995: 128) explains:
It is this comfort Greek-Australians have with being unselfconsciously Greek
while not necessarily wanting to live in Greece, juxtaposed with outsiders’
views of their national chauvinism and their cultural and linguistic loyalties,
that in many ways characterises what it is to be Greek in Australia.
Certainly, in the present study, it was found that the Greeks were the most
comfortable of the various migrant groups with their dual nationality.
Anglo symbolic separation. This is the hypothesis that, as in other situations
when migrants arrive, the established population moves away from the migrants’ ways of speaking to assert their “nativeness”. The kinds of patterns
observed in the study do not support the notion that this process is happening
here. We should take a clue from the fact that the men and women in the ethnic
groups are differentiating to varying degrees; this kind of differentiation is
more likely to reflect a situation of linguistic change. So if the Anglos were
moving away from speakers of other ethnicities, we would expect to see sex
differentiation in them rather than in other ethnicities (this fact does not of
course necessarily entail that the recent migrant groups are participating in a
change, only that it is unlikely that the Anglos are not doing so).
The most plausible explanation is likely to be one that invokes networks
and gender in migrant families and ethnolinguistic vitality and the speaker’s
orientation to the interviewer. The first three provide a powerful motivation
for the pattern of consistent and strong differentiation by the Greek women.
Much, however, relies on the kinds of identity meanings the speakers attach to
the variants we have discussed. A more complete explanation would require
careful analysis of style shifting in the interviews and conversations, and
further experimentation. I would speculate that, for (ay), the men are shifting
to the standard because they are speaking to women, whom they may perceive
as speaking more standardly. The women, on the other hand, may be taking
quite different stances, perhaps showing their affinity with the migrant interviewers through their use of the vernacular, or even attempting to show how
Australian they are. For (ow), there is less metapragmatic awareness, and this
may also reflect a change in progress. The two variables are clearly used
differently by the Italian and Lebanese women.
The explanations advanced here are still speculative; a qualitative analysis will be needed to yield a richer and deeper interpretation. The important
clear finding is that there is certainly less differentiation than in many other
migrant contexts: although there does not seem to be a homogenisation
process at work, neither does there seem to be a specific ethnic sociolect.
Moreover, the Anglo speakers do not seem to be moving away from a
perceived ethnic sociolect. Most variation seems to be in particular ethnic
groups, especially among women. Perhaps this fact simply indicates that
many Anglos have moved to other suburbs in Sydney. But I would hope that,
on the contrary, it suggests that multiculturalism is alive and well in Auburn
and Fairfield.
Bernard, John. 1970. “Towards the acoustic specification of Australian English”. Zeitschrift für Phonetik 2/3:113–28.
Chambers, Jack. 1995. Sociolinguistic Theory. Oxford: Blackwell.
Cox, Felicity. 1996. “An acoustic study of vowel variation in Australian English”. Ph.D.,
Macquarie University.
Eckert, Penelope. 1989. “The whole woman: sex and gender differences in variation”.
Language Variation and Change 1:245–67.
Giles, Howard, Richard Y. Bourhis, and D. M. Taylor. 1977. “Towards a theory of
language in ethnic group relations”. In Howard Giles, ed. Language, Ethnicity, and
Intergroup Relations. London: Academic Press, 307–48.
Harrington, Jonathan, Felicity Cox, and Zoe Evans. 1997. “An acoustic phonetic study of
Broad, General, and Cultivated Australian English Vowels”. Australian Journal of
Linguistics 17:155–184.
Horvath, Barbara. 1985. Variation in Sydney English: The Sociolects of Sydney. Cambridge: Cambridge University Press.
Kerswill, Paul and Ann Williams. 1992. “Some principles of dialect contact: Evidence from
the New Town of Milton Keynes”. Working Papers 1992: Occasional Papers in
General and Applied Linguistics. Department of Linguistic Science, University of
Reading. 68–90.
Labov, William. 1963. “The social motivation of a sound change”. Word 19:273–309.
———. 1984. “Field methods of the project on linguistic variation and change”. In John
Baugh and Joel Sherzer, eds. Language in Use. Englewood Cliffs, New Jersey:
Prentice-Hall, 28–53.
———. 1994. Principles of Linguistic Change: Internal Factors. Oxford: Blackwell.
Milroy, Lesley. 1980. Language and Social Networks. Oxford: Blackwell.
Mitchell, Alexander George. 1946. The Pronunciation of English in Australia. Sydney:
Angus and Robertson.
——— and Arthur Delbridge. 1965. The Pronunciation of English in Australia. Sydney:
Angus and Robertson.
Schilling-Estes, Natalie. 1997. “The linguistic and sociolinguistic status of /ay/ in Outer
Banks English”. Ph.D., University of North Carolina at Chapel Hill.
Tsolidis, Georgina. 1995. “Greek-Australian families”. In Robyn Hartley, ed. Families and
Cultural Diversity in Australia. St. Leonards: Allen and Unwin, 121–43.
Wells, John C. 1982. Accents of English 2: The British Isles. Cambridge: Cambridge
University Press.
The acquisition of colloquialisms
by non-native speakers
Jane M. Curtain
Monash University
The colloquial words and phrases of Australian English (henceforth “AusE”)
were originally influenced by British and Irish English but have more recently
been influenced by American English. Indeed, there is evidence to support the
notion that some traditional Australian colloquialisms are now being displaced
by modern Americanisms, particularly with younger speakers. For this reason,
many newcomers to Australia are not exposed to the idiom of Australia’s past.
The fact that Australia is a multicultural society raises several pertinent
language-related issues. One of these is the role that traditional Australian
colloquialisms may have in creating language barriers between native and
non-native speakers of AusE. Despite the fact that many immigrants will have
competency in more formal English and even in “global” colloquial English,
they find themselves confronted with a language community in Australia
which uses much unfamiliar colloquial language.
The issue is one that has not received attention in the linguistic literature
(perhaps because lexis lacks the systematicity of such core areas of linguistics
as phonology and syntax). And this is an unfortunate situation, given the fact
that colloquial lexis is widely recognised as an inherent and distinctive component of AusE, a dialect which has its origins, at least in part, in the slang
used by the convicts and poor “free” settlers of the 18th and 19th centuries.
This primary aim of the study reported here was to examine the acquisi1
tion of AusE colloquialisms by a particular group of non-native speakers with
similar language backgrounds living in Australia: Malaysian tertiary students.
The linguistic situation in Malaysia today makes such subjects an ideal choice
for this study. Most Malaysians are multilingual: since Malay became the
national language of Malaysia in 1957, English has been acquired as the first
foreign language by most younger speakers (Edwards 1994: 183) and, because it is perceived as being a prestige language, English is taught to all
Malaysian students. Although the official exonormative standard for Malaysian English is British, in recent years Malaysia has been strongly influenced
— like most English-speaking communities — by the American film and
media industries. As a result much American idiom is recognised throughout
the Malaysian community. By contrast, despite Malaysia’s close proximity to
Australia, little Australian idiom appears to be recognised in the country.
The objectives of the study were threefold:
(a) to examine and compare the extent to which traditional Australian and
modern (mainly popular American) colloquialisms are understood by native
and non-native speakers of AusE;
(b) to examine the extent to which the age of a native AusE speaker has an
effect on their understanding of traditional Australian and modern colloquialisms; and
(c) to examine the extent to which the number of years a non-native speaker
has lived in Australia has an effect on their understanding of traditional
Australian and modern colloquialisms.
Methodology of the study
The data used in this study was collected using a questionnaire which was
developed using information obtained from two surveys. Before the surveys
were conducted, 190 colloquialisms were collected from spontaneous interactions with native AusE speakers over a four-week period. These colloquialisms
were then categorised according to whether they were “traditional Australian”
or “modern”, using authoritative dictionaries. One hundred and twenty colloquialisms were deemed “traditional” because they had been in existence in
Australia for over 50 years and were considered Australian in origin.2 Seventy
colloquialisms were deemed “modern” because they had been in existence for
less than 50 years.3 It was obvious that some of the 190 colloquialisms were
more popular than others, because their frequency of usage was much higher.4
2.1. Traditional Australian colloquialisms
The aim of the first survey was to identify the most popular and least contentious of the 120 traditional Australian colloquialisms, those which would in
turn be the most suitable for the main investigation. Sixty participants (26
males and 34 females) took part in this survey; all participants were at least
second generation native AusE speakers, which ensured they all had a relatively good chance of having been exposed to the colloquialisms in the survey.
They ranged in age from 14 to 68 (mean age = 34.13). Under supervision, the
participants were asked to provide definitions for the 120 colloquialisms.
Target definitions, obtained from dictionaries, were used to accept or reject
the participants’ responses, and the results were quantified.
Acceptable definitions of 12 (see Table 1) of the original 120 traditional
expressions were provided (and agreed upon) by all 60 participants. These 12
colloquialisms were, therefore, deemed the least contentious and, consequently, the most suitable for use in the main investigation.
Table 1. The 12 least contentious traditional Australian colloquialisms
(Underlined sections only)
This is a bonzer little joint
He’s got Buckley’s chance of getting it
He gave me an ear bashing all night
Come on, spill your guts
He was as happy as Larry
He’s about to shoot through
He took a sickie yesterday
excellent, attractive, pleasing
a very slim chance; a forlorn hope; no chance
an incessant and long talk; a haranguing
divulge, disclose, reveal all you know; tell all
completely or extremely happy
leave; leave in a hurry
a day’s sick leave, taken whether one is sick or
the city
a break from work for morning/afternoon tea
and/or a cigarette
absolutely, certainly (an expression of
emphatic agreement)
difficult and/or tiring work
She’s very new to the big smoke
Is it time for smoko yet?
We’re having snags for tea
Too right I did!
It was really hard yakka
2.2. Modern colloquialisms
The procedure for the second survey was almost identical to that for the first.
The aim was to discover the most popular and least contentious of the 70
modern colloquialisms, by asking participants to provide definitions for them.
Sixty participants (31 males and 29 females) took part; all were native speakers of English, all except two were native speakers of AusE, and they ranged
in age from 17 to 68 (mean age = 29.86).
Acceptable target definitions of 13 (see Table 2) of the original 70
modern expressions were provided (and agreed upon) by all sixty participants.
These 13 colloquialisms were, therefore, deemed the least contentious of all
those included in this phase, and were deemed the most suitable for use in the
main investigation.
Table 2. The 13 least contentious modern colloquialisms
He went ballistic!
It was ace!
That lecture was really full on!
All he wants to do is get laid
Yeah, I get the picture
I’m out of here
You paid $100.00? You were
Don’t talk shit!
Last Saturday he got smashed
Jeez, you’re really up shit creek,
aren’t you?
I just want to go home and
veg out
What a spunk!
Jeez, you’re a couch potato!
(Underlined section only)
became very agitated or angry
excellent; first in quality
performed without reserve or restraint; carried out to
the utmost
have sexual intercourse
understand; comprehend the situation or circumstances
leaving immediately
taken advantage of; swindled
speak nonsense; tell lies; exaggerate, bluff, boast; talk
in a negative way
drunk; stoned
in trouble; in difficulties; in dire straits
relax into a passive and accepting frame of mind,
especially by watching television
very attractive person; sexually attractive person
dull or inactive person, especially one who spends a lot
of time watching television
2.3. The developed questionnaire
One hundred newly recruited participants (58 native Australian English
speakers and 42 Malaysians) were surveyed in the final phase of the investigation. These participants were classified into four groups, as described in
Table 3.
Table 3. Summary of participants
N = 100
Native Australian English speakers
n = 58
Group 1
native Australian
English speakers
(all students) aged
between 18 & 28
Group 2
native Australian
English speakers
(mostly students)
aged over 50
n = 36
Non-native Australian English speakers
(Malaysian students)
n = 42
n = 22
Group 3
Malaysian students
aged between 18 &
28 who had lived in
Australia for < 2
n = 21
Group 4
Malaysian students
aged between 18 &
28 who had lived in
Australia for > 2 but
< 5 years
n = 21
mean age = 20.5
mean age = 55.6
mean age = 21.2
mean age = 21.4
For Groups 1, 3 and 4, the participants were all students at Monash University;
for Group 2, 15 of the 22 participants (68.2%) were students at Monash, while
seven were employees of the university.
A final questionnaire was developed listing the 25 least contentious
colloquialisms mentioned above: the 12 traditional ones from the first survey
and the 13 modern ones from the second survey. Again, participants were
asked to provide a definition for each of the expressions listed, and target
definitions were used to accept or reject their responses.
The results were scored with each participant receiving a score out of 12
for the traditional colloquialisms and a score out of 13 for the modern
colloquialisms. These scores were then converted to percentages, and parametric t-tests were used to compare the mean percentage of acceptable definitions between both the colloquialism types and the participant groups.
Due to the size and complexity of the original study, it is not possible to
discuss all of the data obtained.5 For this reason, only those collective results
which relate directly to the aims presented in Section 1 will be addressed. The
main findings (see Table 4) — almost all of which were supported by statisti-
cal evidence (p<0.05) — can be summarised as follows:
(a) The native AusE speakers had a significantly greater understanding of
colloquial English expressions (both traditional Australian and modern) than
did the non-native AusE speakers.
(b) The age of the participants had a significant effect on the type of colloquialisms with which they were familiar. In particular:
• Native AusE speakers aged between 18 and 28 had a significantly
greater understanding of modern colloquialisms than did those aged
over 50;
• Native AusE speakers aged over 50 had a significantly greater understanding of traditional AusE colloquialisms than did those aged between 18 and 28.
(c) The amount of time a non-native AusE speaker had been living in
Australia affected the amount of colloquial English with which they were
familiar. In particular:
• Malaysian students who had been living in Australia for more than
two but less than five years appeared to have a greater understanding
of traditional AusE colloquialisms than did Malaysian students who
had been living in Australia for less than two years (This finding,
however, was not statistically supported);
• Malaysian students who had been living in Australia for more than
two but less than five years had a significantly greater understanding
of modern colloquialisms than did Malaysian students who had been
living in Australia for less than two years.
Table 4. Mean percentage of acceptable responses for all participant groups as a function
of colloquialism type
Group 1:
Group 2:
Group 3:
(< 2 yrs)
Group 4:
(>2<5 yrs)
4.1. Speaker origin
The results indicate that the native AusE speakers provided a higher mean
percentage of acceptable responses on both colloquialism types than did the
Malaysian students. These results are hardly surprising: the native speakers
had had many more opportunities to be exposed to colloquial AusE. For the
majority of Malaysians (including all the participants in this investigation),
English is not a native language,6 and the Malaysian students were therefore
unlikely to have had the same degree of competency in colloquial AusE. In
fact, some of them claimed that they did not have a reasonable grasp of
English in general, either formal or informal.7
4.2. Colloquialism type
The age of the speakers had a significant impact on the type of colloquial
English with which they were familiar. With the exception of those in Group
2, most participants provided a higher mean percentage of acceptable definitions for the modern colloquialisms than they did for the traditional colloquialisms. Again, this is not a surprising result: during the time of the investigation
(mid-1996), anecdotal evidence suggested that many of the modern colloquialisms were being used with increasing frequency on the university campus,8
and most of the participants were therefore likely to have been exposed to
them. For the 78% of participants aged between 18 and 28 the modern
expressions probably represented current trends in popular idiom.
Many of the modern colloquialisms were known to be used by actors and
broadcasters in the media, particularly the American media. The 20th century
has seen a greater influence of American culture on Australia: alliances in the
World Wars, the Internet, the Hollywood film industry, the popular music
scene, and general tourism have changed the external influences on AusE
from predominantly British to predominantly American. The Global Village
transference effect seems to be heavily weighted in favour of American idiom.
It could therefore be argued that the results simply reflect this change in the
informal lexis of the AusE-speaking community.
4.3. Age
As mentioned above, age is an important determinant of the class of colloquialisms which one acquires or with which one chooses to be familiar. This is
particularly relevant when one compares the differences between the older
and younger native AusE speakers.
Table 4 shows that the older AusE speakers provided a higher mean
percentage of acceptable responses for the traditional colloquialisms than did
the younger AusE speakers. In contrast to this, Table 4 also shows that the
younger AusE speakers provided a higher mean percentage of acceptable
responses for the modern colloquialisms than did the older speakers.
In order to explain this, it must be noted that the traditional colloquialisms
had been in circulation in Australia for at least 50 years prior to the investigation — almost 30 years more than the average age of the participants in Group
1. The older AusE speakers had had many more opportunities to be exposed to
these expressions. In fact, it is likely that (at least) some of the traditional
colloquialisms were the modern expressions of 50 years ago and they therefore held an on-going attraction for those speakers who had grown up using
For the younger speakers, the modern colloquialisms were probably
much more appealing in terms of motivation for acquisition than were the
traditional colloquialisms. One of the most important functions of colloquial
language is to signify membership to a particular group, profession or social
class, and many of the 13 modern colloquial expressions were originally
obtained from spontaneous conversations with university students. Insofar as
all of the Group 1 participants were university students, these expressions
were most likely indicative of the particular colloquial lexis signifying the
group membership associated with tertiary students, or even similarly aged
speakers in general.
4.4. Period of exposure
The results clearly indicate that a higher mean percentage of acceptable
responses was provided for both colloquialism types by the Group 4 participants than by the Group 3 participants.9 This empirical evidence suggests that
the likelihood of colloquial English being acquired and the motivation for
acquiring it correlated with the relevance of the type of colloquial language to
the age and interests of speakers. Indeed, the Malaysian students’ acquisition
of colloquial English expanded as their length of time spent in Australia
increased. However, because these speakers were undoubtedly exposed to a
higher concentration of modern colloquial English, such as that used by
tertiary students, it is not surprising that they became even more familiar with
modern colloquialisms.
It is likely that (at least some) of the Malaysian participants identified
with Australians of a similar age, and that this association increased over time.
The Group 3 and 4 participants were very similar in age to the Group 1
participants, and as already noted, these younger speakers of AusE also
demonstrated a greater understanding of the modern colloquialisms.
4.5. Individual colloquialisms
Table 5 identifies the percentage of participants who provided acceptable
responses for each of the colloquialisms. Although all 25 of these colloquialisms were chosen because in the preliminary surveys they received 100%
definition agreement from the native English-speaking participants, the results
do not necessarily support the initial consensus. As Table 5 shows, not all of
the native AusE speakers provided acceptable definitions for either the traditional Australian or the modern colloquialisms. For example 33% of the
Group 2 participants failed to provide an acceptable definition for That lecture
was really full on! (in fact they failed to provide any definition at all, thereby
indicating their unfamiliarity with the expression), and 18% failed to provide
the target definition for Last Saturday he got smashed. Interestingly some
participants suggested an alternative — possibly perfectly valid.10 — definition for smashed, namely “bashed up”.
A possible explanation for these differences is that the participants used
in the main phase of the study, who were all Monash students, represented a
slightly different population from those used in the preliminary surveys, a
group including not only Monash students but also seven full-time employees
of the University.
Table 5 also provides evidence that all 25 colloquialisms received acceptable definitions from a considerably lower percentage of participants in
Groups 3 and 4 (Malaysian students) than in Groups 1 and 2 (native AusE
Table 5. Results pertaining to each colloquialism: Mean percentage of participants who
provided acceptable definitions.
Traditional Australian Colloquialisms:
This is a bonzer little joint
He’s got Buckley’s chance of getting it
He gave me an ear bashing all night
Come on: spill your guts
He was as happy as Larry
He’s about to shoot through
He took a sickie yesterday
She’s very new to the big smoke
Is it time for smoko yet?
We’re having snags for tea
Too right I did!
It was really hard yakka
Participant Group
Modern Colloquialisms:
He went ballistic!
It was ace!
That lecture was really full on!
All he wants to do is get laid
Yeah, I get the picture
I’m out of here
You paid $100.00? You were screwed!
Don’t talk shit!
Last Saturday he got smashed
Jeez, you’re really up shit creek, aren’t you?
I just want to go home and veg out
What a spunk!
Jeez, you’re a couch potato!
speakers). In fact, for four of the traditional Australian colloquialisms, less
than 5% of the Group 3 participants provided acceptable responses. These
colloquialisms are:
He’s got Buckley’s chance of getting it
He’s about to shoot through
She’s very new to the big smoke
We’re having snags for tea
The same four colloquialisms were not much more successfully understood by the Group 4 participants, receiving percentage scores of 19%, 10%,
14% and 10% respectively. It would seem that the number of years the
Malaysian students had been living in Australia did not make a huge difference in their acquisition of these particular colloquialisms.
For four of the modern colloquialisms, less than 20% of the Group 3
participants provided acceptable responses. These colloquialisms include:
That lecture was really full on (14%)
Last Saturday he got smashed (10%)
Jeez, you’re really up shit creek, aren’t you? (10%)
What a spunk! (10%)
The same four colloquialisms were more successfully understood by the
Group 4 participants, receiving acceptable percentage scores of 43%, 48%,
24% and 33% respectively. This suggests that the amount of time spent living
in Australia contributes to the recognition of certain modern colloquial expressions.
The findings of the study suggest that certain colloquial expressions, especially traditional Australian colloquialisms, may contribute to the language
barriers between native and non-native speakers of AusE.
Future research should perhaps target a different group of non-native
speakers to see if the performance of these individuals resembles that of the
Malaysians. Given that the range of ethnicities within Malaysia itself is quite
complex,11 and that this complexity might have affected the results of the
present study in some small way, it would be interesting to investigate an
ethnic group which is more linguistically and culturally homogeneous.
The original investigation, as reported in Curtain (1996), examined the use, understanding and personal perceptions of traditional Australian and modern colloquialisms by
native and non-native speakers of Australian English.
Some of the traditional Australian colloquialisms had early origins in Britain or Ireland.
Most of the modern colloquialisms were considerably ‘younger’ than 50 years.
See Curtain (1996) for details on the frequency of the original 190 colloquialisms.
See Curtain (1996) for detailed descriptions and analyses of the participants’ usage,
understanding and personal perceptions of the colloquialisms used in this investigation.
Curtain (1996) also discovered that differences as a result of gender were not apparent.
As mentioned earlier, English is usually the first foreign language to be acquired by
As already noted, it is compulsory for all students to learn English in Malaysian schools
(Samah 1994: 57). Furthermore, the students must have had the required level of
competency in English to be granted entry into Monash University. These facts suggest
that such negative self-appraisals of English competency are prima facie incorrect or (at
the very least) exaggerated.
Approximately 93% of all participants were Monash University tertiary students. The
remaining 7% of participants (all representing Group 2–Native AusE speakers aged over
50) were full-time employees of the university, and had daily interactions with the
Despite these trends, however, no statistical differences were found between these two
groups for the traditional Australian colloquialisms.
It should be noted, however, that the “bashed up” definition was not reported in any of the
dictionaries consulted.
Ethnic Malays and ‘indigenous’ people make up approximately 53% of the Malaysian
population, the Chinese constitute approximately 35%, Indians approximately 11%, and
others (e.g. Indonesians, Thais, Europeans and Australians) 1%. There are four main
languages in Malaysia: Malay, English, ‘Chinese’ and Tamil, but Bahasa Malaysia
(based on Malay) is currently the only official language (Edwards 1994: 182).
Curtain, Jane. M. 1996. “The acquisition of traditional Australian and modern (mostly
American) colloquialisms by native and non-native speakers of Australian English”. BA
(Honours) thesis, Monash University.
Delbridge, Arthur, John R. L. Bernard, David Blair, Pamela Peters and Susan Butler, eds.
1995. The Macquarie Dictionary. 2nd Edition. McMahon’s Point: The Macquarie
Edwards, J. 1994. “Language planning and education in Singapore and Malaysia”. In A.
Hassan, ed. Language Planning in Southeast Asia. Kuala Lumpur: Dewan Behasa dan
Pustaka, 176–194.
Platt, John, Heidi Weber and M. L. Ho. 1984. The New Englishes. London: Routledge and
Kegan Paul.
Ramson, William S. 1970. “Nineteenth-century Australian English”. In William Ramson,
ed. English Transported: Essays on Australasian English. Canberra: Australian National University Press, 32–48.
Samah, A. A. 1994. “Language education policy planning in Malaysia: Concern for unity,
reality and rationality”. In A. Hassan. ed. Language Planning in Southeast Asia. Kuala
Lumpur: Dewan Behasa dan Pustaka, 52–65.
Changing attitudes to Australian English1
David Bradley and Maya Bradley
La Trobe University
There are various ways in which speakers express attitudes to speech variation
within a community. These include unsolicited comments made by a variety
of people and various types of elicited observations. The comments are
usually made by people who are more concerned with and aware of language:
authors, columnists and other journalists, comedians, public intellectuals and
other public figures, and people who choose to make an effort to write a letter
to a newspaper or telephone in to a talkback radio program on language. All of
these groups are expressing popular stereotypes about language, some of them
widely held and some less so. Because they are probably more languageaware than the general population, their comments may be in advance of
general perceptions, and of course may also have some influence in forming
popular attitudes. One measure of changing feelings about Australian English
(hereafter “AusE”) is the rapidly expanding expression of such stereotypes
and the increasingly positive attitude to Australian English in them. We
discuss this in Section 2 below.
In Section 3 we report on more systematic and direct approaches to
measuring changes in attitudes to AusE through time. This is in two parts:
firstly, we interviewed a randomly selected sample of speakers in Melbourne
in 1980. The interview included a variety of direct questions about attitudes to
AusE.2 As many of the same speakers as could be found in 1995 were
reinterviewed with the same questions; changes in answers from this group are
discussed. There are no speakers who changed from expressing positive views
about AusE in 1980 and negative ones in 1995; but 18.5 per cent of the
speakers became more positive about AusE over these fifteen years.
Secondly, we prepared an attitude test (often called a “Subjective Reaction Test” after Labov (1972); hereafter “SRT”) with three AusE stimulus
speakers selected to represent the sociolectal range within AusE, speaking
casually on the same topic. We trialled the stimulus tape with various groups,
and the final version of the SRT has been regularly administered to groups of
university students since 1982; they are asked to listen to the tape and rank the
speakers on various semantic scales which relate to status and solidarity. We
report here on the results for comparable large groups tested from 1984 to
1998, and discuss the significant changes which have taken place in ratings of
these three speakers. Also briefly noted are various other SRTs which have
been carried out on AusE over the last twenty years.
Public stereotyping of Australian English
Because our main focus is the direct measurement of changing attitudes to
AusE, as discussed in Section 3 below, in this section we shall give only a few
salient examples of each type of evidence on stereotyping. These nevertheless
illustrate attitude changes in a rather clear way. For the period up to 1898,
Ramson (1966) has surveyed much of the relevant literature in this area; a
brief update for the current century is presented below.
One unexplored source of direct personal comment on language is in
Australian autobiographies. Since much of an author’s life history is given in
an autobiography, we can usually determine his or her social background and
thus compare his or her views with current attitudes of people of similar status.
Of course all those who work with the written word, and especially those who
have time to edit and polish their work, are certainly more aware of language.
Their comments are particularly useful because they provide insights into an
earlier historical context not otherwise available.
As an example, we will quote from Porter (1963), who was born in
Melbourne in 1911 and lived his childhood there and in a Victorian country
town. He wrote that he was “an unmistakable Australian, albeit of the
Awstralian rather than the Osstralian variety” (ibid. 106), by which he presumably means General to Cultivated rather than Broad. He describes himself
and various family members as showing very substantial shift towards more
cultivated speech in formal settings: a school play, family holiday gatherings
and so on. Insightfully, he remarks that as he shifts from adolescence and
school to adulthood and work, he also shifts his sociolect (ibid. 177). He
expresses ambivalent views, but mainly negative ones, about Broad AusE,
and notes that variation is not just determined by social status:
… the accent is defiantly preserved. Wealth cannot taint it nor education undo
it. … It is an ineradicable and perverse accent, signal at once of the possible
strengths and certified weaknesses of the Australian character … (ibid. 68)
These comments can usefully be compared to those collected from our interview sample in Section 3.1 below.
While journalists have less time to work over every word that they write,
the process of subediting makes them think about language and develop strong
feelings about linguistic issues. Many serious newspapers have regular columnists who comment on language matters, and these are an invaluable source of
data on stereotypes. In a published collection of some of his columns on
language, Max Harris (1989: viii) claims that he had the first such column in
Australia; initially in the Bulletin, later in the Australian and syndicated
elsewhere. Of course he was a literary figure, one of the “Angry Penguins”,
long before this. Some of the Harris columns are superb examples of stereotyping of forms from lower-prestige sociolects, such as his comments on past
participles like known, which in AusE may have a schwa before the final /n/ as
in Irish, Scots and New Zealand English; see Bradley and Bradley (1985) for
further details. Harris (1983) wrote:
The interpolation of a mandatory “e” is caused by degenerative tongue
reflexes. … To say “known” the tongue must move smartly from below the
lower teeth to a point above the upper teeth. The untrained tongue muscle
finds it too strenuous to cope with the exercise and interpolates an air puff and
an “e” to allow the languid lingual mechanism to get there.
Another language columnist with a purely journalistic background was Alan
Peterson, who also collected some of his 1979 to 1986 columns from the
Sydney Morning Herald in a book, Peterson (1986); his attitude to AusE is still
prescriptive but distinctly more positive. The most regular current columnist is
Frank Devine, with his frequent and relatively well-informed column “That’s
Language” in the Weekend Australian. Newspapers also print quite a few
letters to the editor and very occasional guest columns from linguists on
language issues; one topic which is frequently aired is a belief that AusE is
becoming Americanised. These letters tend to go in cycles, with one triggering
many others.
The outpouring of popular books on AusE lexicon is another manifestation of changing attitudes and greater interest. After scholarly early studies by
Morris (1898), Baker (1945) and Wilkes (1978) among others, this swelled to
a flood in the 1980s and has expanded rapidly since. Among the earliest of the
popular genre was Hornadge (1980); another early example is Keesing
(1982). Some such publications are by media personalities, such as the television figure and radio presenter John Blackman (1990).
Another reflection is seen in the representation of AusE in television and
films, especially comedy. If we think back to the “Norman Gunston” character
of the mid-1970s, the Broad speaker was portrayed as stupid and incompetent.
By the 1980s the “Crocodile Dundee” character was being portrayed as much
more competent, although still a larrikin. In current television comedy, there
are still many characters such as the “Dodgey brothers” in the Comedy
Company or the “Sandman” on Good News Weekend, as well as ethnic broad
characters such as “Con the fruiterer”, who are portrayed as stupid; but would
comedy be funny if all the characters were super-competent? However there
are also many leading younger comedians who use broad speech and nevertheless are portraying themselves as highly intelligent and competent.
An ongoing social commentary reflecting one comedian’s attitude to AusE
is in the work of Barry Humphries. Negative views of AusE and its speakers
abound in the early 1970s “Barry McKenzie” character, the chunderous Ocker
in London; or the even more negative role of the somewhat later “(Sir) Les
Patterson”, the venal, lecherous and drunken politician. The gradual metamorphosis of the early “Edna Everage” character from being a stereotyped housewife from a working class area of Melbourne — with a name intended to show
how “average” she is — into a superstar who becomes a Dame, has tea with the
queen and speaks in an increasingly cultivated way that nevertheless still has
traces of its lower sociolectal origins, shows that even the expatriate Humphries
may be becoming more Australia-positive. This is underlined by Humphries’
announcement in early 1999 that “Dame Edna” has renounced her royal honour
and is now in favour of Australia becoming a republic.
The ABC3 is another bellwether of change. It has had a pronunciation
committee for more than fifty years; this receives a flood of letters reflecting
viewers’ stereotypes about AusE and criticisms about perceived linguistic
lapses on the ABC. Their archives are a valuable and largely untapped source.
Until the 1980s, most ABC newsreaders used cultivated to modified sociolectal forms; but the new generation of the 1990s does not, and in 1995 when he
took over, the new ABC Managing Director Brian Johns was interviewed on
one of his own radio stations and said “We don’t want an outdated accent”, by
which he presumably meant the Modified to Cultivated form of AusE that
prevailed at the ABC from its beginnings up to the 1980s.
Over the last ten years, there has been a voracious and growing demand
for linguists to talk on the radio, especially but not exclusively the ABC, about
English usage in Australia, with an increasing variety of regular spots in the
last five years or so. While the linguists do not express popular stereotypes, the
callers do; and the demand for such programs also reflects an increasing
interest in AusE.
We have also seen an expanding proportion of politicians using General
to Broad sociolects. Archival recordings of right-wing politicians such as Sir
Robert Menzies and even Labor figures such as Dr H. V. Evatt show us that, at
least in public speeches, they used to be far towards the Cultivated and
Modified end of the spectrum. Now most politicians are in the General to
Broad range, and many, especially but not exclusively from the left end of the
political spectrum, tend mainly towards Broad. Bob Hawke is one of the
better-documented sociolectal chameleons; his Boyer lectures on the ABC
were General to Cultivated, as a union leader his speech was quite Broad, but
slightly less so later as a politician. Now that he is a business consultant and
retired statesman his speech is moving back into the General to Broad range.
One can only wonder how he spoke as a minister’s son and young Rhodes
scholar at Oxford.
Many Australian public intellectuals for whom language is not a core
issue nevertheless feel impelled to comment on English in Australia; as Robert
Manne writes, it is “in some ways the deepest inheritance of all” from Britain
(1998: 110). We all speak and have opinions about it, so in a sense we are all
experts. Linguists do not have a monopoly on language or metadiscussion of
language issues; and non-linguists’ commentary provides very valuable data
for the understanding of changes in attitudes about it.
Measurement of changing attitudes to Australian English
As is discussed elsewhere in this volume by Yallop, the basic sociolectal
model established by Mitchell and Delbridge (1965) persists and frames our
discussion of variation in AusE. Of course linguists are all aware that each
speaker varies in use along a continuum between Broad, General, Cultivated
and Modified sociolectal forms, though perhaps no individuals normally use
the full range of vowel and other alternatives. However we cannot expect nonlinguists to have such a finely-grained grid of sociolectal terminology, even if
they can react to it as members of the speech community.
In Section 3.1 we attempt to see what ordinary people can say about
variation and change in AusE. This is partly quantitative, based on answers to
specific questions with a clear set of alternative answers. The data is also
partly qualitative: we asked people to discuss language issues. These interviews were done with the same speakers twice, once in 1980 and again in
In Section 3.2 we follow a procedure widely used in sociolinguistics and
the social psychology of language: having prepared a stimulus tape representing several alternative sociolectal patterns, we ask relatively naive people to
judge the different speakers on a variety of semantic differential scales. This
provides a large quantity of readily analysed data. Although some of the
“judges” (the naive listeners) find the task strange or even invidious, it is
almost never difficult, provided that the choice of semantic differential scales
is not too outlandish; and the results are robust and replicable. Again, this is a
longitudinal study, with results for different but comparable groups who heard
the same stimulus between 1984 and 1998.
3.1 The 1980 and 1995 interviews
As part of a larger study in which we interviewed a random sample of 40
Melbourne-born speakers selected from electoral rolls in two suburbs of
Melbourne in 1980 and smaller samples of ten speakers from two suburbs of
most other capital cities over the next few years, we included a number of
direct questions to elicit views and discussion on AusE. In 1995 we sought to
reinterview the 40 Melbourne subjects and were able to find 27; some had to
be reinterviewed by telephone outside Melbourne. The other 13 were deceased or could not be found. We were in fact surprised to achieve such a high
reinterview rate, which represents more than two-thirds after a fifteen year
gap. We are not aware of any previous sociolinguistic investigations which
have done this kind of genuinely longitudinal study, reinterviewing the same
people randomly selected at the beginning, though of course such studies are
not uncommon in other disciplines such as sociology.
On both occasions the interviewees were asked which variety of English
they prefer; suggested alternatives were Australian, British (RP) or American.
One preferred American, both in 1980 and 1995; another preferred Irish, the
accent of her family background but not her own. Of the rest, eleven preferred
British in 1980, eleven preferred Australian, and three felt equally positive
about both British and Australian. In 1995 the 25 divided between British and
Australian preferences differently: fifteen preferred Australian, nine preferred
British and one liked both. All three who said they preferred British and
Australian in 1980 shifted to preferring Australian in 1995, along with one
who preferred British in 1980. One speaker who was British-positive in 1980
shifted to preferring both. If we take these “both” preferences as half Australia-positive, the results can be represented as in Table 1. The five speakers
who shifted their preference away from British in the direction of Australian,
in part or completely, represent 18.5 per cent of the sample, and the overall
increase in Australia-positive responses is 11.1 per cent. Though the sample is
small, the results are suggestive: attitudes have changed since 1980! The 1980
total does not include interviewees who had died or could not be found in
1995. Table 1 also shows a possible connection between attitude shifts and
wars: the oldest Australia-positive speakers of 1980 were born during or
shortly after World War I, and the last Australia-negative speakers were born
within ten years of the end of World War II. These wars were traumatic for
Australia’s connection with Britain: massive casualty rates for Australian
soldiers on the other side of the world; and during the second, British inability
to protect Australia while nevertheless demanding large numbers of Australian troops for service in Europe.
Table 1. Australia-positive answers, per cent and by birthdate, 1980 and 1995
Oldest Australia-positive
Youngest Australia-negative
The interviewees were born between 1899 and 1962. Since the sample was
taken from electoral rolls in 1980, no one then under 18 and born after 1962 was
interviewed; however, we believe that there would be few if any Australianegative speakers among Australians born after 1962. Literary and autobiographical material would suggest an earlier date for the first Australia-positive
speakers, though they might then have been a small proportion of the overall
population. As noted in Section 2 above, Hal Porter, born in 1911, is a possible
example, though he seems at times ambivalent. There is insufficient space here
for full discussion of the qualitative comments made by interviewees; for that,
see Bradley and Bradley (forthcoming).
3.2 The Subjective Reaction Test
There have been many previous SRT studies in Australia; most of them have
investigated perceptions of accented versus unaccented speech, but quite a few
have also looked at AusE itself. Of these, some have had a primarily sociolectal
goal like our study, many have looked at reactions to ethnic-accented speech,
while a few have attempted to determine whether regional differences can be
recognised by Australian judges. Ball et al. (1987) summarise published
studies up to that time, but a great deal of unpublished work has also been done,
much of it by our students.
In this study, we used stimulus passages of casual speech on Labov’s
“danger of death” question from our 1980 AusE interviews. The three speakers are male, of similar age, self-employed professionals living in a high-status
inner southeastern suburb of Melbourne. One uses predominantly Broad
sociolectal forms, another uses primarily General forms, and the third uses
primarily Cultivated forms.
Listeners were played recordings of the three speakers once, then again
with pauses between speakers for them to tick ratings on five-point scales for
three well-known status-related characteristics (job suitability, intelligence
and education) and three solidarity characteristics (friendliness, honesty and
toughness). All listener/judges were also asked to provide personal information about themselves. The judgements reported here are only those of Australian-born judges; however we also have extensive records of responses from
people of a variety of backgrounds, including speakers of other varieties of
English and speakers born overseas in a variety of countries. The personal
information collected also includes country of birth and occupation of parents,
though we will not discuss correlations between judgements by the Australian-born judges and background ethnicity or family occupational status here.
Most researchers using SRTs have preferred to use reading passages; this
gives constant content, but actually makes reading skill and fluency a major
independent variable. Some such studies carried out by students are Poynton
(1979), Schornikow (1981), Finch (1982), Sirianos (1984), Zeccola (1985)
and Spitale (1986). In some SRTs the “matched guise” technique originally
developed for studies of attitudes to bilinguals is used, but stimulus speakers
are asked to assume different sociolectal roles; Berechree and Ball (1979) is
one such study. These have the crucial problem that they rely on the stereotyped behaviour of the stimulus speakers, which may be inaccurate; but at
least this avoids the independent variable of individual voice quality differences.
As far as we know, our topic-controlled SRT which started in 1980 was a
first, though several others have more recently been carried out by students
(Beagle 1995; Al-Hindawe 1998), and there are doubtless others. Beagle
directly investigated the effect of reading passage versus topic-controlled
stimulus, with interesting results.4 Al-Hindawe controlled another important
independent variable, interviewing her judges after the SRT to determine how
judge sociolectal usage correlates with their rankings.
Table 2 presents results for groups of 30 or more female Australian-born
judges between 1984 and 1998, for which statistical tests are useful; groups in
other years were smaller. Table 3 compares overall female and male judgements during this period.5 Because responses on the “toughness” scale
showed a high standard deviation, perhaps partly due to the content of the
danger of death experience reported, this has been excluded from the ‘solidarity’ means below. A response at the high end of the scale (suitable for a person
with a professional job, very intelligent, very well-educated, very friendly,
and very honest) was coded as 1, and the other end of the scale (suitable for a
person with an unskilled job, very unintelligent, very poorly educated, very
Table 2. Judgements of subjective reaction test stimulus by female judges, 1984 to 1998
Broad speaker
General speaker
Cultivated speaker
No. of
unfriendly, and very dishonest) was coded as 5; thus the overall range of
rankings is from 1 to 5, and the mean for a given group of judges can only
range between 1 and 5.6
As Table 2 shows, the speaker with the most Cultivated sociolect is
consistently ranked lowest on solidarity and almost as consistently ranked
highest on status.7 However, the speaker with the most Broad sociolect is
ranked consistently lowest on status, but not highest on solidarity. It is the
speaker using mainly the General sociolect who is ranked highest on solidarity. This is contrary to what some SRTs outside Australia have found, but may
be related to the fact that the General sociolect is also used most frequently by
the majority of the population including presumably most of the judges. They
may therefore be more likely to judge this favourably as it is like their own
Furthermore, overseas findings would predict that the Cultivated speaker
would be ranked substantially higher than the General speaker on the status
scales. In fact this is so in all but the 1991 data, but the differences in these
status rankings are much smaller than those between the General speaker and
the Broad speaker. That is, it would appear that the sociolectal continuum is
judged nonlinearly, with the Broad sociolect ranked much more negatively for
status than General; and discontinuously, with Broad rated not as positively
for solidarity as General, which also far surpasses Cultivated in positive
solidarity rankings. We have also seen a more advanced version of this
tendency in Section 2 above: a decrease in the positive attitude and sometimes even a negative attitude to the Cultivated sociolect, as expressed for
example by the ABC Managing Director in 1995.
Another interesting observation is that the status and solidarity ratings of
the General speaker are rather similar and both positive, unlike the rankings of
the Broad and Cultivated speakers, which are at opposite ends of the scale for
status as opposed to solidarity. This may again reflect the central position of
the General sociolect, and be a factor in its increasing frequency of use.
Table 3 shows that there are substantial gender differences in rankings,
as indeed there are in the use of sociolects. The females show a greater degree
of polarisation than the males on the status scales: females rate the Broad
sociolect significantly lower and the Cultivated sociolect significantly higher
than males on these scales.8 This accords with the worldwide tendency for
females to be more sensitive to status differences in language. There is also a
major difference in that the males rank Broad and General speech very
Table 3. Male and female judgements of subjective reaction test stimulus, 1984 to 1998
Broad speaker
General speaker
Cultivated speaker
No. of
similarly for solidarity, while the females show a much greater difference.
This may reflect the related tendency for women to use General rather than
Broad sociolectal forms, and therefore to empathise more with a male speaking this way; and the tendency for more males to use Broad sociolectal forms
more of the time, and thus to have a more positive view of it.
In this SRT there are insufficient numbers of male judges to say much
about changes through time; but analysis of the raw data on which Table 2 is
based shows that over the fifteen years females show a trend to increase the
status rankings of Cultivated to some degree and General to a greater degree;
there are no obvious trends in the ratings of the Broad sociolect.
One major problem with this study is that its subjects are university
students of linguistics, who are certainly not very representative of the overall
Australian population.9 Many other studies share this design feature. However
it can be argued that the task is not one that can be controlled very consciously,
so this is not a fatal flaw.
Other SRT studies have used high school students as judges. They are
another readily surveyed group, but they are probably also not typical of the
overall community, but in a different way. Some adolescents have been shown
to shift their speech towards a vernacular norm as they move further into their
teens, as shown in Sydney by Eisikovits (1987); so older adolescents may not
be the most appropriate choice either. To document the linguistic shift from
late-adolescent vernacular prestige to adult standard prestige, described in an
Australian setting by Porter (1963) as cited above, a longitudinal study of a
cohort of adolescents as they move into adult life would be highly desirable.
In this paper we have examined changes in attitudes to AusE. This is illustrated by various examples from the media and other areas of the public
domain in Section 2 above. In Section 3 we have described attitude changes
over a fifteen year period: for a group of speakers who were interviewed
twice, once in 1980 and once in 1995; and for successive groups of similar
judges of an SRT which was administered between 1984 and 1998.
These attitude changes suggest that Australians are feeling progressively
more positive about Australian as opposed to other varieties of English speech.
A number of interviewees changed their attitude to a positive one between
1980 and 1995, and we found no speakers with a negative view of AusE born
after 1954. This is not to say that there are no such people; but they are probably
far fewer than in earlier generations. In the SRT, we have also observed
significant gender differences, with females showing greater polarisation in
status ratings: higher than males for Cultivated and lower than males for Broad.
One interesting trend through time is an increasingly positive view of General
speech, with the gap between Cultivated and General status rankings decreasing somewhat over fifteen years.
Other SRT studies, like ours, also suggest that Australia is different: it is
the middle sociolect, General, to which most people feel the greatest affinity;
not the maximally distinctive Australian vernacular, Broad. Broad is still
viewed as having much lower status, and Cultivated is judged as having
higher status than general. For example, Al-Hindawe (1998: 166) concludes
that her female adolescent judges reported maximally favourable attitudes to
the General sociolect in interviews conducted in 1997, even though they rated
the Cultivated sociolect higher on status scales in her SRT.
Interview results indicated that the General sociolect is considered to be both
the nicest and the best sociolect to have. … They are neither too casual nor too
formal, nor are they thought of as being common (‘yobbo’) nor too sophisticated (‘posh’). As far as the factors were concerned, G rated well on all traits
while B and C rated negatively on some. … However, results from the present
study indicated that the most popular sociolect is not necessarily the one with
the most status. It was C, not G, which had the highest rating in terms of
While longitudinal studies are inherently much more difficult to conduct, we
have found the results to be worthwhile. Non-longitudinal studies can never
give us the same degree of certainty about change, even if they are done in
‘apparent time’ with people of different ages. We have shown that individuals
and communities in Australia do change their sociolinguistic attitudes through
We want to thank the subjects who participated in the interviews reported in Section 3.1
and the many students who provided the judgements reported in Section 3.2. Thanks are
also due to colleagues who have commented on earlier presentations of parts of this study
at various seminars and conferences, notably the 1979 and 1987 Language and Speech
conferences, the 1994 and 1995 Australian Linguistics Society meetings, and seminar
presentations at Monash University. Of course all errors are solely our responsibility.
We gratefully acknowledge the support of the Australian Research Council for this
project in 1980–1985 and support from the University of Melbourne in 1979–1982.
The Australian Broadcasting Commission (hereafter ‘ABC’) is the government radio and
television network.
In Beagle (1995), the reading and free speech stimuli were provided by different speakers
of the same sociolect; judges heard only one of the two sets of stimuli, reading or free
speech. The only significant difference between these two conditions was in the rankings
of the two Broad Australian speakers, where the free speech received much higher ratings
(ibid. 36).
Male judgements are not broken down by year as there were too few Australian-born
male judges for statistical tests in most years. This reflects the fact that more female
students enrol in Arts degrees, and that more female than male Arts students choose to
study Linguistics. It may be that this also reflects a greater awareness of language issues
among females. This is a bias in the sample of judges.
Thus, for example, if all of a group of judges coded a speaker as ‘2’ on the job scale, ‘2’
on the intelligence scale and ‘3’ on the education scale, the mean score for status rankings
of that speaker by that group is 2.33. It is of course true that collapsing results inevitably
causes some detail to be lost; but this is necessary to present an overview in the space
available here. In some cases our discussion refers to non-collapsed results which are in
Bradley & Bradley (forthcoming).
We have no solid explanation for the one apparent reversal in status ranking: the
Cultivated speaker was placed lower than the General speaker by the 1991 group of
judges; however due to high standard deviation (up to 1.25 on Cultivated status rankings)
the difference is not significant.
Female judges ranked the Cultivated speaker significantly higher than male judges on all
three status scales; z test J values range from 1.25 to 4.03. For the two solidarity scales
used, the Broad speaker was rated significantly lower by females, with z test values of J
from 1.84 to 6.33. The collapsed totals presented in Table 3 obscure these trends.
The judges were all students at La Trobe University, which draws its students from a wide
range of socioeconomic, ethnic, attitudinal and educational backgrounds.
Al-Hindawe, Jayne. 1998. Attitudes towards Australian English. Ph.D., La Trobe University.
Baker, Sidney J. 1945. The Australian Language. Sydney: Angus & Robertson. 2nd edition
1966, Sydney: Currawong.
Ball, Peter, Cynthia Gallois and Victor J. Callan. 1989. “Language attitudes: a perspective
from social psychology”. In Peter Collins and David Blair, eds. Australian English: The
Language of a New Society. St Lucia: University of Queensland Press, 89–102.
Beagle, Lynda. 1995. Australian Attitudes to New Zealand Accents. Honours thesis, La
Trobe University.
Berechree, Peter and Peter Ball. 1979.”A study of sex, accent-broadness and Australian
sociolinguistic identity”. Paper presented at the 2nd Language and Speech Conference,
University of Melbourne.
Blackman, John. 1990. John Blackman’s Australian Slang Dictionary. South Melbourne:
Sun Books.
Bradley, David and Maya Bradley. 1985. “The phonetic realisation of a morpheme boundary in Australian English”. In John Clark, ed. The Cultivated Australian: Festschrift in
Honour of Arthur Delbridge. Beiträge zur Phonetik und Linguistik 48: 333–340.
———. forthcoming. Changing Australian English.
Eisikovits, Edina. 1987. “Variation in the lexical verb in Inner-Sydney English”. Australian
Journal of Linguistics 7/1: 1–24.
Finch, Christine. 1982. “Subjective reaction to rural and urban Victorian speech”. University of Melbourne Working Papers in Linguistics 8: 23–26.
Harris, Max. 1983. “Oz speech: it’s a horsepiddle case”. Weekend Australian Magazine 6–
7 August: 4.
———. 1989. The Australian Way with Words. Port Melbourne: William Heinemann
Hornadge, Bill. 1980. The Australian Slanguage. North Ryde: Cassell Australia.
Keesing, Nancy. 1982. Lily on the Dustbin. Slang of Australian Women and Families.
Ringwood: Penguin.
Labov, William. 1972. Sociolinguistic Patterns. Philadelphia: University of Pennsylvania
Manne, Robert. 1998. The Way We Live Now. The Controversies of the Nineties.
Melbourne: Text Publishing.
Mitchell, Alexander George and Arthur Delbridge. 1965. The Pronunciation of English in
Australia. Revised edition. Sydney: Angus & Robertson.
Morris, Edward E. 1898. A Dictionary of Austral English. London: Macmillan.
Peterson, Alan. 1986. Words Words Words. The Use and Misuse of English in Australia
Today. Revised Edition. Sydney: John Fairfax.
Porter, Hal. 1963. The Watcher on the Cast-Iron Balcony. Sydney: Faber & Faber.
Poynton, Catherine. 1979. “Attitudes towards Australian English”. MA Preliminary thesis,
Monash University.
Ramson, William S. 1966. Australian English. An Historical Study of the Vocabulary
1788–1898. Canberra: Australian National University Press.
Schornikow, Tamara. 1981. “Subjective Reaction Test: attitudes towards Australian speech
styles”. Unpublished Linguistics paper, University of Melbourne.
Sirianos, Patty. 1984. “Individual research project: 2SL”. Unpublished Linguistics paper,
La Trobe University.
Spitale, Josephine. 1986. “Stereotypes in Australian English”. Unpublished Linguistics
paper, La Trobe University.
Wilkes, Gerald A. 1978. A Dictionary of Australian Colloquialisms. Sydney: Sydney
University Press.
Zeccola, Lilian. 1985. “Attitudes to sociolectal variation of Australian English”. Unpublished Linguistics paper, Monash University.
A. G. Mitchell and the development of
Australian pronunciation
Colin Yallop
Macquarie University
A. G. Mitchell became well known from the 1940s for his description and
defence of Australian English. In 1946, at a time when many Australians
believed there was no such thing as Australian pronunciation, or, if there was,
it deserved only to be suppressed in favour of “correct” pronunciation, he
published his Pronunciation of English in Australia (Mitchell 1946). In the
1950s and 1960s, as the McCaughey Professor of Early English Literature and
Language at the University of Sydney, he continued his interests in Australian
pronunciation, collaborating with Delbridge in a revision of his 1946 work
(Mitchell and Delbridge 1965a) and in a survey of the speech of young
Australians (Mitchell and Delbridge 1965b). The survey remains a benchmark
for empirical studies of Australian speech patterns and Mitchell’s name remains well-known, both for his descriptive work and for his advocacy of
Australian English as a legitimate variety of English (see Delbridge this
From 1965 until his retirement in 1975 he was the founding ViceChancellor of Macquarie University, and, while he gave support and encouragement to Delbridge and the other members of the team preparing the first
Macquarie Dictionary, he had little opportunity to continue his own research
and writing. But in his 80s, he came back to Australian English with renewed
vigour, reading widely and voraciously in Australian history, demography and
sociology, taking particular account of work published in the 1980s and
1990s, and participating in seminars at Macquarie University.
Mitchell gave an interim account of his reflections on this reading in a
public lecture at Macquarie in October 1993 (Mitchell 1995). At the time of
his death, in 1997, he was still working on a much larger manuscript on the
origins and development of Australian English. He had generously indicated
that he trusted others to edit and publish this work and, while that work of
assembling and editing his notes is in progress, this chapter is an attempt to
summarise the main features of his recent work, particularly in relation to
Mitchell’s perspectives
Mitchell’s interest in historical geography and demography is evident in his
1993 lecture, both explicitly in his comments about the rewards of multidisciplinary research (Mitchell 1995: 2) and more generally in the extent to which
his lecture cited various kinds of population statistics (Mitchell 1995: 5, 8,
13–4, 17–8, 21–3, 25, 27). He was keen to base any hypotheses about the
origins of Australian speech on the most accurate data available.
Mitchell pursued sources of information about the convicts and their
origins, and about the children of the colony. He was disturbed by some of the
popular overgeneralising about the depravity of the convicts and the colony,
and in his appeal for a more balanced understanding of Australian history
there are echoes of his earlier spirited defence of Australian English against
unreasonable accusations of ugliness and slovenliness.
Mitchell also looked at evidence of mobility, both of the contemporary
population of the British Isles and of ex-convicts and immigrants in Australia.
He came to the view that free immigration was a significant factor from the
1830s, relevant to the emergence of General Australian as the accent of the
Finally, Mitchell read widely about the development of various kinds of
schooling and training in Australia. He concluded that education has always
been of concern in Australia, even if not always effectively or equitably
implemented. Where education follows British (particularly English) models,
especially from the late 19th century, when Received Pronunciation begins to
be firmly established as “educated pronunciation”, the conditions favour a
form of speech accommodating towards RP.
The following sections will deal briefly with each of these topics.
The convicts and their origins
There has been considerable discussion of the nature and origin of distinctively Australian pronunciation (see for example Mitchell and Delbridge
1965a: 20ff., Bernard 1969; Blair 1975; Bernard 1981; Horvath 1985: 29ff.,
Trudgill 1986: 129ff.). There is some difference of opinion about the possible
co-existence of different pronunciations in early New South Wales, and about
the appropriateness of terms like “mixing bowl” and “homogenisation”, but
there is general agreement
(1) that Australian pronunciation is relatively similar to forms of speech
found in southeastern England (and relatively dissimilar from, say, forms of
speech characteristic of southwestern England, northern England, or Scotland
or Ireland); and
(2) that a distinctive Australian accent (or accents) must have been
emerging quite early in New South Wales.
Mitchell does not take issue with either of these points, but looks to historical
and demographic data to flesh them out. Most of the earliest immigrants were
of course convicts. In all, about 123,000 men and 25,000 women were
transported to New South Wales and Van Diemen’s Land over more than fifty
years (Robson 1994: 8). When the first census was taken of the civil population of New South Wales in 1828, the composition of the total adult population of 30,822 was as follows (Borrie 1994: 30, 32–3):
in bondage
free by pardon or servitude
born free
arrived free
15 668
7 530
3 503
4 121
Children under 12 are not included in these figures, but we will return to them
in Section 4 below. For the moment we note that convicts and ex-convicts
dominate the civil adult population in 1828.
Industrialisation and urbanisation in general, and the pre-eminence of
London in particular, had already had, and were continuing to have, a profound
effect on British society. The convicts came from all over the British Isles, but,
according to Shaw, “the majority of English convicts sent to Australia came
from the cities, especially from London and Middlesex, and as time went on,
from the growing industrial towns of Lancashire” (Shaw 1966: 151). Of those
transported from England, “about a third came from the London metropolis
until 1819, about a quarter in the 1820s…” (Shaw 1966: 152). Nicholas and
Shergold (1988a) examine the birthplace records of nearly 20,000 convicts
transported to New South Wales (admittedly covering 1817–40, not the earliest
period) and compare the distribution of convicts’ origins with the distribution
of the British population in the comparable period. They conclude that most of
the convicts came from “the heartland of England” and from eastern Ireland: in
England, the (urbanised) counties of Middlesex and Warwickshire are overrepresented in convict birthplaces, while rural or “fringe” counties like Cornwall
and Cumberland are underrepresented; and in Ireland western counties like
Clare, Galway and Donegal are underrepresented, while Dublin, with less than
5 per cent of Ireland’s population at this time, accounts for a quarter of the Irish
convicts. “About 45 per cent of all convicts were born in urban areas; 71 per cent
of the Scots, 48 per cent of the English and 34 per cent of the Irish” (Nicholas
and Shergold 1988a: 46).
According to Robson’s calculations, 17 per cent of men transported had
been tried in London, 7 per cent in Lancashire and 5 per cent in Dublin; none
of the other counties of Britain and Ireland accounts for more than 4 per cent
and dozens are represented by less than 2 per cent (1994: 155). The figures
for female convicts and their places of trial are even more concentrated on
London: 20 per cent of female convicts were tried in London, 8 per cent in
Lancashire, 7 per cent in Dublin, and the rest were again spread over many
counties represented by 4 per cent or less (Robson 1994: 162).
Mitchell does express some reservations about the notion of a “mixing
bowl”. The figures we have just quoted suggest a picture of a population
which is diverse in origin, but predominantly urban, with by far the largest
minority coming from London and southeast England. We do not know as
much as we would like to about the pronunciation of English in the British
Isles in the late 18th and early 19th centuries, and we certainly do not know
exactly how Londoners and people from other urban centres spoke at that
time, but it is indisputable that regional diversity was still strong in the British
Isles and it is reasonable to suppose that a rich variety of accents could be
heard among the convicts. Many of them would have spoken “old” rural
dialects, each of these represented by only a very small minority of speakers;
some of them would have spoken the newer emergent urban dialects; probably
some of the Irish were bilingual in Irish and a variety of (Irish) English (cf.
Horvath 1985: 40). Only London English (or Middlesex or southeast urban
English, however it might best be characterised), with a proportion of speakers
perhaps somewhere between 10 and 20 per cent, could have stood out as a
major dialect.
Now we know that many adults accommodate to other dialects when
exposed to them but that there are also limits on such accommodation
(Trudgill 1986: 31ff.). And to the extent that adult convicts in New South
Wales did accommodate, the only form of speech that was represented substantially enough to seem like a model to accommodate to, was that of the
London area. No doubt London would also have been the subject of some
admiration or respect, whether as the capital city or as the largest city by far in
the British Isles or as a paradise for pickpockets (cf. Shaw 1966: 40). But
whether adult convicts made any lasting or radical accommodation of their
speech or not, it is the children born in New South Wales or arriving at a very
young age who have no other model than what is spoken around them. It is
they who are in any real sense the first Australian English speakers.
The developing colony
These children of the early colony grew up in a settlement where convicts
were the majority and kept arriving. Transportation to New South Wales
continued until 1840 and to Tasmania until 1852. Even when transportation
ceased, it took years for the last arrivals to win a pardon or serve their time.
But convicts steadily did win pardons, conditional or unconditional, and
did complete their sentences. By 1821 emancipists were playing a significant
part in the economy of the colony. A petition presented to Governor Macquarie
in 1821 claimed that emancipists had large holdings of land, a total of 29,028
acres under cultivation and 212,235 acres in pasture, compared with free
immigrants’ holdings of 10,787 acres of land under cultivation and 198,369
acres in pasture (Borrie 1994: 34–5). As Borrie points out, these figures were
compiled by emancipists but, even allowing for exaggeration, the contribution
of ex-convicts was impressive.
Indeed, the colony was never simply a prison. Officers of the New South
Wales Corps, for example, were not just jailers or soldiers. Whatever the
official view, they were also businessmen, perhaps opportunistic and greedy
entrepreneurs, but nonetheless contributors to the economic development of
the colony. Moreover, the convict system itself required a major investment
not just in bureaucracy and buildings, but also in shipping and victualling and
other services. And most convicts were workers, bringing some trades and
skills to the colony or at least contributing their labour to public works such as
roads and harbours (Nicholas and Shergold 1988b).
As we have already noted in the previous section, convicts and exconvicts formed the majority of the adult population recorded in the 1828
census. But there was now also a substantial number of children under 12.
Whatever the parentage and economic circumstances of these children, their
civil status was of course that they were free. Thus convicts were actually less
than half the population of New South Wales in 1828, and even the total of
convicts and those now pardoned or freed amounted to less than two thirds of
the total civil population. An amplified version of Borrie’s figures from the
1828 census (1994: 30, 32–3) is as follows:
ex-convicts (pardoned or freed)
born in New South Wales (adults over 12)
arrived free (adults over 12)
children under 12
15 668
7 530
3 503
4 121
5 780
In this environment, children must have spoken something that was similar,
though probably not identical, to the contemporary speech of London or
southeast England or at least to the “accommodated” version of it that might
be heard most commonly among the adults in the colony (cf Blair 1975: 25–6,
Trudgill 1986: esp.135ff.). For Mitchell, this is the earliest form of Broad
Australian, emerging at a time when New South Wales was still quite isolated,
with a population essentially being built up by a one-way process of transportation, but with significant numbers of children being born in the colony.
The “coarse intellectual clay”
Mitchell deplores sweeping generalisations about the coarseness of this early
society, generalisations that too easily encourage a view that Australian
speech was cradled in vice and vulgarity. He examines recent evidence about
the skills and abilities of the convicts and the early colonial children, and
rejects some of the more emotive comments about the inferiority of the
convicts and the children born in the colony.
Hughes, for example, describes the convicts as “runts” (Hughes 1987:
174) and refers to “the coarse intellectual clay of Sydney” (Hughes 1987:
179). But Nicholas and Shergold (1988b: 78ff.) refute Hughes’s use of the
word “runt”: it is not at all surprising that early 19th century convicts were
shorter than the current population of Australia, and the relevant comparisons
are contemporary ones. After examining the records carefully, Nicholas and
Shergold conclude that in fact “the convicts were not short by contemporary
British and Irish standards” and they were “of the same stature as the recruits
into the Marine Society and the British army in the early nineteenth century”
(1988b: 82).
Nicholas and Shergold also examine literacy rates. The evidence is
detailed and complex and cannot be adequately summarised here, but Nicholas and Shergold’s general conclusions are that the literacy rate of Irish
convicts was comparable to that recorded in the Irish census of 1841; and that
the English convicts (of whom perhaps as many as three quarters could either
read or read and write) may actually have had a higher level of literacy than
the population of England at the time (Nicholas and Shergold 1988b: 75). The
greater proportion of urban people among the convicts may contribute to this.
In the light of such evidence, Mitchell argues that the convicts were
people with a respectable rate of literacy, many of whom possessed skills as
farm workers, tradesmen or service workers. As for the children, Mitchell
suggests that while there were clearly appalling conditions for many of the
children in the early colony, so there were by today’s standards in many other
parts of the world, including Britain and Ireland. These conditions do not
automatically imply that convicts and ex-convicts were incapable of stable
family life or that they did not care for the welfare and future of their children.
Various measures to establish education for children and care for “orphans”,
memorials sent by convicts to the governor, and efforts to win apprenticeships
for their children, do not justify a melodramatic picture of a totally irresponsible and depraved society. Oxley’s survey of female convicts transported to
New South Wales suggests that they were “quite highly skilled and highly
literate members of the working class” (1988: 94) and she dismisses the old
claim that they were “members of a professional criminal class” (1988: 95).
Robinson gives some detail of apprenticeships in early New South Wales
(1985: 122ff., 140ff.) and offers some examples of apparently successful
marriages between native-born women and ex-convict men (1985: 163–4).
Such evidence is to some extent anecdotal and Robinson recognises that there
were also some social disasters, but the evidence is enough to indicate the
existence of relatively stable families in which children would have grown up
to be responsible and productive.
Mitchell does not argue that the speech of these native-born in early New
South Wales must have been identical to modern Broad Australian, any more
than it must have been exactly the same as some form of English spoken in
London or southeast England. But nor was it some lowest common denominator produced from a “mixing bowl” of dozens of different dialects. It was the
result of, firstly, a process of eliminating most of the small minority dialects in
favour of the most common minority of speakers (from London and southeast
England), secondly processes of accommodation which are now beyond our
observation but which cannot have had radical effects on a basically southeastern English form of speech, and thirdly the emergence of the native-born
as a significant new group.
Mobility in the British Isles
There was considerable mobility of the population in the British Isles well
before transportation to New South Wales began. The agrarian revolution, the
early stages of the industrial revolution, the clearances and dispossessions in
Scotland and Ireland, all contributed to major movements of people seeking
work. We have already seen that the convicts came from many parts of Britain
and Ireland, but they were not only of diverse origins, they were relatively
mobile. Nicholas and Shergold’s figures, for example, show that in their
sample, 58 per cent of the Scottish convicts had left their county of birth
before their trial, and likewise 57 per cent of the Welsh, 41 per cent of the Irish
and 35 per cent of the English convicts (Nicholas and Shergold 1988a: 54ff.).
Some of them moved considerable distances. Confirming the substantial
movement from rural into urban areas (for example from Ireland into urban
centres in England and Scotland), Nicholas and Shergold make a detailed
analysis of their sample and show that of the English convicts who had moved
before their conviction, more than half had moved 70 miles or more, and more
than one quarter 140 miles or more from their county of birth (1988a: 54–5).
It was this mobility that was already contributing to adjustments or
accommodations in the dialects of Britain (cf. Trudgill 1986 on linguistic
“urbanisation” and “koineisation”), another reminder that we should be cautious about assuming that we know how people were speaking in, say, London
or Manchester in 1800. But another point about this mobility is that it may
have prepared people for Australia, both by making them more adaptable in
general and by preparing them for mobility within Australia.
Mobility in Australia
In the earliest years of New South Wales, mobility was of course constrained.
While travel by sea was possible, there was no route through the Blue
Mountains until after 1815. But once people had access inland from Sydney
and from later settlements such as Port Phillip, mobility within Australia was
extraordinary, and full of what Mitchell calls “bumpiness”, created by rapid
pastoral expansion and short-lived but dramatic phenomena such as goldrushes. Net overseas immigration into Victoria in the period 1851–60 is
estimated at around 337,000, in itself an impressive movement of people; but
in roughly the same period the population of the colony rose by around
440,000, from 97,500 in 1851 to just under 540,000 by 1861. Some of this
increase must of course be ascribed to births, but given the disproportionate
numbers of single males, it is clear that large numbers were moving into
Victoria from elsewhere in Australia (Borrie 1994: 68–9). The effect of this
on Tasmania’s economy was disastrous: in two years, 1851–2, Tasmania lost
over a third of its non-convict male labour force, principally to the goldfields
(Borrie 1994: 80–1).
A later and equally dramatic example is what happened to the population
of Western Australia after the discovery of gold in that colony. Western
Australia had by 1861 reached a total population of only 15,936. In 1881 the
population reached 29,708 and in 1891 49,782. But in the next ten years, as the
gold drew fortune hunters, the population more than trebled to 184,124 in
1901 (Borrie 1994: 150).
This mobility is a significant factor in the absence of major regional
variants in Australian pronunciation. Victoria, for example, grew rapidly and
drew large numbers from New South Wales and Tasmania (many of whom
were presumably already speakers of an “Australian English”) as well as from
overseas. And in Western Australia, even if a local accent had emerged by the
1880s, it would have been swamped by the massive numbers arriving in the
1890s from the eastern colonies, particularly Victoria.
The significance of free immigrants
There were some free immigrants from the earliest years, but it was after about
1830 that their numbers and impact become significant. We have already
mentioned the more than 300,000 immigrants entering Victoria from overseas
in the ten years from 1851. Even before that, according to Sherington (1980:
114), the number of free immigrants arriving between 1829 and 1850 exceeded the total number of convicts arriving between 1788 and 1841. Borrie’s
figures for New South Wales show that by 1837 the number of immigrants
(3,477) exceeded the number of convicts landed (3,425). After 1837 only a
few thousand more convicts landed in New South Wales, the last of them
arriving in 1841. Free immigration had become the major source of arrivals
(Borrie 1994: 50).
The significance of free immigration is not only in the growing numbers
but also in the arrival of more women and more families. Despite the waves of
adventurous single males attracted by gold strikes, there were also immigrants
seeking to engage in trade and commerce, many of them already married and
bringing young children, most of them by no means illiterate paupers. Relatively stable family life, already beginning in an essentially ex-convict society,
was further strengthened, contributing to a society in which natural increase
played a major part.
At the same time, the 19th century sees what Borrie calls “the evolution
of the unique human being”: improvements in hygiene and the control of
infectious diseases begin to lower the mortality rate dramatically, while the
birthrate declined towards a no less unprecedented low level (Borrie 1994:
303ff.). As Mitchell puts it, this “unique human being” was now able to delay
death and control fertility.
Thus immigration (including both convicts and free settlers) accounts for
75 per cent or more of population increase until about 1860. In the latter part of
the century it was natural increase that accounted for most of the population
increase. (For comparison, in the 1930s, natural increase had risen to account
for around 95 per cent of Australia’s population growth, although the higher
levels of immigration after the Second World War pushed this percentage
down again.) Selected figures for the Australian-born and foreign-born (pre-
dominantly British and Irish in these statistics) are as follows (Borrie 1994:
born in Australia
428 954
1 422 533
2 913 997
born outside Australia
723 152
827 541
859 904
1 152 106
2 250 074
2 913 997
It is in this environment that Mitchell sees General Australian emerging,
becoming the majority accent perhaps between 1870 and 1890. Among the
great numbers of British and Irish immigrants, coming, like the convicts, from
various parts of the British Isles, there would have again been processes of
accommodation, in the context of awareness of southeastern English and
urban “middle class” norms that were already different from those of the late
18th century. And in Australia itself, urbanisation and family life and middle
class aspirations would have exerted their own influence. Given the evident
predominance of General Australian in the latter part of the 20th century,
Mitchell doubts whether it could have been somehow spun out of Broad to
become a majority form.
Mitchell notes (again drawing on figures from Borrie 1994: 74ff.) that
according to the 1857 census of Victoria only 17% of the colony’s population
were born in Victoria and 4% elsewhere in Australia. By 1861, the percentage
of Australian-born had risen to 30%, but those born in the British Isles still
constituted a large if temporary majority (59%). Many of the British-born
were moving into the suburbs of Melbourne, either from the goldfields or
directly from Britain, and in some of these suburbs they reportedly outnumbered the native-born by two to one. In the 1861 census, the Melbourne urban
area is recorded as having 127,000 people (already substantially bigger than
Sydney and beginning a sustained period as Australia’s premier city). In the
Melbourne suburbs there may have been fertile ground for a contrast between
a new and rapidly increasing General Australian and an older Broad. The two
forms (or 19th century versions of them) may have co-existed in these circumstances, with a tendency for General to be seen as new, urban and prestigious
and Broad as older and conservative, but without a sharply defined regional or
social justification for the contrast. If this is true, it may help to explain firstly
why there is still relatively easy exchange among the features of the two forms
(such that a particular speaker may for example have mostly Broad features
but one or two diphthongs that are strikingly General) and secondly why there
is still a tendency (though no more than that) to associate Broad with rural
male speakers.
Education, cultivation
We have already mentioned Mitchell’s contention that, notwithstanding the
indisputable harshness of life in early New South Wales, there is evidence of
some concern for the welfare and education of children. Governor Phillip had
been told to reserve land for a church and school in every township. Even in
1788 there were thirty-six children in the new settlement, seventeen of them
children of convicts and nineteen children of marines, and as early as 1789 the
Reverend Richard Johnson arranged for a convict woman to conduct a school
in a hut in Sydney, and in 1791 he put another convict woman in charge of
school in Parramatta (Barcan 1980: 9). It would be difficult to argue that
education was dear to the hearts of all of the early arrivals — the burning down
of the colony’s first church and schoolhouse in 1798, for example, was taken
as evidence that the convicts were a wicked and abandoned lot (Shaw 1966:
76–7) — but the colony was not without some concern for education and
The first private school was opened at Parramatta in 1800 and the first
boarding school, also in Parramatta, in 1804 (Barcan 1980: 15). In 1801
Governor King opened an institution for “orphan girls”. His wife is said to
have attended daily and to have acted as a “matrimonial agent”, finding
approved husbands for the older girls (Kociumbas 1992: 30). By March 1803
this Female Orphan School was providing education to 54 girls aged from 7 to
14 (Barcan 1980: 13). While much of this “education” was no doubt simple
and rough, it did train some young people in useful skills, and it does
demonstrate concern and activity, qualifying the notion that the colony was
totally without aspirations to culture and civilisation.
By the 1830s private schools were being established on the model of the
British corporate schools or colleges, the first of them being the Australian
College, opened by the Presbyterian Reverend J. Dunmore Lang in Sydney in
1831, and the King’s School opened by the Head of the Church of England,
Archdeacon Broughton, in Parramatta in 1832 (Barcan 1980: 47). There
followed a variety of schools, from so-called “private venture schools” which
were popular with parents who wanted a commercial and practical emphasis,
to nondenominational corporate colleges to Roman Catholic schools, including some for girls (who might be taught “graces and refined behaviour, speech
and deportment”).
In the 1850s and 1860s, the period we have mentioned earlier as a time of
rapid development of a Melbourne version of English suburban respectability,
Australia already had English-style private schools, often under an English
headmaster. There was bitter controversy about the funding and supervision
of education (see for example Turney 1992: 55–7). But whatever its formalisation, education was often set in a context of the need to civilise, along British
— or rather more narrowly, English — lines. William Wilkins, for example,
was brought out from Britain at the age of 24 to be headmaster of the Fort
Street Model School, a teacher training school, and he was by most accounts a
remarkable and energetic man; he was also in no doubt about the “civilising
influences of a comfortable and virtuous home” or about the need “in a new
country” for good schools to “prevent the people from retrograding in civilisation” (Turney 1992: 92, 94).
Mitchell notes also some of the “colonial voices” documented by Webby
(1989). Eliza Brown writes home from Western Australia in the 1840s to her
father in England: her young son “has no idea of being a settler in this country”
and “his thoughts revert to England from whence we brought him”; Eliza
would like him “to jostle it with other boys at a public school for about four
years, Eton, Westminster, Harrow, or the High School in Edinburgh, then if
we are prosperous we might send him to College…” (Webby 1989: 121). On
a property in Victoria in 1846, a young George Gordon McCrae describes the
tutoring he is receiving: on December 24th “we went to lessons and as usual
learned a part of the French history. We read a chapter in the Bible and a piece
from Milton’s Paradise Lost. I almost finished the third book of the Aeneid…”
(Webby 1989: 124). This was a society in which Britain or England was
likely to be constantly held up as the source or medium of civilising education.
We cannot infer from this kind of evidence that Received Pronunciation
was part of the requisite civilisation at this time. British people educated
before the 1870s, even aristocrats and Oxbridge graduates, often had identifiably regional speech (Honey 1989: 24–5). It was not until around 1870 that a
substantial number of British “public schools”, mostly inspired by Arnold’s
reforms at Rugby, had achieved the necessary discipline and effectiveness to
attract and hold the children of the rich and powerful, to accommodate them as
boarders and to instil into them a deep commitment to RP as a badge of status
and education (Honey 1989: 26ff.). We need not suppose that RP was a
significant feature of educated Englishness until quite late in the century.
Thus for Mitchell it is not necessary to assume, as Bernard (1981: 20)
and Horvath (1985: 37) seem to, that there was some kind of “cultivated”
English or “proto-cultivated” Australian in early New South Wales. There
were certainly individuals with prestige and status (of various kinds) and we
have already noted that certain urban speech, particularly that of London, must
have enjoyed some status. But given that RP does not become predominant in
Britain until well after 1850, it seems likely that there was considerable
diversity of accent among even the British-born elite of the colony: Macquarie
probably had a noticeable Scots accent and Bourke possibly an Anglo-Irish
accent, for example.
It is interesting in this connection that 19th century comments on Australian English (as documented in Blair 1975) are not highly critical. There may
have been references to a “detestable snuffle” or “a Cockney drawl” but there
is little evidence of the white-hot indignation of the 1930s and 1940s, when
the ABC Weekly was recording the views of those who “suffered acutely
listening to Australian voices”, who condemned Australian speech for its
“impure vowels”, who described popular Australian speech as “cribbed,
cabined and confined” (Mitchell and Delbridge 1965a: 68–9). It is perhaps
not until the first half of the 20th century, when RP became the pronunciation
of the British boarding schools and became the obligatory accent of English
archbishops, Guards officers and BBC newsreaders, that educated Australians
felt a need to imitate RP or adapt their speech to it.
10. Conclusion
In outline, Mitchell sees Broad Australian as established in New South Wales
by the 1830s, not so much from a “mixing bowl” as by a process of local
levelling and accommodation towards norms deriving from London or the
London area. He suggests that General Australian owes much to immigration
from Britain and became established by the 1870s or 1880s in rapidly developing suburban Australia. The cultivated end of the spectrum he ascribes to
“external” influences especially in contexts where an English model of pronunciation is regarded as normative.
Mitchell certainly does not overlook his own earlier caution about distinguishing too sharply among these three varieties. Each is variable in some
details from speaker to speaker, and Australian pronunciation is a continuum,
not a set of three sharply contrastive accents (cf Mitchell and Delbridge
1965a: 15). If Mitchell is correct in his inferences, the origins and motivation
for the three varieties, each different but all emerging in circumstances of
competing norms in an evolving social structure, are part of the explanation
for the characteristic distribution and associations of Australian forms of
If the conclusions of this paper seem elusive, it may be that they await the
fuller treatment that Mitchell himself was working towards. But it may also be
that there is a subtlety about Australian speech patterns that is indeed hard to
define. At a ceremony to celebrate Mitchell’s life and achievements, held at
the University of Sydney in late 1997; Bernard Martin spoke of Mitchell and
other teachers in the Faculty of Arts at the University of Sydney in the 1950s
in terms that provide a fitting reflection at the end of this paper. These
teachers, were, in Martin’s words, “often mildly Anglophile; and yet, without
any great parade of it, they were firm in their Australian identity (the actual
phrase ‘Australian identity’ was one they didn’t use). They had seen the
metropolis; but usually they took little trouble to acquire its gloss, not thinking
that they really needed it” (Martin 1997: 3).
The words are thought-provoking, both as a tribute to Mitchell the man,
and for what they say about Mitchell’s kind of Australianness: there is some
admiration of England or Britain but a mild admiration, without any sense of
needing its gloss, there is firmness in Australian identity but without any great
wish to parade it or even identify it by name. Perhaps it is in just such
counterpoint that Australian speech gets its voice.
On the one hand, this chapter is intended as a tribute to Mitchell the scholar, on the other
it serves to draw attention to important issues in understanding the origins and development of Australian English. Pending the publication of Mitchell’s manuscript, any
inadequacies in this paper are my responsibility and not Mitchell’s.
Barcan, Alan. 1980. A History of Australian Education. Melbourne: Oxford University Press.
Bernard, John R. 1969. “On the uniformity of Australian English”. Orbis 18: 62–73.
———. 1981. “Australian pronunciation”. In The Macquarie Dictionary, 18–27
Blair, David. 1975. “On the origins of Australian pronunciation”. Working Papers of the
Speech and Language Research Centre, Macquarie University July 1975, 17–27.
Borrie, Wilfred D. 1994. The European Peopling of Australia. A Demographic History
1788–1988. Canberra: Research School of Social Sciences, The Australian National
Corcoran, Kris and Stephen Nicholas. 1988. “Statistical Appendix — Convicts transported
to New South Wales 1817–40”. In Nicholas 1988: 202–24.
Honey, John. 1989. Does Accent Matter? London: Faber & Faber.
Horvath, Barbara M. 1985. Variation in Australian English. The Sociolects of Sydney.
Cambridge: Cambridge University Press.
Hughes, Robert. 1987. The Fatal Shore. A History of the Transportation of Convicts to
Australia 1787–1868. London: Collins Harvill.
Kociumbas, Jan. 1992. The Oxford History of Australia Volume 2. 1770–1860 Possessions. Melbourne: Oxford University Press.
Martin, Bernard K. 1997. Address at a ceremony to celebrate the life and achievements of
Alexander George Mitchell, The University of Sydney, 3 November 1997.
Mitchell, Alexander G. 1946. The Pronunciation of English in Australia. Sydney: Angus &
———. 1995. The Story of Australian English. Sydney: Dictionary Research Centre,
Macquarie University.
——— and Arthur Delbridge. 1965a. The Pronunciation of English in Australia (revised
edition). Sydney: Angus & Robertson.
——— and Arthur Delbridge. 1965b. The Speech of Australian Adolescents. A Survey.
Sydney: Angus & Robertson.
Nicholas, Stephen, ed. 1988. Convict Workers. Reinterpreting Australia’s Past. Cambridge: Cambridge University Press.
——— and P.R.Shergold. 1988a. “Convicts as Migrants”. In Nicholas 1988a: 43–61.
——— and P.R.Shergold. 1988b. “Convicts as Workers”. In Nicholas 1988b: 62–83.
Oxley, Deborah. 1988. “Female Convicts”. In Nicholas 1988: 85–97.
Robinson, Portia. 1985. The Hatch and Brood of Time. A Study of the First Generation of
Native-Born White Australians 1788–1828. Melbourne: Oxford University Press.
Robson, Leslie L. 1994. The Convict Settlers of Australia. 2nd edition. Carlton: Melbourne
University Press.
Shaw, Alan G. L. 1966. Convicts and the Colonies. A Study of Penal Transportation from
Great Britain and Ireland to Australia and other parts of the British Empire.
Melbourne: Melbourne University Press.
Sherington, G. 1980. Australia’s Immigrants 1788–1978. Sydney: Allen & Unwin.
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Basil Blackwell.
Turney, Clifford. 1992. William Wilkins. His Life and Work. Sydney: Hale & Iremonger.
Webby, Elizabeth, ed. 1989. Colonial Voices. St Lucia: University of Queensland Press.
Lexicography and national identity:
the Australian experience
Arthur Delbridge
Macquarie University
The concept of nationality must be very present in the minds of today’s
Australian lexicographers, because the two most important dictionaries of
Australian English (henceforth “AusE”) published in the last fifteen years
both claim on their title pages to be the Australian national dictionary. Now
this may seem simply to reflect a crass commercial motive on the part of
publishers anxious to promote their product, and there could be some puzzlement among users over which to choose. But in fact they are both genuine and
supportable claims, though different.
The Australian National Dictionary
Edited by Dr W. S. Ramson for the Oxford University Press, and published in
1988, the Australian National Dictionary (henceforth “AND”) has as its
subtitle A Dictionary of Australianisms on Historical Principles. It has a word
list of about 6000 main entries, many of them with subdivisions, in the Oxford
fashion, especially those forming simple combinations and collocations that
do not require further definition and other more special ones that do, each
entry judged to be distinctively Australian. The entries consist mainly of
citations drawn from the close reading of nearly ten thousand books and
papers having Australian associations, from the time of the first British
settlement in the country (1788) up to the present. The generous provision of
citations from this reading provides three sorts of information: they establish
the chronology of each word’s use, substantiate the definition, and illustrate
the range of registers within which the word has been used. There were
difficulties, of course, in establishing strong claims for a distinctly Australian
origin for many words, even with citations that antedated those of British or
American dictionaries. It therefore seemed best to the editor to interpret the
concept of Australianism liberally, “not making undue claims but including
many words which are of undoubted significance in the Australian context but
about the precise origin of which there remains uncertainty.” (Ramson
1988: vi). One of the most interesting features of this dictionary is the
inclusion of some four hundred words borrowed from Aboriginal languages,
all with an etymology supplied, usually identifying the source language. This
had never before been attempted comprehensively in Australian lexicography.
Since at the time of the first settlement in Botany Bay in 1788 there must have
been more than two hundred Aboriginal languages in use, the editor’s ability
to identify the source language of Aboriginal words taken into English usage
reflects the fairly advanced state of present knowledge of the patterns of the
early contact between the indigenous people and the new settlers. It is now
commonly acknowledged that socially and politically it was not a contact that
has reflected much credit on the settlers or their administrators. But it was the
period when Aboriginal contact was lexically most productive.
Since this is a dictionary of origins, recording the first known use of
individual Australianisms, it also identifies the origin of many Australian
institutions, practices, beliefs and national attitudes. It records the new names
that had to be found, especially by the early settlers, for the unique features of
this antipodean landscape, and its local flora and fauna. This need is referred
to by Australia’s first lexicographer, Edward E. Morris, writing in 1896, in a
dictionary called Austral English: “there never was an instance in history
when so many new names were needed…for never did settlers come, nor can
they ever again come, upon Flora and Fauna so completely different from
anything seen by them before.” The need is one that persists and will persist,
for the preoccupation of Australian historical lexicography is still the recording the names of Australian novelties, whether flora, fauna or other. And the
need applies especially to the “other”: whatever new thing becomes part of
what is identified as part of the Australian way of life, whether originating in
Australia or not. Historical lexicography maps the nation as it changes lexically: every page of AND tells the reader something about Australian history
and the Australian way of life.
Let us consider just one example: dole (-bludger). The distinctively
Australian sense of dole emerged only during the Depression of the 1930s. It
referred to unemployment benefits, and it appeared in many combinations:
dole-ticket, dole-workers, dole stations, dole towns. One of the combinations
which developed later but is still current is dole-bludger, referring to one who
exploits (or is perceived as exploiting) the system of unemployment benefits
by avoiding gainful employment. Bludger and its back-formation verb bludge
came into AusE, according to Ramson, as a survival of British slang bludger,
a shortened form of bludgeoner. It has had a range of senses: cadging,
scrounging, borrowing but not returning, being idle especially at someone
else’s expense, living off the efforts of others. Dole-bludger, among those
who use the expression, projects an unfavourable view of some of the unemployed, especially if young, who expect the dole to maintain them in idleness.
Naturally there are others in the community who see the matter differently and
have no use for the expression. The dictionary simply records the usage with a
good range of citations, eight of them dated between 1976 and 1986. For
bludger alone, it has about 40 citations, grouped around five main senses, and
from 1856 to 1982.
Not surprisingly, AND has become an indispensable tool for historians
and literary scholars, and to a lesser extent a source of interest among the
general population. Accordingly, it is easy to justify its name as the national
dictionary, though some have suggested that since the work of producing it
was done in the Australian National University, in Canberra, the Australian
National Capital, that was justification enough.
The Macquarie Dictionary
The Macquarie Dictionary was written at Macquarie University, in Sydney,
first published in 1981, and now in its third edition (1997). Its title page
describes it as “The National Dictionary”. The justification for this is that, at the
time work began on the Macquarie, in 1970, it was quite obvious that
Australians wishing to consult a dictionary of English had available to them
only dictionaries which were focused on the use of English in the northern
hemisphere — dictionaries such as the Oxford English Dictionary, the Concise
Oxford, Webster’s, and the Random House. After two hundred years of the use
of English in Australia there was still no dictionary that focused on the
Englishes of Australia, and particularly on AusE, which is the national though
not the official language. The dictionaries then available in Australian
bookshops gave information based on English and its patterns of usage in
England and America. In those dictionaries people might certainly find scattered and peripheral references to some of the oddities of Australian terminology, such as that the Australian word for a ranch, or perhaps farm, is station, as
in cattle station, sheep station. Definition 5e for station in Webster’s Third
International says simply Austral: RANCH. It offers no explanation at all of
how large grazing properties in Australian came to be called stations. It fails to
record important combinations such as mission station, Aboriginal station,
stock station, out-station, right of station, home station, and station rouseabout,
which between them go some way towards providing an account of Australian
rural life. Clearly Webster’s is not the place for this to be done: the word
International in its title must be taken to mean “American, with occasional brief
reference to the English usage of some other countries.”
Undoubtedly, then, there was a need for a dictionary that would serve
Australians, as some of these foreign dictionaries had so imperfectly done, as
a first-port-of-call reference book, for any and all the words that Australian
users might want to look up in the course of their education, their occupation,
their leisure reading. But it had to be focused on the usage of the Australian
community itself, rather than on some other community. It had to be a
dictionary not simply of Australianisms, like AND, but a dictionary suitable
for widespread Australian use, with a comprehensive word list in which all the
pronunciations, all the spellings and all the definitions of meaning would be
taken from the use of English in Australia, and in which AusE itself became
the basis of comparison with other national varieties of English. Its publisher
would not try to sell it in other English-speaking countries, except to people
who were interested in it as an Australian dictionary, in which for example
they could find explanations of words that turned up in their reading of
Australian books, especially novels, poems, and plays that represented Australian perceptions and the intimacies of Australian life, in Australian idiom.
Since this dictionary would be holding up a mirror to Australian usage both in
speech and writing it would give something like equal coverage to the whole
range of registers found in community life: formal, informal, colloquial, slang,
standard, non-standard, even crass and vulgar (at least so far as in a general
dictionary the aim is to meet the interests of that part of society that actually
buys dictionaries). It would of course be an international dictionary in the
sense that English is an international language, with different communities
sharing the common resources, lending and borrowing as need or fancy
dictated. If there were Australianisms in it they would not be labelled as
Australian, any more for example than one would expect the Oxford dictionaries to label words of English origin as Brit..
The third edition of the Macquarie Dictionary has some 2500 pages and
about 100,000 headword and secondary headword entries, including around
20,000 encyclopaedic entries, mostly for names and places, ranging across the
whole world, though with a special interest in those with Australian associations. For these the admission criteria are somewhat lowered, in order to give
local names and words a degree of preference. For example, for the names of
Australian cities and towns the main criterion for entry is a resident population
of just 1000 people, whereas for cities and towns outside Australia it is
100,000. The population of Australia is small, around 18 million, but it is very
much an urban population concentrated in a small number of quite large cities,
like Sydney and Melbourne, and spread out over a large number of rather
small towns. Neverthess the international focus of encyclopaedic entries is
strong: if you want to check the spelling of the name of the last prime minister
of Indonesia, or of Italy, or to know what might tempt you to visit Assisi, it is
there, in its proper alphabetical place. Similarly with spelling, variants are
readily entered, and whenever two spellings seem to enjoy about equal popularity in Australian usage, they are entered as equal, though some priority is
given to any that may be distinctively Australian. For example, in words such
as sympathise, which is commonly spelt either with an -ise or -ize, both
variants are entered as acceptable, but with priority of entry to the -ise form,
largely because that is the form that has for many years been recommended in
the Style Manual published by the Australian Government Publishing Service.
It is, if you like, an Australianism. Ordinarily, however, the criterion for
entering variants is simply the level of currency in Australian writing. So
equal standing is given to honor and honour, but since in the corpus count of
newspaper and book publications honour has the edge on honor it is given
priority in the order of listing. Moreover, the variant put first in the headword
list is the form that is also used throughout the text of the dictionary.
Since the phonology of AusE differs from that of British and American
English, the Macquarie also became a pronouncing dictionary, with all its
pronunciations given in a version of the International Phonetic Alphabet.
Every headword and many secondary headwords are given a phonemic tran-
scription, together with significant variants. Throughout the Australian community there is a spectrum of allophonic variation, which applies particularly
to the vowel system, whose sociological and regional distribution has long
been identified and analysed. However in the Macquarie this spectrum of
difference is only tacitly recognised in a way that lets any Australian users
interpret the phonemic symbols of a given pronunciation in terms of their own
habitual place in the vowel spectrum. Thus the single transcription for beat
could be interpreted as [bit] if you happen to speak Cultivated AusE, but [bit]
for a General speaker, and [bəit] for a Broad speaker. No one of these is
assumed to be better than or preferred to the others.
I may by now have persuaded the reader of the Australianness of the
Macquarie Dictionary, but perhaps not yet of its national standing. When the
first edition was launched in 1981, its first print run of 50,000 copies sold out
in three months, which encouraged the publisher in his view that the population really did want its own entirely home-made Australian dictionary.
By the end of its first year it was adopted by the Australian Broadcasting
Commission as its general guide for pronunciations; soon it was adopted by
the Australian Education Commission as the guide to spelling for Australian
schools; it was taken up by the Hansard reporters of the Commonwealth
Government as its “Bible” (their word!); it came into common use in the law
courts, as the dictionary that described Australian as distinct from English and
American usage; it was adopted by a very large number of organisations,
societies, and institutes; and it continues to be the common household dictionary of choice, not necessarily in its largest format but in one of that stable of
dictionaries, the Concise Macquarie, the Pocket, the Junior, the Student, the
My First, and so on, depending on the needs of the users for whom each was
subsequently written. As these appeared so did a corresponding set based on
its companion piece, the Macquarie Thesaurus. The success was partly a
matter of timing: the Macquarie was the first, and there have since then been
other comparable dictionaries produced by international publishers that are in
some degree Australianised.
The Australianness of the lexicon
When it comes to the central task of defining word meanings, the Macquarie
and AND fulfil their national interests in different ways. For example, AND
has only one sense listed for the noun house, defined as ‘The principal
residence on a rural property…’, with many supporting citations. This is the
only sense of the word which is claimed as an Australianism. By contrast the
Macquarie lists 29 senses of the noun and the phrases and collocations with
which it is found in our various Australian corpuses. With only one or two
exceptions these are much the same senses that may be found in dictionaries
focused on other varieties of English.
Many words found in the common core of world English develop Australian senses. Recently a distinguished scientist who is not a native speaker of
English reported that he had received a note from an Australian colleague
which puzzled him, because the use of the word elitism in the note seemed not
to make sense in its context; it seemed to contradict the rest of the sentence. He
looked up a number of dictionaries, all of which gave him something like:
elitism 1a. the belief that society should be governed by a select group
of gifted and highly educated individuals b.such government
2. pride in or awareness of being one of an elite.
Then he looked up the Macquarie, and found this:
elitism 1. practice of or belief in rule by an elite
2. Consciousness or pride in belonging to a select or favoured group
3. Snobbery; undemocratic sentiment.
That clinched the matter for him. His colleague’s note now made sense, for the
dictionary had shown that the word elitism has been indigenised in Australia.
The word has gained a derogatory sense, such that in Australian conversation
a quite unflattering view of elitism may be projected. The Macquarie 25million word corpus throws more light on the words elite, elitist and elitism,
yielding 190 citations, mostly from recent newspapers. Of these citations,
- 45% related to outstanding sports people, and were not at all derogatory;
- 35% related to people and institutions regarded as pretentious, undemocratic and snobbish, and were predominantly derogatory;
- 15% related to people in institutions and at the top end of their field
(outstanding by reputation and achievement, whether intellectual, artistic,
or organisational), and were generally approbatory:
- 5% related to government by non-elected groups, mostly in countries
outside Australia, and were strongly derogatory.
This sort of derogation among Australians is related to the so-called tall poppy
syndrome, a desire to diminish in stature those people who have attained
excellence, the tall poppies of the community, who, in the view of some, need
to be “cut down to size”. So there does seem to be a characteristic Australianness in the concept of elitism; not necessarily distinctive, but clearly characteristic.
Australian English
There is still the question of what we mean by “Australian English”. The first
publisher of the Macquarie refused to have these words in the title of the
dictionary, because he believed the community already equated them with
Australian slang, which was something of an object of shame at least in the
minds of the careful people, people even mildly affected by the “cultural
cringe” which for so long has been held to be one of the determinants of
Australian social mores and attitudes. Throughout the 19th century much was
written about the English used in Australia, none of it complimentary. The
general view is epitomised in the words of William Churchill, a member of the
American Philological Society, who visited Australia in 1911 and described
AusE as “the most brutal maltreatment that has ever been inflicted on the
mother-tongue of the great English-speaking nations” (Churchill 1911: 14).
Even as they became conscious of differences between their language and that
of the mother-land, Australians tended to accept the view that theirs was a
“pretty crook” form of English, a distortion of English, distinguished only by
its ugliness and its carelessness. The view of educationists, and the local
dignitaries who gave admonitory speeches to pupils and teachers on Empire
Day, and even of the Sydney branch of the English Association in its declaration of its intellectual and literary objectives, was that something must be done
about the state of English in Australia. It is true that towards the end of 19th
century there was a surge in national feeling (prompted by the number of
adults in the population who were native-born beginning to exceed the number who had immigrated) which produced a small manifestation of interest,
even pride, in some of the words and phrases of Australian idiom. The
Bulletin, a radical weekly journal, encouraged its readers — especially those
from the country — to write in with interesting local expressions, words like
fair dinkum, larrikin, bonzer, offsider, fair cow, battler, and bludger. However
this expression of pride in the local idiom did not make much of a dint on the
general feeling that the local English was “pretty crook”. At this time the
population of Australia was almost entirely monolingual. Immigrants typi-
cally came from England, Scotland, Wales and Ireland. In 1901, when the
various states of the country joined together in federation, to become the
Commonwealth of Australia, only 5% of people in the Census of that year had
a mother tongue other than English. The Census did not include the Aboriginal people, whose various languages had had no impact on the dominance of
English. By this time the Australian accent was already well and truly established, and the novelties of the local lexicon not only entrenched but multiplying. Yet the cultural cringe about local English was deeper than ever.
The beginnings of a new attitude only emerged in the early 1940s.
Alexander George Mitchell, a young Australian scholar, just back from his
PhD program at London University, made his first attempt at analysing the
phonology of AusE. He declared that AusE (a term hitherto non-existent) was
not a careless distortion of any other variety of English, but rather a variety in
its own right, developed by perfectly normal processes in the course of the
past 150 years, and that indeed there was nothing wrong with it. However
Mitchell found it very hard to get a hearing for his words. There was an
enormous outcry, even from the language professions, speech teachers, broadcasters, theatre people, and the public at large. That man, they said, must have
rubber ears if he thinks there’s nothing wrong with the sound of our excruciating vowels! He is insulting the Australian nation because he believes that any
old thing is good enough for us. But Mitchell persisted and in the end he won.
He persuaded the Australian Broadcasting Commission to relinquish the view
expressed by the chairman of the board in 1941 that “Every quest for announcers has revealed that the number of men most suitable have been Englishmen.”
(The Sun, 11 September, 1941). Australian speech (he had said) was slightly
objectionable in its monotony and “in the throatiness and distortion of its
vowels, due to a tendency to speak with the lips and teeth closed.” (The Sun,
11 September, 1941). To combat this sort of thinking Mitchell later engaged in
a huge survey of the speech of senior secondary school pupils all over
Australia, and in his analysis he laid out the reality of the Australian accent
and its personal, sociolinguistic and regional distributions (see Mitchell and
Delbridge 1965). He initiated university teaching and research in Australian
English. He made it possible for actors to offer their directors the Australian
accent as one of the dialects in their repertoire of dialects for dramatic roles.
He encouraged others to devote some of their energies to the same cause,
resulting in theses, articles in journals and programs in general education.
Most of the academic editors of the two national dictionaries described in this
paper had been his students. Mitchell’s pursuit of his scholarly interest in
AusE ended only with his death in 1997.
During the 20th century two enormous changes overtook this formerly
monolingual country in terms of its population. One was the granting of
citizenship to people of Aboriginal lineage, and a consequent rise in the degree
to which their place in Australian life became the subject of their own urging,
and of public debate. The study of their languages also took a quantum leap.
The other change was the introduction of a major immigration program at the
end of World War II , and on to the present time. Now, at the end of the
century, almost one in three of the population of Australia were born overseas,
or are the children of people born overseas. So the Australian community has
become both multilingual and multicultural, and not just in a de facto way, but
as a consequence of government policy. In the multicultural policies adopted
by the Australian government since the mid-seventies, language has been a
central issue. In 1987, after consulting extensively with the linguistics societies of Australia among others, the Commonwealth government introduced its
National Policy on Languages (Lo Bianco 1987), with four basic tenets for
balanced development of the nation’s many languages, all of them enjoying
absolute complementarity, and all viewed as economic and educational assets
to be nurtured. The policy’s four tenets were:
– English for all,
– Support for Aboriginal and Torres Strait languages,
– A language other than English for all (through both mother tongue
maintenance and second language learning),
– Equitable and widespread language services.
In addition, there were special provisions for the support of Asian languages,
since immigration from Asia had become a major part of the immigration
process. More recently there has been a strengthened strategy to promote
literacy and language learning under the aegis of an Australian Literacy and
Language Policy. Here the word literacy means “literacy in English”, and the
word English specifically means “the form of English generally used in
Australia, Standard Australian English”. This represents a huge step forward
from the situation in the 1940s when AusE did not even exist as a concept, and
the general view was that the use of English in Australia was really misuse.
Standard Australian English
The notion of a standard for AusE is being steadily developed, with financial
support from a body responsible for developing the languages policy, Language Australia. There is an annual conference, with the name “Style Council”, directed by the Dictionary Research Centre of Macquarie University, at
which language themes are the subject of papers and discussion in groups of
cross-disciplinary people concerned with, amongst other things, the language
of law, language and the media, education, and computing. The proceedings
are widely published. A national bulletin is issued twice yearly, called Australian Style, on issues in Australian style and the use of English in Australia,
with contributed articles and questionnaires on spelling and other practices,
through which close touch is kept with the range of variables in usage and the
pattern of their distribution. There have also been a number of style manuals
dealing with Australian usage, one of them published by the Government
Publishing Service. The most recent, and the most interesting, is a corpusbased style manual, The Cambridge Australian English Style Guide. Written
by a linguist, Pam Peters, it embodies the notion that language variables
present themselves to users in different ways, and that any alleged standard
must always be a moving target.
The Australianness of any writing or speaking depends on the subject
matter, and the relation of the participants to each other. Formal relationships
may call for formal language, and formal Australian writing is not all that
different in its lexicogrammar from British English or even American English
in the formal mode. The language of the quality press in Australia would
ordinarily be acceptable for the quality press of the linguistic superpowers,
especially with a few idioms or spellings revised. But there is often a deliberate lapse into informality in (say) The Sydney Morning Herald, especially in
Even in Peters’ Style Guide there is informality in the writing: it’s likely, it’s
usual, it’s, always with an apostrophe, as if she were speaking. The best
writers make great play with informality, whether in fiction or poetry or
informative prose. Poets and dramatists, especially, like to plummet out of a
stretch of high style, down into a highly colloquial phrase, even a single word,
even a vulgar word, but certainly a slang idiom. David Foster, in The Glade
Within the Grove, a novel which won the Miles Franklin prize, has an
introductory monologue in which he writes admiringly of the old Norse sagas,
and claims that he too is writing a saga:
And though it could be claimed that the saga of which I speak has not
concluded, in that hewers of wood and drawers of water still live in that far,
remote Valley of Erinungarah, then I must retort that Icelanders live in
Iceland today, as Poms do in the Old Dart, which proves only that the body
outlives the spirit.
(Foster 1996: xxxv)
Both the styles here, the patrician prose and the slang are written tongue-incheek as if to say: “Look, I can master them both!” But ultimately this is a
feature of Australianness. As subject matter comes closer to the intimacies of
Australian life and times, the peculiarities of the idiom come into play,
necessarily. David Malouf, one of Australia’s foremost novelists, born in
Australia but not of Australian parentage, speaking in a recent broadcast,
compared English language with Australian language: the one that stayed at
home, and the one that got transported and indigenised:
Insofar as we are a people here, and insomuch as we have a culture, it is
absolutely rooted in that language. The language is what holds us together.
You know when people are always looking round for what defines our
Australian identity, or defines us as a community, or a nation, or whatever it
is, it seems to me to reside less in particular characteristics than in the fact that
we share that language with one another and have changed that language in
ways that fit us, but fit us socially rather than fit the land. That seems to me to
make the way language exists here something both more precious, because it
is the source of our cohesion as a people, but also something that we are selfconscious about in a way that a speaker of the language in England may not
have to be.
(Malouf 1996: 12)
Slang and the colloquial style are perhaps the dominant, certainly the most
notorious, element in the Australianness of AusE. In what Patrick White
called the “novels of social realism”, written in the earlier part of this century,
the attempt to capture the Australian voice relied mostly on the colloquial,
especially in fictional dialogue. Indeed it was often so overdone that what was
intended to be the colourful background to character and action became the
foreground, with the paint laid on too thick. The most comprehensive account
of colloquial Australianisms is the Dictionary of Australian Colloquialisms,
edited by G. A. Wilkes when he was the foundation Professor of Australian
Literature at Sydney University, but now in its 4th edition. It is a citationbased dictionary, with Australian fiction the source of a high proportion of the
citations. It is here that the mysteries of expressions such as don’t come the
raw prawn, kangaroos in the top paddock, brown bomber, Gloria Soame,
shout, and wouldn’t it are revealed.
The standing of Australian English
In the 1940s the concept of AusE, only so recently proposed, was greeted with
mistrust and scorn. Now at the end of the century it is fundamental to the
government’s language policy; it has been supported by a range of dictionaries
and style manuals that have gained the trust of the community; it is the variety
of English learnt by non-English-speaking migrants; it is one of the main
languages of Australian literature; and it is accepted as the expression of
Australian identity. Internationally it is noticed and acknowledged. Tom
McArthur, editor of The Oxford Companion to the English Language and of
the journal English Today, wrote an editorial to the January 1996 issue of that
journal entitled Third in the Pecking Order? (McArthur 1996: 2). The first
two places were of course given to the United States and the United Kingdom,
but of the possible candidates for the third place McArthur gives his reasons
for concluding that “the third place appears to lie elsewhere, in the recent
striking emergence of Australia”.
This is a revised version of my paper presented at MAVEN 97, an international conference on The Major Varieties of English held at Vaxjo, Sweden, 20–22 November 1997.
The conference version of the paper was included in a volume of the conference
proceedings (Lindquist et al. 1998).
Australian Government Publishing Service. 1988. Style Manual for Authors, Editors and
Printers. Fourth Edition. Canberra: Australian Government Publishing Service.
Baker, Sidney J. 1945. The Australian Language. Sydney: Angus & Robertson.
Churchill, William. 1911. Beach-la-Mar, the Jargon or Trade-speech of the Western
Pacific. Carnegie Institution of Washington Publication No. 154. Washington, Carnegie
Collins, Peter and David Blair. 1989. Australian English: The Language of a New Society.
St Lucia: University of Queensland Press.
Delbridge, Arthur, John R. L. Bernard, David Blair, Pamela Peters and Susan Butler, eds.
1997. The Macquarie Dictionary. Third Edition. Sydney: Macquarie Library Pty Ltd.
Foster, David. 1996. The Glade Within the Grove. Sydney: Random House Australia.
Lindquist, Hans, Staffan Klintborg, Magnus Levin and Maria Estling, eds. 1998. The Major
Varieties of English (Acta Wexionensia Humaniora 1) Vaxjo University.
Lo Bianco, Joseph L. 1987. National Policy on Languages. Canberra: Australian Government Publishing Service.
Malouf, David. 1996. “An Interview with David Malouf”. Australian Book Review. Sept.
McArthur, Tom. 1996. “Third in the pecking order?” (Editorial) English Today 12(2): 2.
Mitchell, Alexander G. 1940. The Pronunciation of English in Australia. A lecture privately printed for members of the Australian English Association.
Mitchell, Alexander G. and Arthur Delbridge. 1965. The Speech of Australian Adolescents:
A Survey. Sydney: Angus & Robertson.
Morris, Edward E.1898. A Dictionary of Austral English. (Second impression, 1972).
London: Macmillan, and Sydney: Sydney University Press.
Peters, Pamela. 1995. The Cambridge Australian English Style Guide. Melbourne: Cambridge University Press.
Ramson, William S. 1988. The Australian National Dictionary: A Dictionary of Australianisms on Historical Principles. Melbourne: Oxford University Press.
Turner, George W. 1992. “English in Australia”. In Robert Burchfield, ed. The Cambridge
History of the English Language (Vol. 5) Cambridge: Cambridge University Press.
Wilkes, Gerald A. 1996. A Dictionary of Australian Colloquialisms. Fourth Edition.
Sydney: Sydney University Press
Australian English in interaction
with other Englishes
Brian Taylor
University of Sydney
While in recent years there has begun to be some scholarly discussion of the
impact of especially American English (“AmE”) and to a lesser extent of
British English (“BrE”) on Australian English (“AusE”), there is a need for a
chronologically longer and linguistically broader survey of the interaction of
AusE with other Englishes, perhaps even with certain other languages such as
Gaelic, and one that examines the possibility of influences from AusE as well
as on AusE since the beginning of the colonisation of Australia by Europeans.
The present chapter will attempt such a survey along with a brief review of the
scholarly literature within the confines of the space available.1
The interaction of Englishes in the early formation of Australian
English, 1788–1820
The interaction of Englishes in Australia began as soon as the thousand or so
military, naval and convict personnel who arrived on the First Fleet set foot on
the soil of Sydney Cove on 26 January 1788. Amongst these people there were
undoubtedly speakers of a number of varieties of BrE, though mainly of
English English (“EngE”) varieties, and they were joined over the next
decades by speakers of varieties of Scottish English (“ScotE”) and Irish
English (“IrE”).
When the first really scholarly study of the varieties of AusE was under-
taken by A.G. (Alex) Mitchell over 150 years later they were found by that
time to have been not regional ones (for AusE showed then and still shows
today relatively little regional differentiation), but rather sociolectal ones,
what Mitchell called “Educated Australian” and “Broad Australian” (Mitchell
1946: passim). On the basis, however, of a huge survey of the speech of
Australian teenagers that Mitchell undertook with Arthur Delbridge around
1960, this was expanded to three sociolectal varieties: “Cultivated Australian”, “General Australian” and “Broad Australian” (Mitchell and Delbridge
1965: passim).
Mitchell and Delbridge and subsequent scholars then faced the question
of where this regionally uniform variety came from — whether largely from a
single variety of British English originally, or from the interaction of a number
of British Englishes — and when it was formed. The literature on this is
surveyed in Cochrane (1989) and can be summarised, mainly from there, as
Mitchell (1946) and Mitchell and Delbridge (1965) see AusE as developing in Australia mainly from a mixing of London English with IrE. Bernard
(1969) posits an amalgam of Englishes, dominated by London English and
developing after arrival in Australia (which he refers to as “Proto-Broad”),
and he claims that this amalgam developed in each new major centre of
settlement around the continent independently from all the rest. Blair, using
comments from early printed sources, concludes that “the amalgamation of
British dialects as a basis for Australian pronunciation was completed in the
colonies rather than in London” (Blair 1975: 26).
The other writers on the subject that Cochrane deals with differ from the
above views by seeing the accent as having already developed in Britain, with
Collins (1975) assuming a mixing occurring in the Southeast Midlands, with
Turner (1960) and Gunn (1972, 1975) seeing it as mainly London English,
and Hammarström (1980; see also Hammarström 1985) seeing the basis for
the accent as London English alone.
Cochrane himself, after reviewing all the evidence comes down on the
side of the latter group, seeing the accent as essentially deriving from the
London accent of around the late 18th and early 19th century, with only
peripheral influence from other British varieties or from developments occurring within Australia itself.
Horvath, whom Cochrane does not mention, concludes in her sociohistorical critique of the above hypotheses that in Australia’s early “extremely
polarised society” two varieties, Proto-Broad and Proto-Cultivated, emerged,
the former through the interaction of non-standard EngE varieties, with possibly an admixture of IrE, the latter out of RP (Horvath 1985: 37).
In his examination of dialect mixing in AusE Trudgill (1986: 129–46)
goes beyond accent to grammatical and lexical features, but here too influence
from other varieties such as ScotE and IrE is essentially peripheral.
So, in the majority opinion, there seems to have been little interaction of
Englishes in the laying down of the basic shape of AusE between the beginnings of European settlement in 1788 and the end of the first phase of
linguistic development around 1820.
Interactions in the middle decades of the 19th century
During the decades of the 19th century from 1820 on there was an influx of
immigrants from Britain into Australia, with the ratio of free immigrants to
convicts increasing until the transportation of convicts to the eastern Australian colonies ceased altogether in 1840. However around the middle of the
century there was also migration, often only temporary, to and from North
America and New Zealand. This was in addition to the visits made by well-todo Australians back “home” to Britain, where they may well have checked up
on the acceptability of their own speech by observing their “betters” there.2
That there were various Englishes in contact with the established form of
AusE around the middle of the 19th century can be directly inferred from what
seems to be a somewhat neglected potential source of information, namely the
newspaper reporter Charles Adam Corbyn’s reports from the Sydney Police
Court, many of which he republished in book form in 1854 (Corbyn 1854/
1970). While AusE itself seems to be reproduced in his reports in a fairly
standard form, the following brief samples suggest that a number of other
varieties are distinguishable as well. I have identified them by the name of the
person Corbyn claims to be quoting and by the variety I think it is.
Mrs O’Brien, IrE (p.43): “On Monday night my blessed husband went to St.
Pathrick’s, a’cos ’tis a taytotaler he is. I send my darter arter a pound and ahalf of pork sasingers jist to have reddy ’gainst he cumed home, for that
taytolling [sic, B.T.] work allers makes him mortal hungry;[…].”
Tommy Hopkins, London (Cockney) English (p.62): “I lodges long on
muster and mother Stewart and I seed that ere lady a getting hup stares. I wur
’pon the bed, a’having a lie down, as is nat’ral nuff seeing as how master
Stewart vos avay: cos, don’t ye know, ven the cats avay, the mices does play.
[…] she picks up vot they calls The Morning Chronicle in Lunnon […]”. (But
NB: to “Mr Henry Lookalive Sloman[,] emigrated to this colony from […]
Dyot-street, Bloomsbury-square” is attributed the form “werry” for very
Mrs Melville, ScotE (p.65): “I’m na a natif av Saydnay, and theerfur am
unable to gie ye the geographical posthun av Market-lane, but mayhap ye ken
the Sheriff, and if ye’ll just gang to him I dinna doot he’ll ensthruct ye as to
the desirable spot.”
Phillip Macedon Mealby, (American?) Black English (p.49): “The gemman
in dat [witness] box dere, wid him head and shoulder ‘tuck above de
board,[…] Habing some sovereigns and tinking de ’pearance of de coloured
gemman in de box [suspicious… (Corbyn switches to standard English)].
Day or two afterwards, little New Yorker tell him two sovereigns gone. As he
subspect dat coloured nigger, he tell him so.”
John Hilt, a “little New York Creole” (i.e. the “little New Yorker” referred to
in the preceding), AmE (p.49): “My name’s John Hilt; I’m Merican, I come
from States New York, I reckon. I’m thirteen years and a quarter old. I hang
out at old Byers’s. Now go a head [sic, B.T.], old ’un, if you’ve got any more
questions to be arter axing me, as time’s precious.”
“Mr John Chinaman”, Chinese Pidgin English (p.97): “How many years you
go givee me? me no care -------- (In fact Mr John Chinaman had picked up the
swearing portion of the British language with remarkable facility.) You givee
me five or six years, an soon as I comee out, I settle that ----------- Tartar
superhumanary. Yes and me settle dat ‘terpreter too!”
Until someone analyses them for their degree of possible genuineness, these
and the other allegedly verbatim non-standard texts in Corbyn’s reports can
only serve to show some at least of the varieties of English interacting on the
streets of Sydney around 1850, but it is perhaps significant that it is IrE and
what appears to be London Cockney English that predominate by far in his
3.1 IrE and AusE
Scholars have long been of the opinion that IrE had virtually no impact on
AusE at first, in spite of the large numbers of Irish convicts sent to Australia.
However, this attitude is beginning to change.4 The late Alex Mitchell said in
a recently published lecture: “I have long thought,[…], that linguists have paid
too little attention to the Irish […]. In relation to the many influences that made
Australian English what it is the Irish have been ubiquitous”. Mitchell went on
to point out that the Irish constituted over one quarter of those arriving in
Australia during the period of transportation, and in 1846 were the most
widespread of the four main groups — colonial-born, English-born, Irish-born
and Scottish-born — in Australia (Mitchell 1995: 2f.).
While Mitchell does not provide any actual instances of IrE influence,
attempts have been made in the last few years to find them.5 Horvath, on the
basis of a personal communication from Lesley Milroy, cites the following
features as common to AusE and Belfast English: “1. yous /juz/ stressed; /jəz/
unstressed; 2. adverbial but: I went to the store but; 3. the expression Good on
you; 4. epenthetic (ə) in film [filəm]” (Horvath 1985: 39). She adds there:
“And I would very tentatively suggest that the intonation pattern to be discussed in Chapter 9 might well have its roots in Irish English”, by which she
apparently means the High Rising Tone (cf. Section 5.3 below) dealt with in
her Chapter 8, though she only mentions there Belfast English as one of six
BrE varieties reported as containing this intonation pattern (Horvath 1985:
In a paper given in 1992 I explored various aspects of possible Celtic
influence on AusE and was able with reasonable certainty to show that the
AusE phrase of praise or congratulation Good ón you was, because of its
eccentric accentuation, calqued on Irish Gaelic maith ort (maith ‘good’, ort >
or ‘on’ + tu ‘you’) and so derived from that source (Taylor forthcoming: 63),
and that the 2nd person pronoun singular:plural contrast you:yous(e), found in
IrE and in varieties of English spoken where the Irish have emigrated in large
numbers was calqued semantically on the Irish Gaelic contrast tú:sibh, with
yous being modelled on English noun plural formation, though AusE had
developed different rules of use from those in IrE (Taylor 1997: 267–9;
Taylor forthcoming: 61f.).6 Presumably both have been mediated to AusE by
IrE rather than developing here directly from Gaelic.7
In his Regensburg master’s thesis Clemens Fritz analysed a corpus of 359
letters written in the 19th century, of which a large proportion derive from
“Irish Australia”, and concluded that “[t]here was no unified use of grammar
in nineteenth century Australia”. Fritz recommends an investigation of the
Irish Australian letters to establish how IrE and AusE interacted at the time
(Fritz 1996: 88, 89f.; see also Fritz 1998).8
3.2 Early interaction with AmE
There were contacts with speakers of AmE from about 1800, through the visits
of American ships, especially whaling vessels, to Port Jackson (as Sydney
Harbour was then usually called), but such contacts had no demonstrable effect
on AusE. In the mid-19th century many Australians took themselves off to
America after gold was found in California in 1849. Most returned to Australia
when the gold rushes began here in the 1850s, and with them came many
American miners — some 16,000, it is estimated — to try their luck in
Australia. Ramson, who has made the major scholarly study of the AmE
borrowings, differentiates between the pre-1850 borrowings, mainly of words
like township which were connected with land surveying and settlement and
often mediated by the colonial administrators in London, and those after 1850
such as digger ‘prospector’. Some of the borrowings, like bushranger and
digger, underwent further semantic development after their arrival in Australia
(Ramson 1966: 132–51; 1981: 32).9
3.3 Early AusE influence on New Zealand English (NZE)
From the time of earliest white settlement in New Zealand, “in the days of
sealers and whalers [when] New Zealand was simply part of the Australians’
area of operations” (Turner 1970: 84), but especially from the 1840s on,
many Australians went to that pair of islands to settle and farm sheep, dig for
gold or cut timber, as many had done in Australia. The proximity of New
Zealand to Australia and the number of Australians who went there, led to
“New Zealand, […], [being] part of a total Australasian settlement” (Turner:
ibid.),10 produced relatively heavy borrowing from AusE into NZE, and
resulted in NZE being closer to AusE than any other variety (cf. Bennett
1970: 80).
World War I would have provided a period of particularly intensive
interaction between AusE and NZE, since the Australian and New Zealand
Army Corps, the ANZACs, fought together throughout it. I was struck when
reading an early draft of the Dictionary of New Zealand English (“DNZE”) by
the fact that the earliest cited example of the use of the plural pronoun form
youse in NZE was from 1918 and in a military context when a non-commissioned officer addresses his men with the words: “Youse ‘roughies’ ‘ave
really ter get ter know me better, or we won’t be friends.” This appeared to me
to be a grammatical borrowing from AusE due to that interaction, all the more
likely because Australians generally think that it was Scots rather than Irish
who settled New Zealand along with the English. However, I have since
learned on a visit to New Zealand that there were two major waves of Irish
immigration there, and on the appearance of the DNZE in print I found that an
earlier example, spelt uniquely yewse, had been inserted, and it was from a
newspaper item dated 1908 (DNZE, s.v. youse), so the word may have been
mediated to NZE directly from IrE.11
Interactions from World War II to the end of the 1980s
Research into and publication on AusE interaction with other Englishes began
to increase in the 1970s and 1980s, but this was mainly concerned with the
influence of AmE on AusE. While Sussex looked in particular at the role of
AmE in the Australian media, especially television (Sussex 1978, 1985,
1989), I examined a limited range of items, such as hopefully, which had come
into AusE from AmE, but which I regarded as being loans in AmE from
German (Taylor 1978). Just over a decade later (Taylor 1989), I threw my net
much wider and, using especially newspapers and magazines as my sources,
looked at foreign influences on AusE (notably lexical, but also phonological,
morphological and syntactic). I was able to demonstrate that, in addition to the
very strong lexical influence from AmE on AusE, during the 1970s and 1980s
there was a degree of countervailing BrE lexical influence attributable very
much to British television programs screening in Australia at that time. Space
constraints prevent me from reviewing the details here. Cornelius followed up
soon afterwards with a limited attempt at an empirical study of the degree to
which a mixed group of 26 informants, (14 Australian speakers, plus 12
speakers of BrE, AmE and non-native speakers of English), could identify
correctly on linguistic grounds Australian versus American articles in a single
issue of each of two magazines, one Australian (The Bulletin) and one American (Newsweek), published at the same time. While he was able to say in his
summary: “This survey has highlighted a number of significant differences
between AmE and AusE”, a significant conclusion suggested by his data was
that “the dividing line between AmE and AusE (if such a thing exists) is
somewhat blurred” (Cornelius 1989: 89).
Interactions in the 1990s, especially as reflected in newspapers
In Taylor (1989) I relied mainly upon Sydney newspapers and magazines as
primary sources for Americanisms used in AusE, supplementing these with
further observations of my own and of others (e.g. Gunn 1970: 60). In the
remainder of the present chapter, Sydney newspapers will be used rather as
secondary sources: using an array of articles, columns, letters to the editor,
cartoons, and so on, I will consider how readers, columnists, cartoonists, and
even some scholars whose opinions have been reported, have reacted in the
1990s to the interactions of AusE with other Englishes.
5.1 Ongoing AmE influence on AusE
A selection of letters-to-the-editor will serve to indicate the kinds of attitudes
newspaper readers express in regard to AmE influence. Sometimes a complaint about AmE is couched in a more general bemoaning of American
cultural influences, as in the following chain of letters. Daniel Greer in a letter
entitled “US-Centric” in the Sydney Morning Herald (hereafter “SMH”) berates the newspaper for its front-page and further three page coverage of
President Clinton’s election win and then goes on (SMH 9/11/96, p.42):
In the recent debate about immigration and its effects on this country no-one
has raised the alarm against the insidious invasion of US culture. As we sit
down to another evening of sitcoms, eating macaroni and cheese, wearing our
baseball caps backwards, and saying things like “Yo, dude” and “Hi, guys”,
will any of us really care that we have become the 51st state of America?
This explicitly triggered Margaret Grove’s letter under the title “Fed up”
(SMH 12/11/96, p.16), in which she complains:
My kids have traces of American accents, my youngest threw a tantrum when
I wouldn’t allow her to go “trick or treating” on Halloween, and I’m sick of
reading Australian magazines with American spelling.
To this Patrick Sutcliffe responded in turn approvingly in a letter titled “Petrol,
please” (SMH 18/11/96, p.16), in which he complains “[…], I was incensed to
see that great Aussie [comic]strip The Potts talking about ‘gas’ for the car on
the way to Bullamakanka!”12 He goes on to say that no one of his “wide range
of friends and acquaintances in both business and social life […] calls it
‘gas’”, and that he fully supports Margaret Grove “about how we should stop
the American invasion”.
Similarly, Ronald Spain had earlier in the year complained about guy
(SMH 15/5/96), saying: “Surely it is time our brightest young persons desisted
from their use of this north American import in place of older, more colourful
words”, and asking: “Have they never had a dekko at its real meaning?”13 He
then proceeds to recommend the “older, more colourful words”, namely bloke,
coot and cove and scorns guy as meaning “an artificial construction set up for
ridicule”. P.F. Gill of Tasmania in a letter titled “Invidious Americanisms”
begins by complaining that Australia’s biggest retailer, Coles-Myer, “[…] is
to open a chain of stores to be labelled with that illiterate American jingoistic
name World 4 Kids or W4K”, and at the end of the letter Gill expands the
complaint: “I note also with concern that Coles Myer is adding more American expressions into our Australian culture promoting terms such as ‘cookies’
instead of our revered ‘bickies’ or ‘diner’ instead of ‘cafe’. It is a great shame
and a cause for great concern when an Australian organisation continually deAustralianises our heritage.”
In one set of four letters (SMH 8/7/93, p.10) titled “Bêtes noires that
really get one’s goat” we find again this fear of the “de-Australianising” of
AusE by AmE. The letters decry pronunciations such as “ree-search” for
research and “pry-merrily” for primarily, and words such as guy (in fact
already borrowed in previous decades; cf. Taylor 1989: 229, 240) which the
writer points out applies to “blokes, cobbers and now includes girls”.
Many columnists also take a stand. There is in the columnist Phillip
Adams’s article “It’s your shout” (“The Weekend Review” of The Weekend
Australian 17–18/2/96, p.2) a somewhat hysterical reaction to American
influence, which he claims involves some 10,000 words and threatens to
replace indigenous AusE lexical items with AmE ones. He urges his readers to
campaign to save and protect this threatened lexis by adopting and nurturing
one AusE word. He himself will adopt for nurturing drongo ‘a slow-witted or
stupid person’, “the somewhat affectionate pejorative [which] has been replaced by dickhead”. Adams’s article, triggered, incidentally, by the appearance of Johansen (1996), to which he had written the Foreword, is accompanied
by Steven Moore’s illustration containing in a Scrabble-like layout seven of the
AusE words he wants rescued: bludge ‘to evade responsibilities; to impose on
others’, ratbag ‘rascal, rogue; person of eccentric or nonconforming ideas
or behaviour’, sanger ‘sandwich’, dinkum ‘true, honest, genuine’, furphy
‘rumour, false story’, sheila ‘girl or woman’, and bonza ‘excellent, attractive,
pleasing’ (also bonzer).14
A variation on the de-Australianisation theme is the alleged perniciousness of what is often called the “politically correct” euphemism deriving from
AmE, as alluded to in two brief items: Andrew Main’s “Twisted lingo” (SMH
7/1/93, p.16) and Matt Condon’s “You’re not fired” (Sun-Herald — hereafter:
“S-H” — 10/3/96, p.108). Such euphemisms are attributed to in themselves
praiseworthy American efforts to spare the feelings of handicapped people by
avoiding the use of traditional words like blind, deaf, dumb and crippled,
which have often acquired unkind connotations such as ‘stupid’, but these can
easily spill over into what are felt as ridiculous circumlocutions. Thus blind is
replaced first by visually impaired, later by visually challenged, which in turn
can lead on, claims Main, to such (intentionally?) laughable creations as
vertically challenged for ‘short’ and hair disadvantaged for ‘bald’ (cf. “the
former girlfriend of celibacy-challenged rock star Michael Hutchence” in
Rosalind Reines’s “See me in Oz” (S-H 28/3/96, p.29)). On the other hand
Matt Condon deals with euphemisms exploited, especially in business, to
gloss over the reality of what might be called “bloody-minded” actions done
to people, so that “to lose your job”, which used to be expressed by older terms
like to be fired or to be sacked, is now replaced by Condon’s long but not
exhaustive list of nineteen euphemisms ranging alphabetically from bumped
via de-hired and involuntarily separated to vocationally relocated (cf. also
Terry Smyth’s “Axed, whichever way you say it”, S-H 23/6/96, p.43).15
But not everyone is as frightened or as derisory of AmE influence as the
letter writers and the columnists. Some scholars whose work in the area of
AusE is reported think the degree of influence commonly decried is in fact
much exaggerated.
Susan Butler, who works on Australia’s Macquarie Dictionary, points out
in her long article titled “Hasta la vista, strine, the lingo’s gone feral” (SMH 7/
12/96, Spectrum p.4s) that some earlier AmE borrowings, such as bushranger,
jinx, and phoney, have become so well integrated into AusE that its speakers are
no longer aware that they are AmE in origin. But she asserts too that while
clearly AmE words like dude meaning ‘man, person’ have been borrowed into
AusE, such borrowings may not always last and AusE words may be revived
(such as grouse, meaning ‘good’, which Butler claims is “[…] making a
comeback in the playgrounds of Australia”). Her point is, incidentally, amusingly illustrated by a two-frame cartoon from earlier in the year (SMH 1/3/96,
p.24) in which in the first frame a bespectacled, bald and heavily jowled man
is saying: “hello children”, over the caption “Politician slightly out of touch
with today’s young people”, and in the second frame, now sporting sunglasses
(AmE shades) and a turned baseball cap, he is saying: “Yo dudes and dudettes”,
over the caption “Politician completely out of touch with today’s young
Another long article, by Roy Eccleston on pp.4–6 of The Weekend
Australian Review supplement to The Weekend Australian, 10–11/1/98, titled
“Hey, you Septic Tanks: Australian English strikes back” on the Review’s
cover and “War of the Words” on p.4, takes a similar line, as is clear from its
subtitle “If you fear for the future of Australian English amid the onslaught of
American films, television and pop culture, don’t.” This article uses not only
Susan Butler as a source but also Bruce Moore as editor of the Australian
Concise Oxford Dictionary. Their combined opinions are summed up in
statements such as on p.4:
The health of Australian English is good. Bonzer and cobber might be in
trouble, but g’day and fair dinkum are well entrenched.
[…] while we’re adopting American and British slang, we’re also refashioning and resuscitating former favourites. Things might be cool, wicked and bad
— all Americanisms for good — but they might also be grouse, a revived
5.2 Continuing BrE influence on AusE through TV and backpackers
Although the older British television series referred to in Section 4 above
have long since ceased to run in Australia, other, newer series have picked up
where they left off and continue the influence of BrE on AusE. The article
“Invasion of the UK ratings snatchers” by Sue Williams (The Australian 16/6/
97, p.12) discusses how popular British television productions are becoming
in Australia, especially police dramas such as Heartbeat, A Touch of Frost,
Inspector Morse, Pie in the Sky and quite especially The Bill, three episodes of
which are currently screened weekly in Sydney, though she also mentions
series that expose Australian audiences to non-EngE varieties of BrE, such as
Hamish MacBeth (ScotE) and Ballykissangel (IrE).
While Williams makes no reference in her article to linguistic influences,
Frank Devine in his column “That’s language” (“The Weekend Review” in
The Weekend Australian, 16–17/8/97, p.10) refers explicitly to the linguistic
influence of The Bill in his opening sentence:
Although subtitles with regional English accents would sometimes be useful
for us addicts of The Bill, the lingo of Sun Hill police station has eased its way
into our vocabulary. We all know a ‘tom’ is a prostitute. What else would you
call her?
The series title is a BrE phrase meaning ‘the police’, and is often heard in the
episodes in the form the old Bill. That this phrase had “eased” its way into
AusE vocabulary, was demonstrated when former Australian Governor-General, Mr Bill Hayden, handed over office to his successor, Sir William Deane,
in February 1996 and one newspaper twice had the headline: “The old Bill
hands over to the new “ (SMH 17/2/97, pp.1, 6), a pun on the BrE phrase.
But words like this and tom, meaning ‘prostitute’, often only penetrate the
receptive competence of AusE speakers rather than their active performance.
The answer to Devine’s question, “What else would you call her?”, is arguably AmE hooker rather than BrE tom, for Steve Warnock in his article
“Fighting filth” (S-H 14/4/96, p.13), about prostitutes soliciting customers
along a public highway in Sydney, has a need to express the concept ‘prostitute’ 15 times, but to avoid monotony he alternates General English (GenE)
prostitute (used 8 times) with AmE hooker (7 times), not BrE tom.16 That said,
the word hooker is not so integrated into AusE that a very large and prominent
building in Sydney’s Central Business District built by the late property
developer Mr. L.J. Hooker cannot carry its name “Hooker House” on a huge
sign for all to see because of some risk that it might be thought of as the
headquarters of a prostitutes’ association or a mega-brothel. This possibility
does not seem to suggest itself to the populace at large.17
However, a word that can be employed usefully, and especially one that
fills a semantic gap, may well transfer into the active AusE vocabulary. In the
1990s a striking instance of this has been the BrE exclamation cheers, due
perhaps not only to its frequent use in The Bill, but perhaps also to its use by
young British backpackers visiting Australia, and possibly too to its importation by young Australians returning from their visits to Britain.
As elsewhere, cheers has already long existed in AusE in its use as an
informal toast, but during the 1990s it has acquired or developed certain new
uses, a couple of which compete with pre-existing AusE terms, while a couple
of others may be filling gaps where there was previously no generally accepted term at all. Thus we encounter it now in the following uses:
as an expression of farewell competing with goodbye and AusE synonyms such as hooroo and hóoray (NB the last not with the second syllable
stress of the GenE cheering expression hooráy);
as the signing off formula at the end of letters and faxes and quite
especially of e-mail messages in place of the ritualised and insincere sounding
yours sincerely;18 and
as a competitor to thank you used casually.
A further use I have noticed frequently is as a response to an expression of
thanks. BrE and AusE speakers do not normally have a routine way of
responding to an interlocutor’s expression of thanks, as for example German
does with bitte or Italian with prego. While some may at times say: “You’re
welcome”, this tends to be regarded by many to be a rather self-conscious
Americanism and so may not be readily used. But now I often hear younger
Australians in particular responding with cheers. I also have recent experience
of one elderly Australian man using cheers to indicate that it is in order for
those at the dinner table to begin eating, thus as an equivalent to such
Continental expressions as guten Appetit or bon appétit, traditionally lacking
in present day Anglo-Saxon societies. So this BrE word has undergone sudden
and dramatic semantic extension over the last few years and in doing so is
showing signs of filling at least one, if not two semantic and pragmatic gaps in
AusE, possibly as indigenous AusE developments.
It should be added that, especially among younger AusE speakers, AmE
cool is a competitor with cheers in the sense of ‘thanks’ and as a response to
expressed thanks; however, it is also a competitor with yes and OK as a term of
affirmation or acceptance, a sense that cheers does not seem to have acquired.
5.3 Purported AusE influence on BrE through TV programs
Apart from the presumably temporary adoption of a few Australian slang
terms by some of the thousands of American servicemen in Australia during
World War II mentioned in various of S.J. Baker’s sources (Baker 1945:
284f.), there are virtually no claims that AusE has influenced AmE.19 But
there have recently been claims that, for the very first time, AusE is possibly
having a degree of influence on some varieties of BrE, through Australian
television “soaps” about the lives of everyday Australians.
One such allegedly influential Australian series is Neighbours, screened
daily in Britain over a long period. This claim is made in Simon Kent’s article
“Poms aghast as teens adopt Aussie speech. G’day, G’day, what ‘ave we
‘ere?” (S-H 31/3/96, p.113) and attributed to Professor Barbara Bradford of
London’s School of Oriental and African Studies. Young Britons are said to
be using not only items of AusE lexis, but also AusE pronunciation, including
the much researched intonational feature HRT (the “High Rising Tone”, or
”High Rising Terminal”) whereby Australian women in particular have a
rising, so a question, intonation instead of a falling one at the end of statements
(cf. Horvath 1985: 118–132; Guy and Vonwiller 1989). However, in Christopher Henning’s “Our soaps are sizzling in the UK”, subtitled “The Poms are
talking with Aussie accents” (SMH 30/11/96, p.26) Bradford is reported as
being more circumspect and saying that this HRT or “Upspeak” has been a
feature of young people’s speech for some thirty years in various parts of the
English-speaking world and that British youth are “picking up a feature of
young people’s speech, which happens to be a feature of Australian young
people’s speech”, though she grants that it is “quite possible that Australian
soaps have triggered this effect”.
But the very success of some Australian television productions has led to
their being remade for the overseas market with a concomitant linguistic
adaptation to the particular market area. Stephen Dabkowski in his article
“Britons love our Mother and Son” (SMH 31/5/97, p.30) reports that the
Australian series Mother and Son is being remade for Britain with a “Bristol
burr”, a Bristol accent, and a very brief item titled “Eel be right” (The Sunday
Telegraph 3/3/96, p.188) reports that “the new cop drama Blue Eelers, or Water
Rats as it is described in the program details,” has for “an American version
ordered for export” undergone “much re-recording of the dialogue[…]”. This,
of course, prevents any possible AusE influence on the other varieties.
5.4 AusE influence on NZE, and NZE non-influence on AusE
We implied in Section 2.3 above that there had been two phases of influence
of AusE on NZE, though there has probably been in reality ongoing linguistic
influence from Australia onto its little neighbour. The result has been the
striking similarity of the two varieties referred to earlier.
However, in the last few decades, NZE has undergone an autonomous set
of changes in its vowel system, changes so profound that they have sometimes
been referred to as the “Second Great Vowel Shift” (e.g. Bauer 1979) . The
high front vowel // has been retracted to a central position // or /ə/ (cf. ScotE
Wully for the name Willy) and a later drag-chain effect has raised the position
of /ε/ towards // and of /æ/ towards /ε/ (but see now Trudgill et al. 1998). This
has resulted in some degree of miscommunication between Australians and
younger New Zealanders and has led to Australians making fun of New
Zealanders’ pronunciation. In May 1995 The America’s Cup was won from
the New York Yacht Club, for only the second time in 144 years, by a New
Zealand yacht and there was much good-humoured teasing of New Zealand in
the Sydney newspapers. Some of this was linguistic in that it played on the
vowel system of NZE, as in the following two cartoons (SMH 15/5/95, pp.10,
a two-frame one with an American in the first frame saying: “I’m not
bitter” and a New Zealander in the second clutching the Cup and saying:
“We’re heaps bitter”; that is ‘much better’, playing on the raising of /ε/
towards //;
an Australian male asks a fellow countrywoman: “What’s a yuppie?”,
and she answers: “The sound of a New Zealander watching the America’s
Cup win”; that is,”yippee”, which plays on the replacement of // by the
retracted vowel (<u> being one common device for the graphemic realisation
of centralised vowels in the reproduction of non-standard English).
The retracted vowel is in fact the shibboleth by which Australians identify
New Zealanders, and is referred to metalinguistically by saying that New
Zealanders say sex (more consistent with cartoon 2 above would be sux) when
they mean six.
One can imagine then the great shock when in a 1989 newspaper interview the linguist David Blair said that, to judge by the vowels being used
amongst young people in Sydney’s northern suburbs, Sydney English was
beginning to resemble NZE and that this was perhaps due to the large number
of young New Zealanders emigrating to Australia. One can imagine the great
relief when in further interviews in 1996 Blair said that more recent research
suggested that he had been mistaken.20 The relief is expressed jocularly by
Leonie Lamont in her article title “Thenk hivvins we’re not talking like Kiwis
after all” (SMH 2/12/96, p.3) by her imitating the New Zealand accent in print,
while there is a slightly coarse pun on the metalinguistic term vowel movement
for ‘vowel shift’ and the quasi-medical term bowel movement in columnist
Terry Smyth’s title of his article “Sydney’s great vowel movement” (S-H 5/1/
97, p.39).
AusE in interaction with the PC, the CD-ROM and the Internet
While the development of the personal computer and its software led, from the
1980s on, to a vast amount of mainly AmE technical vocabulary entering the
other Englishes (and other languages), this did not have much of an effect on
non-technical language, except in the occasional instance of semantic extension in a verb like default from its old senses of ‘to fail (to carry out certain
obligations)’ to the new computer-inspired one (not yet found in the 1981
Macquarie Dictionary) used of people, namely ‘to revert automatically to a
particular behaviour’.
In the 1990s, the vastly increased access of the general populace to
computers in Australia and the rapid advances that have led to the availability
of audio- and video-interactive computer programs, computer games and CDROMS and to potentially universal access to the Internet have provided AmE
with another powerful medium besides television to influence younger AusE
speakers. This is brought out in Ann Glover’s article titled “Learning the
lingo” (SMH 23/8/97, “icon” Section p.16), where she says: “The TV, software programs and the Internet can make it difficult for Aussie kids to be
Aussie kids.” The problem is brought out more precisely where she says:
Parents and teachers often complain about accents and pronunciation, as well
as US cultural overload. Toddlers are not only reciting their “A to Zees”, but
are also saying “zeebra” and “gas station”. Some programs even confuse the
pronunciation of “ant” and “aunt”.
Pronunciation has become more of an issue in recent years as children are
using multimedia computers with high quality sound. In the past words on the
screen could be read or interpreted in the children’s own way. Now we hear
not robotic voices, but American voices.
The article reports that “there are a few Australian software designers battling
the tide” and producing materials sensitive to the needs of Australian children,
though “[m]any local CD-ROMs can’t compete with the animation, colour
and sound of the imports that have had megabucks thrown at them.” Another
alternative is to use titles “appearing as ‘World English’ […] rather than as
‘American’ editions.”
As to teenagers communicating internationally over the Internet, Frank
Devine in his column in The Australian Magazine supplement to The Weekend
Australian (21–22/6/1997, p.44), itself based on an item from the Australian
National Dictionary Centre’s biennial journal Ozwords (Anon. 1997), observes that American youth and, following them, Australian youth are using
existing English words with meanings totally confusing to older speakers,
such as bad meaning ‘good’, and babe meaning a ‘male’. However, according
to this information, Australian youth have at least managed to get the charac-
teristic AusE greeting G’day out over the Internet and into what Devine calls
“worldwide teenage parlance”.
Assessment of the situation and prospects for further research
AusE has over its relatively brief history been a receiver of linguistic material
from other Englishes rather than a giver of it. The previously mainly BrE
influence on AusE has to a very great extent been replaced by massive and
increasing AmE influence thanks to film, television and the multimedia computer and the Internet, though there has also been limited renewed BrE
influence through various series of television programs.21 The only substantial
AusE influence on another major variety has been on NZE, for any alleged
influence on young people in Britain through Australian television series is at
worst illusory and at best only transitory fashion.
Many members of the Australian public and many journalists think the
influence of AmE is threatening the survival of AusE as a distinct national
variety. On the other hand, linguists like Butler and Moore writing or written
up in newspaper articles are far less pessimistic. Peters has also in a more
scholarly medium expressed her scepticism about the degree to which AusE is
losing ground to AmE. For instance, she notes on the basis of corpus evidence
that AusE tends strongly to favour irregular past tenses of the type burnt rather
than the regular type like burned in AmE (Peters 1993: 23–25; 1994: passim).
She points out, too, that AusE has long incorporated AmE borrowings without
speakers seeming to feel that they are doing their variety violence. She also
finds that the lists of alleged AmE loans she has compiled from Sussex (1985)
and Taylor (1989) differ quite widely (Peters 1998: 43). She feels too that in
the case of many developments, such as the use of the past participle gotten for
got, there are factors within Australia itself that can, at least in part, account
for them (op. cit. 33). And what is taken over has “been restructured into
Australian paradigms and denationalised”. She concludes: “I would venture to
say that Australian English is no more likely to become Americanised than
English English became ‘Frenchified’ as a result of 300 years of occupation”
(op. cit.: 41). This is in stark contrast to Sussex, who, while conceding that
“Australian English will continue to sound phonetically like Australian English” (Sussex 1985: 402), concludes his most recent publication on the subject
with the claim that: “Australians as a nation don’t seem to place a great deal of
emphasis on their variety of English as part of the national culture” (Sussex
1995: 3).
One of the problems bedevilling all our attempts to make accurate statements about the interaction of AusE with other Englishes, though especially
with AmE and BrE, is the nature of our data. I have worked mainly with
newspaper material. Peters has used computer corpora, and, in an interesting
four-way comparison with Canadian English, both corpora and magazines
(Peters and Fee 1989). So far, however, these corpora have been based on
printed — and therefore, for the most part, edited — material. As Peters
admits, corpora of spontaneous speech need to be compiled and used too
(Peters 1993: 27). Sussex has, on the other hand, used radio, film and especially television as his sources (and has recently been awarded a large grant to
pursue his research further). But what we lack are empirical surveys of
speakers themselves. Cornelius (1989) is the result of one very limited survey
that could give only vague results. Clyne (1992) looks at the interaction of
individuals’ Englishes with AusE in Australia, including the effect of AusE on
a speaker of AmE, while Rogers (1981) had examined AusE influence on the
intonation of two young BrE speakers. Somewhat more ambitious is Oishi
(1998), based on a pair of surveys done in 1997, involving 138 and 43 items
respectively, in which he looked not only at BrE and AmE, but also at GenE.
For the first survey he had 39 Queensland informants, and for the second a
mixture of 21 informants — 7 students in Sydney, and 5 (Australian?) students
and 9 (Australian?) teachers at his University in Japan. However, though this
is a move in the right direction, Oishi’s selection of informants was necessarily haphazard and his items were not always appropriate; for example, asking
students in Brisbane about subway versus underground when the city possesses no such railway (Oishi 1998: 13, item 57).
Any sort of adequate survey would at least have to:
1. use informants throughout Australia and ideally not just in the main
cities, but also the country, which may prove to be more resistant to
cultural change and so to foreign linguistic influences,
2. compare the different socioeconomic areas of the main cities, and
3. compare the different age groups.
Clearly any such survey would be a considerable undertaking and one wonders just how long any of the results would remain stable as the increasing
speed of technological development brings the inhabitants of the global village ever closer together and subjects potential informants to ever more
intense exposure to AmE, for those younger members of our society communing day in, day out for hours on end with their AmE-speaking computers are
being interactively exposed to that variety in a way earlier generations never
were with passive TV watching.
The chapter is based on lectures given in late 1997 at the Universities of Bamberg and
Würzburg, Germany, the Åbo Akademi in Turku, Finland, and at the MAVEN 97
conference at Växjö University, Sweden. Taylor (1998) is a shortened version of that
paper with a somewhat different emphasis from the present chapter.
I recall hearing years ago — where, I have now forgotten — the hypothesis that the
reason that <a> in words like dance, example, and grasp is pronounced /æ:/ amongst the
Sydney working class and /a:/ among the better off was because the ancestors of the latter
had picked up the new pronunciation on their visits back “home”, while the rest were too
poor to have such an opportunity and so stayed with the older “colonial” pronunciation
(cf. Bradley 1991; Taylor 1997: 264).
Kniezsa (1997) attempts to evaluate the role of Cockney in the formation of AusE
pronunciation on the basis of undervalued quasi-linguistic observations made by Samuel
McBurney in the 1880s.
In what follows I have not dealt with Troy (1992) for, while her paper is of some
relevance, it is principally concerned with Australian Pidgin and so falls outside the scope
of the present study.
While answering questions after his lecture on 12 October 1993, Mitchell said he thought
that the quality of AusE /a/ might have been influenced by IrE, but this is not mentioned
in Mitchell (1995).
Interestingly, yous is found in Tasmania as well as New South Wales, although few or no
Irish convicts were sent to that island, which suggests that yous may have already been
part of Sydney convict English around the time Tasmania was colonised from 1803 on.
However, I have not been able to find any Adelaide speaker who knew it from that city.
While the east coast centres all had convicts sent to them until 1840, and Western
Australia from 1850 to 1868, South Australia did not.
In July 1998 Anne-Marie Whitaker gave a paper (Whitaker 1998) in which she complained that Australian lexicographers have tended, in spite of the large proportion of
Irish speakers in Australia in the early days, to give precedence to speculative etymologies deriving from EngE dialects rather than to a possible origin from Irish. She then
proceeded to suggest such Irish origins for a number of words (e.g. chook ‘fowl’, ‘hen’
derived from a Gaelic word of almost identical pronunciation used to call out to hens).
While she admitted that she was a historian and not a linguist, and some of her etymologies did seem doubtful, lexicographers would do well to heed Whitaker’s main argument,
namely that Irish origins should always be considered in assessing possible etymologies.
In dealing briefly with the suffix -o in AusE (such as in garbo < garbageman, Robbo <
Robertson), Fritz notes that the suffix also occurs in Dublin speech (Fritz 1996: 42 note
161). I had independently come to the conclusion that borrowing of this suffix from IrE
was a preferable explanation to the “calling-out” origin of it suggested in Taylor
(1992: 553; cf. Taylor 1998: 60).
The journalist Sidney J. Baker, who first looked at the influence of American slang on
AusE (Baker 1943), subsequently attempted a brief but broader survey of “Oversea
influences”, including “American influence” (Baker 1945: 277, 280–8), claiming on
p.277 that certain words had been recorded earlier in Australia. However a word of
warning is sounded by Ramson (1966: 133), who advises “Nor should we too readily
draw conclusions (as Baker does) about the histories of words which are at present
recorded earlier in American English than they are in Australian”, and cautions against
“words which have been borrowed separately into American and Australian English from
British slang or regional dialect use.” (ibid.)
From an allegedly 1961 source Orsman cites the following in his Introduction to the
DNZE: “…There has been significant trans-Tasman exchange of migrants, especially in
the 1860s, the period c.1885–1893, with two periods of loss to Australia (1885–91 and
1967–8) [and with further loss from NZ to Australia from the 1980s on, B.T.]” (DNZE
1997: vii note 3) and later, with reference to the Australian National Dictionary, mentions lexical material shared with AusE (p.ix).
Later examples of the use of youse in the DNZE mark it as particularly characteristic,
almost a shibboleth, of Maori English today, though it is not in Australia regarded as
mainly a feature of Aboriginal English, rather of “uneducated” speech generally.
Bullamakanka, a name made up from Aboriginal Pidgin bullamakau ‘head of cattle’, and
the suffix -nka, sometimes found on place names in Aboriginal languages, refers to a
proverbial, though imaginary, remote Australian town (cf. also the synonymous Woop
Woop, modelled on reduplicated place names of Aboriginal origin like Wagga Wagga).
It is striking how often in letters and articles defending AusE from AmE encroachment,
expressions that are thought to be Australianisms are employed (here dekko ‘look’).
Some of the words are not, however, Australianisms at all. They include lingo ‘language’
(also AmE) and rhyming slang phrases such as Septic Tanks (‘Yanks’), which is BrE
Cockney in origin, and porkies (used in connection with the 1998 Federal election; cp.
porky pies < pork pies ‘(politicians’) lies’), deriving from a significant feature of British
cuisine but one that is virtually unheard of in Australian cuisine.)
The definitions are taken from the The Macquarie Dictionary 1981: s.v.
These kinds of euphemistic neologism seem to have had their origin particularly in the
Vietnam War and to have continued into the Gulf War; one of the most infamous is
collateral damage for ‘damage incidental to the main targets, but usually involving
civilian deaths’ (see Clark 1998: s.v. Collateral Civilian Damage).
Although tom does not seem to have caught on in AusE in this sense, I have in late 1998
noticed the word punter, which in The Bill is used in the senses ‘prostitute’s customer’
(AmE john), then ‘member of the public’ (= potential police customer?), twice being used
in Sydney newspapers in the more general sense, one of these with reference to the voters
in the upcoming 1998 Federal election.
Relevant here is a brief letter under the title “Local lingo” (SMH 9/11/96, p.42) recalling
“the heading on the (back) sports page of one of the afternoon [Sydney] tabloids in the
late ‘60s, early ‘70s, which stunned American visitors: ‘Manly Hooker Flattens Raper’”,
where hooker actually refers to the ‘player whose job is to hook the ball out of the scrum’
and Raper to the Rugby League footballer, Johnny Raper.
A listener to a talkback program on language on which I was being interviewed on the
Sydney radio station 2BL on 3/9/97 said that during World War II he and his Australian
wireless operator colleagues in New Guinea used cheers as a term for signing off, so it
appears to have been present already in a restricted variety of AusE, probably borrowed
then too, however, from BrE speakers.
Frank Devine in his “That’s language” column makes a possible case for one of Baker’s
wartime AmE borrowings into AusE, jane meaning ‘woman’ (Baker 1945: 123, 288),
actually being of early 20th century Australian origin (The Australian Magazine 21–22/
11/95, p.29). See also Maurer (1944) on the influence of AusE rhyming slang on AmE.
Blair was, however, not the first to comment upon the retraction of the vowel [I] in
Sydney English, for it is already mentioned in Bradley (1980: 76).
In the case of both the Australian “soaps” in Britain and the British (London-based)
programs in Australia a pre-condition for interdialectal influence would appear to be the
frequency of exposure to the speakers of the the other variety over a long period which
leads to a sustained emotional engagement with them, though without any active linguistic interaction, of course.
AND = Australian National Dictionary, The. A Dictionary of Australianisms on Historical
Principles. 1988. Ed. by William S. Ramson. Melbourne: Oxford University Press.
Anon. 1997. “Do they still roll the Jaffas down the aisle?” Ozwords 3(1): 1–2.
Baker, Sidney J. 1943. “The influence of American slang on Australia”. American Speech
18: 253–6.
———. 1945. The Australian Language: an Examination of the English Language and
English Speech as Used in Australia from Convict Days to the Present, with Special
Reference to the Growth of Indigenous Idiom and its Use by Australian Writers.
Sydney, London: Angus & Robertson.
Bauer, Laurie. 1979. “The second Great Vowel Shift?”. Journal of the International
Phonetic Association 12(1): 48–9.
Bennett, J.A.W. 1970. “English as it is spoken in New Zealand”. In Ramson, ed. 1970, 69–
Bernard, John R.L. 1969. “On the uniformity of Australian English”. Orbis 18: 62–73.
Blair, David. 1975. “On the origins of Australian pronunciation”. Working Papers
(Macquarie University Speech and Language Research Centre) July 1975: 17–27.
Bradley, David. 1980. “Regional differences in Australian English phonology”. University
of Melbourne Working Papers in Linguistics 6: 73–9.
———. 1991. “/æ/ and /a:/ in Australian English”. In Jenny Cheshire, ed. English around
the World: Sociolinguistic Perspectives. Cambridge: Cambridge University Press, 227–
Clark, Gregory R. 1998. Words of the Vietnam War: the Slang, Jargon, Abbreviations,
Acronyms, Nomenclature, Nicknames, Pseudonyms, Slogans, Specs, Euphemisms,
Double-talk, Chants, and Names and Places of the Era of United States Involvement in
Vietnam. Jefferson, N.C.: McFarland.
Clyne, Michael. 1992. “Australian English in contact with other Englishes in Australia”. In
Rüdiger Ahrens and Heinz Antor, eds. Text — Culture — Reception. Cross-cultural
Aspects of English Studies. (= Forum Anglistik, n.F. Bd 8) Heidelberg: Winter, 305–15.
Cochrane, G.R. 1989. “Origins and development of the Australian accent”. In Collins and
Blair, eds. 1989, 176–86 (notes 318; references 341f.).
Collins, Henry E. 1975. “The sources of Australian pronunciation”. Working Papers
(Macquarie University Speech and Language Research Centre) January 1975: 115–28.
Collins, Peter and David Blair, eds. 1989. Australian English: the Language of a New
Society. St Lucia: The University of Queensland Press.
Corbyn, Charles Adam. 1854/1970. Sydney Revels of Bacchus, Cupid and Momus; Being
Choice and Humorous Scenes at the Sydney Police Office, and other Public Places,
During the Last Three Years. Sydney: Printed by Hawksley & Williamson (“Presented”
by Cyril Pearl, with additional, post-1854 items, Sydney, London: Ure Smith, 1970.)
Cornelius, Steve. 1989. “A comparison of magazine articles from the ‘Bulletin’ and
‘Newsweek’ magazines based on readers’ perceptions of differences between American
and Australian English”. Working Papers (Macquarie University Speech Hearing and
Language Research Centre) 1989: 71–109.
DNZE = The Dictionary of New Zealand English: a Dictionary of New Zealandisms on
Historical Principles. 1997. Ed. by H.W. Orsman. Auckland: Oxford University Press.
Fritz, Clemens. 1996. “Early Australian letters: a linguistic analysis”. Unpublished MA
thesis, Universität Regensburg.
———. 1998. “Letters from early Australia”. Dialectologia et Geolinguistica 6: 25–42.
Gunn, John. 1970. “Twentieth century Australian idiom”. In Ramson ed. 1970, 49–68.
———. 1972. “Early Australian pronunciation”. Linguistic Communications (Department
of Japanese, University of Melbourne) 6: 36–40.
———. 1975. “A possible source of some Australian sounds and their relationship with
Received English”. Talanya (Journal of the Linguistic Society of Australia) 2: 5–15.
Guy, Gregory and Julia Vonwiller. 1989. “The high rising tone in Australian English”. In
Collins and Blair, eds. 1989, 21–34.
Hammarström, Göran. 1980. Australian English: its Origin and Status (= Forum
Phoneticum 19). Hamburg: Buske.
———. 1985. “On the origin of Australian English”. In John E. Clark, ed. The Cultivated
Australian: Festschrift in Honour of Arthur Delbridge. Hamburg: Buske, 369–72.
Horvath, Barbara. 1985. Variation in Australian English: the Sociolects of Sydney. Cambridge: Cambridge University Press.
Johansen, Lenie ‘Midge’. 1996. The Penguin Book of Australian Slang. Ringwood: Viking
Kniesza, Veronika. 1997. “A nineteenth-century report on the pronunciation of Australian
English”. Dialectologia et Geolinguistica 5: 65–77.
Macquarie Dictionary, The. 1981. St Leonards: Macquarie Library.
Maurer, D.W. 1944. “ ‘Australian’ rhyming argot in the American underworld”. American
Speech 19(3): 183 -195.
Mitchell, Alexander G. 1946. The Pronunciation of English in Australia. Sydney: Angus &
———. 1995. The Story of Australian English. Macquarie University, NSW: Dictionary
Research Centre.
——— and Arthur Delbridge. 1965. The Pronunciation of English in Australia. Revised
ed. Sydney: Angus & Robertson.
Oishi, Itusuo. 1998. “American and British influences on Australian English”. Seikei
Hôgaku 47 (March 1998): 1–36.
Peters, Pam. 1993. “American and British English in Australian usage”. In P.H. Peters, ed.
Style on the Move: Proceedings of Style Council 92. Macquarie University: Dictionary
Research Centre, 20–7.
———. 1994. “American and British influence in Australian verb morphology”. In Udo
Fries, Gunnel Tottie and Peter Schneider, eds. Creating and Using English Language
Corpora: Papers from the Fourteenth International Conference on English Language
Research on Computerized Corpora, Zürich 1993 (Language and Computers, 13).
Amsterdam, Atlanta: Rodopi, 149–58.
———. 1998. “Australian English”. In Phillip Bell and Roger Bell, eds. Americanisation
and Australia. Kensington: University of New South Wales Press, 32–44.
——— and Margery Fee. 1989. “New configurations: the balance of British and American
English features in Australian and Canadian English”. Australian Journal of Linguistics
9(1): 135–47.
Ramson, William S. 1966. Australian English: an Historical Study of the Vocabulary,
1788–1898. Canberra: Australian National University Press.
———. ed. 1970. English Transported: Essays on Australasian English. Canberra: Australian National University Press.
———. 1981. “The vocabulary of Australian English”. In The Macquarie Dictionary.
McMahons Point, NSW: The Macquarie Library, 28–33.
Rogers, Inge. 1981. “The influence of Australian English intonation on the speech of two
British children”. Working Papers (Macquarie University Speech and Language Research Centre) 3(1): 25–42.
Sussex, Roland. 1978. “North American English as a prestige model in the Australian
media”. Talanya (Journal of the Linguistic Society of Australia) 5: 36–41.
———. 1985. “Linguistic evidence on the Americanization of Australian English: preliminary report.” In John E.Clark, ed. The Cultivated Australian: Festschrift in Honour of
Arthur Delbridge. Hamburg: Buske, 395–402.
———. 1989. “The Americanisation of Australian English: prestige models in the media”.
In Collins and Blair, eds. 1989, 158–168 (notes p.317).
———. 1995. “Americanisms roll into Australian English”. Australian Style 3(2): 2–3.
Taylor, Brian. 1978. “German as a source of American prestige forms entering Australian
English”. Talanya (Journal of the Linguistic Society of Australia) 5: 1–14.
———. 1989. “American, British and other foreign influences on Australian English since
World War II”. In Collins and Blair, eds. 1989, 225–54 (notes 318–21; references 345–
———. 1992. “Otto 988 to Ocker 1988: the morphological treatment of personal names in
Old High German and colloquial Australian English”. In Claudia Blank, ed. Language
and Civilisation: a Concerted Profusion of Essays and Studies in Honour of Otto
Hietsch. Frankfurt-on-Main: Lang, vol.2, 505–36.
———. 1997. “The inner-city working class English of Sydney, Australia, around 1900: a
linguistic critique of Louis Stone’s novel Jonah”. In Heinrich Ramisch and Kenneth
Wynne, eds. Language in Time and Space. Studies in Honour of Wolfgang Viereck on
the Occasion of his 60th Birthday. Stuttgart: Steiner (Zeitschrift für Dialektologie und
Linguistik–Beihefte Nr. 97), 258–70.
———. 1998. “Australian English and its interaction with other Englishes”. In Hans
Lindquist, Staffan Klintborg, Magnus Levin and Maria Estling, eds. The Major Varieties of English. Papers from the MAVEN 97. Växjö: Växjö University (Acta
Wexionensia, Humaniora 1), 59–68.
———. Forthcoming. “Syntactic, lexical and other transfers from Celtic in (Australian)
English”. In Geraint Evans, Bernard Martin and Jonathan Wooding, eds. Origins and
Survivals: Proceedings of the First Australian Conference of Celtic Studies (held at the
University of Sydney, 5–9 July 1992). Sydney: The University of Sydney, Centre for
Celtic Studies, 47–70.
Troy, Jakelin. 1992. “‘Der Mary this fine cuntry is there is in the wourld’: Irish English and
Irish in late eighteenth and nineteenth century Australia”. In Tom Dutton, Malcolm
Ross and Darrell Tryon, eds. The Language Game. Papers in Memory of Donald C.
Laycock (= Pacific Linguistics, Series C — 110). Canberra: Department of Linguistics,
Research School of Pacific Studies, The Australian National University, 459–77.
Trudgill, Peter. 1986. Dialects in Contact. Oxford: Basil Blackwell.
———, Elizabeth Gordon and Gillian Lewis. 1998. “New-dialect formation and Southern
Hemisphere English: the New Zealand short front vowels”. Journal of Sociolinguistics
2(1): 35–51.
Turner, George W. 1960. “On the origin of Australian vowel sounds”. AUMLA (Journal of
the Australasian Universities Language and Literature Association) 13: 33–45.
———. 1970. “New Zealand English today”. In Ramson, ed. 1970, 84–101.
Whitaker, Anne-Marie.1998. “Lexicography as cultural genocide: the Irish influence on the
Australian language”. Unpublished paper given on 25th July at Literature and Politics in
the Celtic World: Third International Celtic Conference, University of Sydney, 23rd26th July 1998.
A Geolinguistics of short A
in Australian English
Barbara M. Horvath and Ronald J. Horvath
University of Sydney
Short A in English
The variable linguistic phenomenon we discuss here goes by a number of
names but is most often alluded to by example. We are concerned with the
variant pronunciation of a word like dance as either [dæns] or [dans] which
Americans often refer to as short or flat A vs broad or long A (Ferguson 1972).
Wells (1982: 133–135) describes these variable pronunciations first of all by
noting that there is a group of words in English — which he calls the “BATH
lexical set” — which in various dialects may be pronounced either with the
vowel used in the TRAP lexical set (short A [æ]) or the vowel used in the
PALM lexical set (broad A [a]). This leads him to speak of “short-BATH”
dialects and “broad-BATH” dialects. What is unusual about the phonological
variation of the BATH lexical set is that although one can describe the
phonological conditioning to a certain extent in those places where there is
variation, one cannot predict that the conditioning will apply in the case of any
particular word. For instance, in British Received Pronunciation (“RP”) broad
A is used in demand, command and slander, so that under normal circumstances one could expect broad A in other words followed by [nd], such as
grand, stand or gander — but in fact RP has short A in these words.
Furthermore, the variation between dialects, both within and between nations,
is also complex. Some dialects can be described as basically short A dialects
and have few occurrences, or none at all, of the broad A variant: General
American is one such dialect. RP is a broad A dialect and, with some
exceptions, broad-BATH accents are found in the south of England, the West
Indies, and the southern hemisphere (Wells 1982: 135). However, even
within broad A dialects, the distribution of the short A and broad A within the
lexical set varies from one place to another. All of this complexity means that
it is often very difficult to predict with any confidence the pronunciation of a
particular word in a particular place.
Wells (1982: 135) handles all of this complexity by first of all defining
the BATH lexical set as all the words in which General American has short A
and RP has broad A. He then has three other groups of words: the first contains
words that are pronounced with the TRAP vowel by speakers of otherwise
broad A dialects, with Australia being one of these; the second group contains
words that are pronounced with the TRAP vowel in a normal broad A dialect,
such as in the north of England; and finally, the last group of words are those
that all have short A in General American but which fluctuate between the
TRAP and PALM vowels in RP.
Let us look at what Wells proposes for Australian English (henceforth
“AusE”). He gives the following list of words and says that they are “sometimes” said with the TRAP vowel:
dance, advance, chance, France, lance, glance, enhance, prance, trance, entrance v.;
grant, slant, aunt, chant, plant, advantage, vantage, chantry, supplant enchant;
branch, blanch, ranch, stanch, stanchion;
demand, command, remand, slander, chandler, commando, Alexander, Sandra, Flanders;
example, sample;
chancel, chancellor, Frances, Francis, lancet, answer.
His suggestion here is that Australian English has some phonetic conditioning: preceding a nasal + consonant (ns, nt, ntʃ, nd, mpl), “many” Australians
use the TRAP vowel, while “other” Australians have broad A “in all the
BATH words, as do New Zealanders and South Africans.” (Wells 1982: 233).
Researchers have long recognised the existence of variation between
short and broad A in AusE. This is what Mitchell (1958: 66) had to say about
“If heads were counted in Australia there is little doubt that the majority
would be found to use the æ pronunciation. The position seems to be that a is
the older pronunciation and æ the more recent. a seems to be commoner
among older people, and those who are self-consciously conservative in their
habits of speech. æ is commoner among younger people.”
Our study will empirically examine some of these claims. However what is
clear from this short overview of the BATH lexical set is that in order to
predict the pronunciation of this variable sound, you have to know where you
are and what the word is.
Many articles and books make reference to national varieties of a language,
such as Canadian English, New Zealand English and Australian English
(McArthur 1998). Part of the justification for recognising these as linguistic
entities is either that they represent more or less widely accepted standard
varieties, or that there is so little variation within each country that there is
only one dialect. In Australia it has long been held by researchers that there is
“a greater uniformity of pronunciation extending over a wider expanse than
anywhere else in the world.” (Bernard 1981: 19). Mitchell and Delbridge
(1965a: 11) recognised variation within AusE, but they noted that the same
varieties occurred throughout the whole country. Since this pioneering study,
there have been descriptions which have focused on sociolinguistic variation
(e.g. Horvath 1985) and geographical variation (e.g. Bryant 1985, 1989).
Bryant’s research has identified a number of dialect regions based on lexical
differentiation, while others have studied the regional variation of phonological variables: the vocalisation of /l/ (Borowsky and Horvath 1997; Horvath
and Horvath 1997), and vowel variation, including that with short and broad
A, from both a regional and sociolinguistic perspective (Bradley 1989).
In this chapter we shall concentrate on the geographical aspects of the
variability of short A in Australia. We call the approach “geolinguistics” after
the suggestion made by Chambers and Trudgill (1980). In the early work in
dialect geography by researchers such as Orton (1978) in England and Kurath
(1949) in New England maps were used to display relatively unanalysed data
and it was hoped that these would reveal regional dialects (via the emergence
of an isogloss or a bundle of isoglosses that would suggest where some dialect
boundary could be located). Explanations for the patterns would then be
sought in terms of something like settlement history. This approach conformed with the practices of regional geography of the same era — the 1930s,
1940s and 1950s. With the advent of sociolinguistics in the 1960s came a
variety of criticisms of dialect geography, including sampling and data collection methods as well as methods of linguistic analysis and interpretation.
Geolinguistics draws upon a more contemporary meaning of geography and
its proponents espouse the view that geography has an important role to play
in accounting for, and indeed explaining, language variation. In the following
section we show that in order to predict the occurrence of short A in Australia,
it is necessary to identify where a speaker lives. We show that Australia
consists of a set of speech localities that differ from one another, sometimes
quantitatively and sometimes qualitatively. We will use our conceptualisation
of geolinguistics to address the question of how we should define AusE.
Geolinguistics is an approach to linguistic variation that takes into account
three factors: linguistic, social, and geographic. By “linguistic” we mean the
analysis of language variation along the lines sociolinguists have been developing since Labov’s earliest work (Labov 1966). By “social”, we mean taking
into account how social structures, especially gender, social class and age,
constrain variation and change. By “geographical”, we mean investigating
how the concepts of place, space and scale contribute to an understanding of
linguistic variation and change. Geolinguistics thus adds a contemporary
conception of geography to the field of sociolinguistics.
Table 1 presents a comparison of geolinguistics and dialect geography.
Current work linking sociolinguistics and geography is being developed in
different ways by several others (Trudgill 1974; Kretzschmar 1994; Bailey et
al. 1993; and Britain 1991).
Short A in Australia
Two empirical studies of short A are reported on here. The first one, which
deals only with Sydney, is based on a large number of words, mostly drawn
from Wells (1982: 135), and examines Wells’ description of AusE. The
second is based on a much smaller list of words, but incorporates a survey
taken in five major cities.
3.1. The pilot study: Sydney
A study was conducted1 in Sydney which was designed to test Wells’ predictions for AusE and to see whether or not Sydney fell into Wells’ “many
Australians” or “other Australians” category referred to above. Thirty two
speakers were asked to read a word list and a passage with the words taken
primarily from Wells’ list; a smaller number of other words were investigated
Table 1. Dialect Geography and Geolinguistics
Dialect Geography
1. Sample
Use of a very small number of
speakers, often a single
speaker, selected according to
predetermined criteria (older,
rural, etc.)
2. Type Of
3. Geography
Qualitative analysis
4. Mapping
Direct mapping of language
Use of isoglosses for
identifying dialects
Use of a sample of speakers
(sufficient in number to allow tests
of statistical significance to be
used) from a speech locality representing its gender, social class and
age structure
Quantitative and qualitative
Recognition of the possibility that
place, space, and scale are possible
constraints on linguistic variation
Mapping of sociolinguistically
analysed features
Characterisation of the geographical dimension of language change
in progress
Recognition that both individuals
and groups can be analysed and
5. Research
1930s-1950s Regional
Geography framework
6. Individual vs
Vulnerability to the “individual
Speech Locality fallacy” (the assumption that an
individual lect captures the
variation in a speech locality)
with a sample of only six speakers. Tables 2 and 3 show the results of that
investigation. Table 2 lists all the words which were consistently pronounced
with a broad A and another set that were consistently pronounced with a short
A. Table 2 indicates that in Sydney the following words do not vary between
short and broad A as Wells suggests they might: dance, advance, chance,
plant, advantage, branch, Alexander, example. They are all pronounced with
a short A.2 Table 3 contains the set of words that were variably pronounced,
and provides clear indication that the environment of a following nasal +
consonant is a potential one for variability, as is the environment s + consonant.
Tables 2 and 3 also indicate the pronunciations given in the Macquarie
Dictionary for all of the words. The Macquarie includes pronunciations with
both short A and broad A when it is believed that a word has a variable
pronunciation in educated AusE, with the variant judged to be the more
common listed first.3 Bearing in mind that the Macquarie is concerned with
“Australian English” and not just “Sydney English”, let us compare how good
its predictions are for Sydney. Referring now to the list of broad A words in
Table 2, only castle does not vary as the dictionary predicts for AusE; it
would be fair to conclude then that this word is one that is likely to vary
regionally. The list of non-varying short A words, however, contains quite a
few words the dictionary suggests do vary for AusE: chance, advances,
dancing, plants, and advantage.
Table 2. Broad A words and short A words in Sydney
(100 % [a])
(100% [æ])
laugh, laughter, staff, giraffe, calf, halfway (a)
raft, after, craftsman (a)
can’t (a)
class, brass (a); castle (a>æ)
basket, ask, mask (a)
nasty, fast, last, past, disastrous, bastard (a)
path (a)
banana (a)
Alexander, sandal (æ)
chance, advances, dancing (all æ >a)
romance (æ)
plants, advantage (both æ >a)
branches (æ)
example (æ)
masculine (æ)
Table 3 gives a list of words that are variably pronounced either with short or
broad A. According to the Macquarie the relevant vowel in some words
(sample, slander, trans-, and ranch) is unvaryingly [æ] and in others (shan’t,
casket, pastor, pastoral, and contrast) it is unvaryingly [a]. Nevertheless, an
examination of the percentages involved will show that the Macquarie got it
right, most of the time. However, for gasp, hasp and grasp, all of which the
Macquarie suggests are variable, we find that in Sydney at least, gasp is more
frequently pronounced with a broad A, hasp with the short A (as the dictionary
shows), but grasp is more likely in Sydney to have an [a] than an [æ]. The
Sydney speech locality conforms to Wells’ “many” Australians in that they
conform to the specific examples he gives (on p. 233), pronouncing staff, path
and last with broad A and dance, grant, and example with short A. However
Wells also predicts that these same Australians will have broad A in words
that have the vowel followed by a nasal (see the last five lines of the list of
words given by Wells in Section 1 above). Table 3 shows that many of these
are variable in Sydney.
Table 3. Words with variable short A/broad A pronunciations in Sydney
Variable Short
A/Broad A
Macquarie No. of Short
Dictionary A/Total No. of
Short A
a> æ
æ >a
æ >a
æ >a
æ >a
æ >a
æ >a
æ >a
æ >a
æ >a
æ >a
3.2. The national survey
What the pilot study of Sydney tells us is that if we want to find out about the
pronunciation of short A in Australia, we can use the Macquarie Dictionary as
a reasonable guide to the variable pronunciations of the words in AusE, but we
cannot know whether the variability is the same wherever English is spoken in
Australia or whether the variability is an indication of regional differences in
pronunciation. Since the Macquarie Dictionary only includes the standard or
educated variant of AusE, it may be that some of the variation in pronuncia-
tion is the result of nonstandard or uneducated pronunciations. In order to
begin to explore some of these questions, we undertook to survey a small
number of BATH words in the context of a larger study of regional variation
in AusE.
3.2.1. Methodology
Six words were chosen for the study: dance, advance, plant, giraffe, grasp and
mask. They were included intermittently in a word list which was primarily
designed to study the vocalisation of /l/ (Borowsky and Horvath 1997; Horvath
and Horvath 1997). The sample was constructed so that it included five major
cities; eleven speakers from Adelaide were included serendipitously, but the
survey was not undertaken there. An attempt was made to sample a range of
people in each city so that we could study the effects of age, social class and
gender on the use of short A. Table 4 shows how the 185 speakers were
distributed in the six cities and according to age, gender, and social class. The
survey was ordinarily conducted in public places and speakers, who supplied
minimal demographic information, were asked to read a wordlist and a series
of passages into a tape recorder. The data were coded using Language Coder4
and were subjected to statistical analysis. A Varbrul analysis was performed on
the following five independent variables: 1) word (dance, etc.); 2) place
(Brisbane, Sydney, Melbourne, Hobart and Mount Gambier); 3) social class
(Middle and Working); 4) age (in four groups as shown on Table 3); and 5)
gender (men and women). Social class was determined solely by the speaker’s
occupation or by the occupation of the main breadwinner in the family. We
were looking to see whether a set of words were pronounced with short or broad
A. The Varbrul analysis identifies the statistically significant independent
variables that contribute to the overall likelihood of, in this case, a short A being
produced and assigns a weight to each variable which indicates its contribution
to the overall probability. Our analysis identified two factors that were very
strongly associated with the short A: word and place. Of the three social factors,
gender and age were eliminated because they were not statistically significant
and social class was found to be weakly associated with the short A.5
Table 4. The Sample
No. of Speakers
Mount Gambier
Middle Class
Working Class
Social Class
3.2.2. Results
The results of the statistical analysis show that gender and age play no
significant role in the variability of the use of short A in these six words in
AusE. This indicates that we are not studying a linguistic change in progress
because these two social dimensions are always implicated in such a change.
Both the word itself and the place, however, are very important factors in
understanding the variability of the use of short and broad A, while social
class plays a statistically significant, if weak, role. The analysis confirms that
in order to predict how a word in the BATH lexical set will be pronounced in
Australia you need to know what word is being said (see Table 5 below),
where the word is being said (see Table 6 below), and something about who
is saying it (see Table 7 below).
Table 5 shows the variation in the pronunciation of the six study words
for the sample as a whole. All words are variable but some are clearly more
variable than others. Those with a following nasal are more likely to be
pronounced with a short A, but even within this subset, the particular word is
still important. Dance is much less likely than either advance or plant to be
heard with a broad A. Those with a following fricative are more likely to be
pronounced with a broad A; however, while there are a reasonable number of
Australians who would use short A in grasp, only a tiny number would use
short A giraffe, and hardly any at all would use it in mask.
Table 5. The six words in Australian English
Number of Short A
% of Short A
Once we add in information about locality, the picture becomes much clearer.
Table 6 is an implicational table in which the places and the words are
arranged so that the incidence of short A decreases across the rows and down
the columns. The speech localities are strongly polarised: Brisbane is more
firmly a short A dialect than any of the others and Mount Gambier and
Adelaide are broad A dialects, with the word dance being exceptional. As we
know by now, these kinds of lexical exceptions are typical of the BATH word
set. The phonetic conditioning is maintained when we separate the data out by
speech locality. As we can see with Mount Gambier, a following nasal +
consonant does not play an important role in the use of short A. We also see
that in most places both giraffe and mask are overwhelmingly pronounced
with the broad A. In the study one speaker in Sydney, but five in Hobart, used
short A for giraffe, and for mask the only [mæsk] pronunciation was produced
by a young girl in Brisbane. Grasp shows an interesting result; only in Sydney
and Hobart is the short A used much.
Table 6. Percentage of short A in the six words × place
Mt Gambier
As we noted earlier, social class predicts only weakly — but significantly —
that a short A will be used. Table 7 is an implicational table in which the rows
and columns are arranged from most to least likely to have a short A. It
indicates that working class speakers are more likely to use short A than broad
A in these words in AusE. However, it is interesting to note that grasp with the
short A is preferred by middle class speakers
Table 7. Percentage of Short A × Social Class
Let us now compare our findings with those reported by Bradley (1989: 263–
264).6 Bradley found regional, social, stylistic and age differences; we did not
find age differences that were statistically significant and our data collection
methodology did not allow us to investigate stylistic variation (that is, variation induced by the formality of the speech situation). In addition, Bradley
found that in the words that vary (mainly in nasal + consonant contexts), short
A was more frequent in Sydney than in Melbourne (best illustrated in our data
with grasp) and that Adelaide had the least amount of short A (to which we
can add Mount Gambier). Although he would classify Brisbane and Hobart
together, tending to the use of the short A, our results, on this small set of
words, show Hobart patterning more like Sydney (see the data for grasp and
The national and the local scale
The type of linguistic variation exemplified by short A is not like that found in
much of the work in sociolinguistics. First of all, it is not associated with the
usual social characteristics of gender and age and is only weakly associated
with social class. Sociolinguists have generally focused on language changes
in progress in which one expects to find generational variation as well as
differences in the use of a linguistic feature by men and women.7 Some
sociolinguistic variables are not changes in progress but are used to mark
social differences within a speech community. Our data does not show this,
except in the slight preference for short A by working class speakers. How-
ever, others have found a preference for broad A to be a marker of formality or
socioeconomic class (Bradley 1989) or “a sign of pedantry, snobbishness, or
of undue striving for effect” (Mitchell and Delbridge 1965b). We therefore
cannot rule out the possibility of short A becoming sociolinguistically important in AusE. Wells (1982: 79) notes that it is “of indexical importance in
distinguishing high vs popular accents in the north of England.” What we have
with short A in Australia is a geolectal feature rather than a sociolectal one.8
Again, Wells (1982: 79) notes the importance of short A “in distinguishing
northern from southern accents in England as a whole, and in distinguishing
between the two standard accents RP and GenAm” (which we would consider
to be geolects on two different scales, the national and the international).
Two geographical effects have been analysed: the effects of place and
scale. Place effects have been identified using the Varbrul analysis and have
been found to contribute strongly — along with the word itself — to the
likelihood of short A being used. The effects of scale were examined by
comparing the aggregated data for all of the places in Australia (Table 5) with
the pattern of variation in each of the six speech localities (Table 6). Although
we cannot report here on whether the differences between the national and
local scales are statistically significant, or whether the method used to identify
AusE was the best approach to identify the national scale (scaling up through
aggregation), we did discover that national variation was not matched exactly
in any of the six speech localities and was particularly poor in describing the
English of Adelaide and Mount Gambier.
Most work on AusE, which we take to mean the variety spoken at or
representing the national level, has been done with the aim of establishing
AusE as a distinct national variety of English worldwide, an important step in
the construction of a national identity in a postcolonial context. A geolinguistic approach is also concerned with the relationship between the local and
national varieties of a given language. We have compared our results with the
pronunciations listed in the Macquarie Dictionary with the aim of showing
how variation within Australia compares with the representation of AusE in
this dictionary. In future work we will use geolinguistic analysis (that is, the
identification of word, place, and social class effects, as in the present study)
to investigate the relationship between AusE and New Zealand English (the
latter based upon data collected in Auckland, Wellington, and Christchurch)
with the aim of testing scalar effects.
Finally, we will compare the major regional patterns of variability found
in short A (see Table 6) with our recent study of the vocalisation of /l/ in the
same cities (Horvath and Horvath 1997). The following pattern for /l/ was
found (percentage of vocalised /l/): Adelaide 33%; Mount Gambier 33%;
Hobart 26%; Sydney 23%; Melbourne 12%; and Brisbane 7%. In both studies
the two South Australian cities (Adelaide and Mount Gambier) stand out as
the most distinctively different. Most different to Adelaide/Mount Gambier
are Melbourne and Brisbane, with the lowest percent of /l/ vocalisation and a
near categorical short A dialect. Brisbane is the most marginal dialect with
respect to the two linguistic variables studied. Because we argued that /l/ was
a language change in progress, we interpreted Brisbane’s marginal position as
a possible indication that it was in the early phase of this language change.
Between Adelaide/Mount Gambier and Brisbane/Melbourne, linguistically
speaking, are Sydney and Hobart, having intermediate levels of /l/ vocalisation and a more variable pattern with short A. The study of a wider sample of
speech localities in Australia that would include a wider range of variables
will be necessary before we can firmly delimit the geolects of AusE.
Students in a class on sociolinguistic variation collected and coded the data. We are
grateful to Peter O’Carroll for collating the data that is presented here.
Interestingly, romance is also unexpectedly pronounced with [Q] in RP.
We have used (æ >a) to mean that [æ] is listed first in the dictionary and is more common
than [a].
A FileMaker-Pro application designed for the efficient coding of survey data and allowing the construction of datasets that can easily be exported to Goldvarb (Horvath and
Horvath 1995).
We will not be reporting on the results of the Varbrul analysis (Rand and Sankoff 1990) in
this paper; but for those readers interested in the results, the contributions to the probability of short A were as follows (a figure greater than .5 means that short A is promoted and
less than .5 means that short A is inhibited in that factor group): for the Word factor
group: dance .995; advance .892; plant .883; grasp .142; giraffe .056; mask .009. For the
Social Class factor group: working class .567; middle class .420. The Place factor group
showed: Sydney .809; Brisbane .782; Hobart .775; Melbourne .726; Adelaide .074;
Mount Gambier .066. The age and gender factor groups were not significant.
Such comparisons are particularly difficult unless the same set of words is studied, which
is not the case here.
The variation with short A may, of course, be evidence of a sound change that has become
moribund or it may be the result of lexical diffusion. Wells (1982: 233) considers it to be
due to the “ossification of a half-completed sound change”. See Labov (1981) for a
further discussion of the differences between new and vigorous sound changes and those
described as the result of lexical diffusion.
A sociolect (Horvath 1985) is a pattern of linguistic variation that corresponds to social
differentiation within a speech locality, and a geolect is a pattern of linguistic variation
that differentiates a speech locality from others.
Bailey, Guy, Tom Wikle, Jan Tillery and Lore Sand. 1993. “Some patterns of linguistic
diffusion”. Language Variation and Change 5: 359–390.
Bernard, John R. L. 1981. “Australian pronunciation”. In The Macquarie Dictionary.
Sydney: Macquarie Library Pty. Ltd., 18–27.
Borowsky, Toni and Barbara M. Horvath. 1997. “L-vocalization in Australian English”. In
Franz Hinskens, Roeland van Hout and W. Leo Wetzels, eds. Variation, Change and
Phonological Theory, Amsterdam: John Benjamins, 101–123.
Bradley, David. 1989. “Regional dialects in Australian English phonology”. In Peter
Collins and David Blair, eds. Australian English: The Language of a New Society. St
Lucia: University of Queensland Press, 260–270.
Britain, David. 1991. “Dialect and space: a geolinguistic study of speech variables in the
Fens”. Ph.D., University of Essex.
Bryant, Pauline. 1985. “Regional variation in the Australian English lexicon”. Australian
Journal of Linguistics 5: 55–66.
———. 1989. “The south-east lexical usage region of Australian English”. Australian
Journal of Linguistics 9: 85–134.
Chambers, Jack K. and Peter Trudgill. 1980. Dialectology. Cambridge: Cambridge University Press.
Ferguson, Charles A. 1972. “‘Short A’ in Philadelphia English”. In M. Estelle Smith, ed.
Studies in Linguistics in Honor of George L. Trager. The Hague: Mouton, 259–274.
Horvath, Barbara M. 1985. Variation in Australian English: The Sociolects of Sydney.
Cambridge: Cambridge University Press.
Horvath, Jayne and Barbara M. Horvath. 1995. The Language Coder: A File-Maker Pro
Application. Sydney: Black & White, Pty Ltd.
Horvath, Barbara M. and Ronald J. Horvath. 1997. “The geolinguistics of a sound change in
progress: /l/ vocalization in Australia”. In Miriam Meyerhoff, Charles Boberg and
Stephanie Strassel, eds. Working Papers in Linguistics: A Selection of Papers from
NWAVE 25. Philadelphia: University of Pennsylvania, 105–124.
Kretzschmar, William et al. 1994. Handbook of the Linguistic Atlas of the Middle and South
Atlantic States. Chicago: University of Chicago Press.
Kurath, Hans. 1949. A Word Geography of the Eastern United States. Ann Arbor: University of Michigan.
Labov, William. 1966. The Social Stratification of English in New York City. Washington,
D.C.: Center for Applied Linguistics.
Labov, William. 1981. “Resolving the Neogrammarian controversy”. Language 57: 267–
McArthur, Tom. 1998. The English Languages. Cambridge: Cambridge University Press.
Mitchell, Alexander. G. 1958. Spoken English. London: Macmillan.
——— and Arthur Delbridge. 1965a. The Speech of Australian Adolescents. Sydney:
Angus and Robertson.
——— and Arthur Delbridge. 1965b. The Pronunciation of English in Australia. Sydney:
Angus and Robertson.
Orton, Harold et al. 1978. The Linguistic Atlas of England. London: Croom Helm.
Rand, David and David Sankoff. 1990. Goldvarb. Version 2. A Variable Rule Application for
the Macintosh. Montréal: Centre de recherches mathématiques, Université de Montréal.
Trudgill, Peter. 1974. “Linguistic change and diffusion: description and explanation in
sociolinguistic dialect geography”. Language in Society 2: 215–246.
Wells, John C. 1982. Accents of English 1: An Introduction. Cambridge: Cambridge
University Press.
Abbott, I. 143
abbreviations 6, 92, 108
ABC (Australian Broadcasting Commission/Corporation) 274–275, 280, 283,
308, 311
Aboriginal English 3, 10, 134–136, 139,
140, 147, 182–184, 192, 201–222
and education 202
definition of 201
discourse in 216–217
functions of 217–218
grammatical system of 215–216
lexicon of 216
phonology of 217
Aboriginal languages 5, 304
borrowings 152
accent types of AusE 17, 29–31
see also Broad Australian, General
Australian, Cultivated Australian;
Educated Australian, Modified
acoustic analysis 19
acronyms 92, 98, 102, 109
Adnyamathanha 146
affrication 54
age 266, 348
Ahmad, K. 164
Air Traffic Services call-signs 94
airport codes 92
Alexander, D. 202
Al-Hindawe, J. 279, 282
Alyawarr 143
America 265
American English 4–6, 119, 126–128,
130, 155–157, 163, 167–169, 174–
175, 317, 322–327, 333
American influence 4, 5, 113, 117, 153–
155, 157, 168, 259, 324
Andrésen, B. 52, 53
Anglo-Australians 8, 240–252, 255–256
Anglo-English 2
Anglo symbolic separation 255
apodosis 120, 123, 126
Archer, M. 143
Ariss, R. 218
Arthur, J. 134, 202
assimilation 8
attitudes to AusE 271–285
Auer, P. 235, 352
Australasian English 20
Australian Concise Oxford Dictionary
146, 327
Australian Corpus of English (ACE)
116, 118, 119, 122, 127–130, 164,
Australian National Dictionary 134,
141, 146, 158–160, 303, 304, 308–
Australianisms 163, 304, 307
Awakabal 146
Ayers Rock 145
Babbitt, E.H. 52
backshifting 121–123
Bailey, G. 344
Baker, S.J. 49, 54, 163, 274, 329, 336,
Ball, P. 278, 279
Bank of English 164
Barcan, A. 298
Barnhart Dictionary Companion 156
Bauer, L. 20, 21, 40, 41, 164, 330
Beagle, L. 279, 283
Bell, Alexander Melville 52, 53
Bell, Allan 20, 40
belong-construction 189, 190, 198
Bennett, E.M. 143
Bennett, G. 1–2
Berechree, P. 279
Bernard, J.R.L. 7, 9, 17–19, 23, 26–28,
41, 244, 289, 318, 343
Berndt, C. 203–205
Berndt, R. 203–205
Berry, R. 202
bi-gestural articulation 70
Birmingham corpus 164
Black American English 192
Blackman, J. 274
Blair, D. 1, 3, 9, 18, 123, 165, 289, 292,
300, 318
Blake, B. 135
blends 98
Blevins, J. 70
Bloomfield, L. 19
Booij, G. 51
Borowsky, T. 69, 70, 72, 76, 84, 86,
343, 348
Borrie, W.D. 289, 291, 295–297
borrowings 154–157
from Aboriginal languages 133–136,
from AmE 4
Bouma, G. 130
Bourhis, R. 226
Bradley, D. 9, 18, 20, 24, 273, 335, 337,
343, 351, 352
Bradley, M. 273
Braithwaite, R.W. 142
Britain, D. 344
British English 6, 111, 114, 115, 117,
118, 120, 126–129, 154, 160, 163,
164, 167–176, 317, 323, 327–330
British influence on AusE 5
Broad Australian 1, 2, 7, 8, 17, 20, 29,
49, 54, 242, 272, 274–276, 279–282,
292, 297, 300, 318
Browman, C. 85
Brown Corpus 163–176
Bryant, P. 9, 343
Butler, S. 4, 326
Cambridge Australian English Style
Guide 313
Canadian English 128, 334, 343
Cardiff English 53
Cassidy, S. 21, 23
Catford, J. 18
cattle station English 213
centralisation 19
chain shifts 19, 40
Chambers, J.K. 21, 22, 239, 343
change in language 3, 7–8
in vocabulary 154–157
in vowels 19–22, 29, 41
Cherikoff, V. 143
Cheshire, J. 122
Churchill, W. 310
Clark, G. 336
Clark, J.E. 54
cleft constructions 172
clippies 108
Clyne, M. 116, 130, 224, 225, 227, 237,
COBUILD corpus 164
Cobuild English Language Dictionary
Cochrane, G.R. 20, 40, 318
Cockney 2, 49, 300, 320, 335
Coleman, J. 51
Collins, H. 318
Collins, P.C. 113–118, 123, 126, 165,
colloquialisms 259–269, 314
common nouns 7, 90, 94, 108–111
community languages 224
comparative adjective 117
computer use, effect on
vocabulary 331–333
Concise Oxford Dictionary 305
concord 6, 115, 119–120
conditional constructions 121–123
conjunction like 166
consonants 8, 45–67, 69–87
continuous aspect 171
contractions 168–175
convicts 288, 289–91, 292, 294, 296
Cooke, M. 217
Corbett, G. 164
Corbyn, C. 319
Cornelius, S. 323, 334
categories in 165, 167, 168, 172–173
evidence from 4, 6, 163–178
genres in 165–167, 169, 172–173
Coulthard-Clark, C. 147
Cox, F. 7, 10, 17–19, 21–23, 26–28, 29,
37, 38, 41, 237, 244, 245
Cran, W. 163
creoles 10, 181–182, 197–198
creolisation 210, 211
Crystal, D. 117
Cultivated Australian 2, 7, 8, 17, 20, 29,
49, 242, 272, 274–276, 279–282, 300,
cultural cringe 311
“culture shifting” 208
Cunningham, P. 2
Curtain, J. 116, 270
Dabke, R. 6, 90
Dailey-O’Cain, J. 20
“danger of death” question 278
Darambal 141
Daw, B. 144
debuccalisation 47, 48
decreolisation 211
Delbridge, A. 8, 10, 17, 23, 40, 44, 46,
49, 55, 66, 163, 167, 177, 242, 275,
287, 289, 301
Dennis, C.J. 53
depidginisation 211
Dermody, A. 90, 104, 105, 110
Dermody, P. 6, 22
Devine, Frank 273, 327–328, 332, 337
dialects 1, 10–11
dialect geography 348
levelling 1
mixing 319
Dictionary of Australian Colloquialisms
Dictionary of New Zealand English
322–323, 336
Dictionary Research Centre (Macquarie
University) 313
differentiation 243
diminutives 6, 108
diphthongs 18–21, 26–28, 30–37, 38–42
direct speech 174
disputed usage 116–119
Dixon, J. 2
Dixon, R.M.W. 134, 201, 204, 210, 211,
Djingulu 135
Docherty, G. 51, 63
Donaldson, T. 207
Doyle, R. 106, 111
Douglas, W. 202
drag chain, in vowel shift 330
Dutton, T.E. 202, 211–213
Dyirbal 142
Eades, D. 201, 202, 217
Eckert, P. 20, 21, 29, 34, 37, 41, 254
ecology of language 203–204
Educated Australian 318
education 293, 298
Edwards, J. 260, 270
Eisikovits, E. 113, 116, 229, 281
elitism 309–310
Ellis, A.J. 52, 53
Elwell, V. 204, 207
emancipists 291
Enemburu, I. 201
English English 119, 127, 317
English place-names 107
Epistemic have (got) to 126
Estuary English 53
ethnic varieties 223–234
ethnicity 231–234, 236, 243, 253
ethnolects 8, 10–11, 184, 223–227
ethnolinguistic vitality 255
Evans, Z. 17, 40, 245
exclamation marks 174
existential statements 172–173
exonormative standard 260
Fallon, R. 169
Fant, G. 21–22
Farid, M. 65
Fasold, R. 2
Fee, M. 334
femininity 253
Ferguson, C. 341
Fesl, E. 202
Finch, C. 278
Fishman, J. 229
flapping of /t/ 45, 54, 57–58
Flint, E. 202
Follett, W. 117, 118
foreigner talk 211, 227
formal register 168
formants 19, 22–23, 30–32
Foster, D. 313–314
Foulkes, P. 51, 63
Fowler, H.W. 167
Franklyn, J. 53
frication of /t/ 45–48, 54, 58, 62
Fritz, C. 321, 336
Fujimura, O. 70, 71, 73, 76
gender 343, 346, 347, 349, 352
gender differences 254, 348
General American 341
General Australian 17, 20, 22–23, 29,
36, 49, 242, 272, 275, 276, 279–282,
288, 297, 300, 318
genres in corpora 165–167, 169, 172–
geolinguistics 341–355
German 224, 232–234
Gerstaecker, F. 2
Gibbs, G. 202
Gilbert, K. 140
Giles, H. 3, 226
Ginibi, R. 140
glottal stop 47, 52, 53, 62, 63
glottalisation 47, 49–54, 56, 58–60, 62–
Goldlust, J. 224
goldrushes 322
Goldstein, L. 85
Goldvarb analysis 73
Gordon, I.A. 176
Görlach, M. 171
grammar of AusE 113–132
Granger, S. 171
Great Vowel Shift 20, 330
Greek Australian English 228–229
Greek Australians 240
Greeks in Australia 241–255
Green, E. 165
Greenbaum, S. 165
Gunn, J. 318
Guy, G. 55, 330
Hammarström, U.G.E. 318
Hannah, J. 9, 115–117, 120, 126, 128–
Hanson, Pauline 240
Harkins, J. 201, 202, 215, 216
Harrington, J. 17, 18, 22, 23, 29, 30, 40,
Hitchen, M. 202
Hockett, C. 19
Hofland, K. 169
Holmes, J. 20, 40, 65, 165, 171
homogenisation 239, 242, 243, 252
Honey, J. 299, 300
Hornadge, B. 274
Horvath, B. 8, 20, 45, 54, 69, 72, 74, 84
86, 209, 223, 242, 243, 290, 321, 330,
343, 348, 353, 354
Horvath, R. 348, 353
Huddleston, R. 121
Hudson, J. 202
Hudson, N. 116, 117, 119, 123, 125
Humphries, Barry 274
Hundt, M. 164, 168, 171, 175
hypercorrections 118, 119
hypocoristics 7, 89–111
endings 104–107
identity 303–316
immigrants 288, 289, 291, 296–298, 312
imperatives 174
implicational table 350
in-group 225–226, 231
indigenisation 201
informal register 168, 175–176
Ingram, J. 8, 18, 53, 54
interactive writing 174
interlanguage 183, 184, 198
International Corpus of English 165
internet 331–333
interrogatives 174, 191
Irish English 48–51, 105, 106, 109, 127,
317, 318, 320–321
Irish Gaelic 50, 64
Isaacs, J. 143
it-clefting 190
Italians in Australia 241–252
Ito, J. 96
Jespersen, O. 105, 107
Johansen, L. 325
Johansson, S. 169, 175
joking names 95
Jones, Daniel 52
Kala Lagaw Ya 181
Kaldor, S. 202, 207, 210, 214, 215
Kale, J. 202
Kamilaroi 146
Kaye, J. 51
Keesing, N. 90, 274
Kerswill, P. 55, 239
Kiparsky, P. 51
Kjellmer, G. 172
Kniesza, V. 335
Koch, H. 217
Kociumbas, J. 298
koinéisation 10, 214, 295
Koscielecki, M. 202, 205, 207, 213, 214
Kotsinas, U-B. 237
Kretzschmar, W. 344
Kuku Yalanji 141
Kurath, H. 343
Labov, W. 19– 22, 34, 40, 41, 168, 235,
239, 242, 245, 278, 344, 354
Ladefoged, P. 20, 70
Lancaster-Oslo-Bergen (LOB) Corpus
Langton, M. 217
Larrakia 135, 145
laryngealisation 62–63, 65
Lebanese in Australia 244–252
Leech, G. 164, 169
Lefebvre, C. 210, 214
leniting 45
levelling of dialects 48, 113, 209
lexical invariance 190, 198
lexicography 303–316
lexicon 308–310
Liljencrants, J. 20
Lindblom, B. 20, 41
Lindquist, H. 315
literacy 293, 312
of TSE speakers 182–184, 198–199
Lo Bianco, J. 313
London English 52–53, 60
Lowenstamm, J. 51
Luthin, H. 40, 41
Mackenzie Beck, J. 121
MacNeil, R. 163
Macquarie Dictionary 40, 146, 156,
159, 166, 176, 305–310, 326, 336,
345–347, 352
Maddieson, I. 70
Malay 65
Malaysia 260, 264, 267
Malcolm, I. 202, 205, 207, 210, 213–
mandative subjunctive 167–168
Manne, R. 275
manuals of style 114, 116, 121, 123,
125, 127–129
markedness constraints 69, 72–73, 77
Martin, B. 301
masculinity 253
matched-guise technique 279
Matthews, R. 20
Maurer, D. 337
Mayi-Kulan 141
Mayi-Yapi 141
McAndrew, A. 90
McArthur, T. 315, 340
McCarthy, J.J. 72, 96
McCrum, R. 163
McGregor, C. 135
McKenry, R. 202
McMahon, A. 74
McRae, D. 202
media 323
Mees, I. 53
Meredith, Louisa 2
Meriam Mir 145, 181
Mester, A. 96
Meyer, C. 174
migrant groups 239–257
migration effects on AusE 48, 50, 239,
240, 319
Millar, B. 22, 29
Milroy, J. 53
Milroy, L. 55, 254, 321
Mitchell, A.G. 1, 8, 10, 17, 23, 40, 48–
50, 54, 163, 242, 275, 287–302, 311,
318, 320, 321, 335, 342, 343, 352
Mitchell, T.L. 144
Mittins, W.H. 118
mixed code 225
mixing bowl theory 289, 294
mobility 295–296
modal verbs 123, 175
Modified Australian 275, 276
Mohanan, K. 51
Monash Australian English Project 228
monophthongs 18, 20–26, 28, 30–32,
37–42, 185
Moonwomon, B. 40
Moore, B. 98, 327
Morgan, S. 140
Morgan, W. 218
Morris, E. 274, 304
Mossman, S. 2
Mountford, C.P. 143
Muecke, S. 202, 204, 217
Mühlhäusler, P. 90, 202–204, 209–211,
213, 214
multiculturalism 240, 256, 259, 312
multiethnicity 240
multifunctionality in TSE 191, 198
name exchange 196
name taboo 196
National Policy on Languages 312
negation 172
negative contractions 169, 171–172
networks 254
New Englishes 201–202
New South Wales pidgin 211–215
New Zealand English 7, 20, 40, 65, 119,
322–323, 330–331, 333, 343
Newbrook, M. 116–119, 121, 123, 124,
newspapers 334
Ngajan 142
Ngunnawul 145
Nhangka 146
Nhanta 141
Nicholas, S. 290, 293, 294
Nietschmann, B. 193, 197
Nixon, G. 164
Noël-Armfield, G. 53
non-Anglos 240
non-native speakers 259–269
non-restrictive relative clauses 125
non-standard forms 114–115, 122, 125,
183, 184, 186–190, 192, 195, 197–
notional agreement of verbs 164
Nunga English 213
Nyamal 143
Nyungar 141, 146
Oasa, H. 18, 40, 42
Observer’s Paradox 235
ocker 275
Ohala, J. 20, 53, 60
Ohala, M. 53
Oishi, I. 334
Optimality Theory (OT) 69–87
Orsman, H. 336
Orton, H. 343
out-group 225–226
outliers 250–251
Overgaard, G. 168
overgeneralisation 186, 187, 197, 199
Oxford English Dictionary 305
Oxley, D. 293
Ozcorp 166, 176
Pacific Pidgin English 181
Palethorpe, S. 37, 237
Pama-Nyungan 181
pan-ethnic variety 234
Panyjima 146
Partridge, E. 118
passive construction 171, 175
passivisation 190
Pawley, A. 113, 132
personal names 91, 94, 105, 106, 109,
Peters, P. 116–119, 123, 125, 127, 164,
165, 167, 172, 175, 333, 334
Peterson, A. 273
pidginisation 210
Pierrehumbert, J. 62
Pitjantjatjara 139, 146
place-names 89–111
plosive articulation 45, 46, 49, 51, 54,
57–60, 62, 63
“political correctness” 326, 331–333
Porter, H. 272, 281
Poynton, C. 109, 110, 278
prepositional usage in TSE 189
present perfect aspect 131
Prince, A.S. 71, 72, 96
pronouns and contractions 172
pronunciation 287, 288, 290, 300, 325
proper names (nouns) 90, 109–111, 196
protasis 121–122, 126
proto-Broad accent 318, 319
proto-Cultivated accent 300, 319
pub names 93, 107
Pullum, G. 116
punctuation, sentence final 174
puns 91, 109
purpose clauses 191
quantitative statements 173
quasi-existential statements 172
question marks 174
Quirk, R. 163, 167, 171
raising of vowels 19, 20, 24
Rampton, B. 237
Ramsay, J. 103
Ramson, W.S. 134, 154, 155, 272, 303,
304, 322, 336
Random House Dictionary 305
Readdy, C. 202
Recasens, D. 75, 83–86
Received Pronunciation 288, 299, 341
reduction of /t/ 45–67
regional variation 18, 93–95, 108, 231–
regularisation 184
Reid, W. 173
relative clauses 125–126
religion as a variable 231–234
Renton, N. 116–119, 123, 125
restrictive relative clauses 125
Reynolds, H. 203
rhetorical markers in narrative 196
Roberts, A. 143
Robinson, H. 90, 107, 109
Robinson, P. 293
Robson, L. 289, 290
Rogers, I. 334
Romaine, S. 209, 211
Rose, D. 202, 210, 213
Rosewarne, D. 53
Rubach, J. 51
Rumsey, A. 203, 204
Samah, A. 270
Sandefur, J. 201
Sansom, B. 202, 217
Schilling-Estes, N. 245
Schornikow, T. 278
Scottish English 49, 129, 317
Seal, G. 5
sex and language 253
Sharpe, M. 185, 202
Shaw, A. 289, 290, 293, 294
Shaw, G.B. 53
Shergold, P. 290, 293, 294
Shnukal, A. 113, 181
shortenings 108
Siegel, J. 202, 210, 213, 214
simplification 184, 185, 198, 199
Sirianos, P. 278
Slattery, S. 202
Sleep, B. 202
Smolensky, P. 71, 72
social class 348
social networks 93, 94, 95, 108
socio-stylistic variation 18
sociolectal continuum 49
sociolinguistics 343, 351
solidarity 279–281
Sonorant Ranking Hierarchy 99
sonority 103
sound shifts 20
sound spectrography 19, 244
South African English 128
speaker attitudes 46, 60
spelling, ambiguity in 90
spirantisation of /t/ 47
Spitale, J. 279
Sproat, R. 71–73, 76
standard forms 115–130, 167, 175
status 279–281
St Clair, N. 3
Steiner, R. 19
stereotypes/stereotyping 271, 272, 275
Strahan, R. 142
Strang, B. 171
Strong, B.W. 143
style and usage 163
Style Council 313
style guides/manuals 114, 116, 121, 123,
125, 127–129
style shifting 255
styles (formal/informal) 114, 115, 118,
119, 121–122
Subjective Reaction Test 272, 278–281
subjunctive 167–168
suffixation 95–96
superlative 123–124
surnames, hypocoristic forms with /o/
Sussex, R. 113, 323, 333, 334
Sweet, H. 19, 52, 53
Sydney English, short A distribution
syllable 50, 57, 64
coda 72–78, 81–83
harmony 70, 73, 79, 83
nucleus 71, 79–81
onset 71, 79–81
syntactic norms 114–115
syntax 113–132
of TSE 189–192
tapping of /t/ 45–48, 54, 57–58, 61–62
Taylor, B. 4–5, 90, 105, 106, 111, 113,
116, 321, 323, 324, 325, 333, 335,
television 327–330
templates for hypocoristics 96–104, 107
Thaiday, W. 200
Thomas, M. 134
Thursday Island 181, 183, 184, 186,
189, 190, 193, 195
Tiwi 143
Tollfree, L. 50, 52–55, 59, 60, 227
topographic descriptors 92, 107
toponyms 89–111
Torres Strait Creole (TSC) 181, 182,
187–192, 194–196, 198, 199
Torres Strait English (TSE) 182–200
Tottie, G. 171
transfer 184, 190
transference 185, 187, 189, 191
transportation 291
Troy, J. 202, 205, 206, 209, 211–213,
Trudgill, P. 9, 20, 21, 114–117, 120,
126, 128, 129, 214, 289, 291, 294,
319, 330, 343, 344
truncation 95, 96, 102–104, 106–108,
Tsolidis, G. 255
Turner, G.W. 49, 54, 318, 322
Turney, C. 299
Tyneside English 51, 63, 64
Uluru 145
urbanisation 294
variability in TSE 184
variation in consonants 45, 47, 60, 63
verb chaining in TSE 189, 191, 198
Vergnaud, J.R. 51
Vietnamese in Australia 244–252
of /t/ 47–48, 54, 55
of /l/ 69–87, 348
voice quality of TSE 185, 198
Vonwiller, J. 330
vowels 17–42
backness 75, 83, 84
centralisation 20
chainshifts 19–20, 41
fronting/frontness 18–19, 20, 22, 24–
25, 31, 33–34, 37–38, 40–41
height 18–19, 83, 84
length 74, 83, 84
raising 20
spectra 19
Walker, R. 202
Walsh, M. 204, 207, 217
Walsh Dickey, L. 70
Warren, J. 234
Warrgamay 141
Wathaurong 145
Watjari 146
Watson, C. 21, 40
Watson, K. 118
Webby, E. 299
Webster’s Dictionary 305, 306
Wells, J. 8, 21, 40, 50, 54, 62, 64, 245,
341, 342, 344, 345, 352, 354
West Yorkshire English 65
Westergren-Axelsson, M. 173
Western Desert language group 141, 146
Whitaker, A-M. 335
White Australia policy 240
Whitlam, E.G. 3
Wierzbicka, A. 90, 94, 109, 237
Wik-Mungkan 133
Wilkes, G. 274, 314
Williams, A. 239
Wilson, G. 202
Winter, J. 116
Wiradhuri 145
“Wogspeak” 234
Wolfe, P. 20
word endings 90, 92, 96–99, 104–107
World War I 151–153
World War II 152–153, 323, 329, 337
Wright, S. 55
Wunambul 143
Yaeger, M. 19
Yagara 142
Yallop, C.L. 275
Yiddish 224
Yiddish Australian English 229–231
Yindjibarndi 142
Yio, S.K. 117
Yolngu 146
Young, W. 172
Yumpla Tok 181
Zeccola, L. 281
Zeller, C. 20
titles have been published thus far:
G1. LANHAM, L.W. & C.A. MaCDONALD: The Standard in South African English and
its Social History. Heidelberg (Groos), 1979.
G2. DAY, R.R (ed.): Issues in English Creoles: Papers from the 1975 Hawaii Conference.
Heidelberg (Groos), 1980.
G3. VIERECK, Wolfgang, Edgar SCHNEIDER & Manfred GÖRLACH (comps): A Bibliography of Writings on Varieties of English, 1965-1983. 1984.
G4. VIERECK, Wolfgang (ed.): Focus on: England and Wales. 1984.
G5. GÖRLACH, Manfred (ed.): Focus on: Scotland. 1985.
G6. PETYT, K.M.: ‘Dialect’ and ‘Accent’ in Industrial West Yorkshire. 1985.
G7. PENFIELD, Joyce & Jack ORNSTEIN-GALICIA: Chicano English. 1985.
G8. GÖRLACH, Manfred and John A. HOLM (eds): Focus on the Caribbean. 1986.
G9. GÖRLACH, Manfred: Englishes. Studies in varieties of English 1984-1988. 1991.
G10. FISCHER, Andreas and Daniel AMMAN: An Index to Dialect Maps of Great Britain.
G11. CLARKE, Sandra (ed.): Focus on Canada. 1993.
G12. GLAUSER, Beat, Edgar W. SCHNEIDER and Manfred GÖRLACH: A New Bibliography of Writings on Varieties of English, 1984-1992/93. 1993.
G13. GÖRLACH, Manfred: More Englishes: New studies in varieties of English 1988-1994.
G14. McCLURE, J. Derrick: Scots and its Literature. 1995.
G15. DE KLERK, Vivian (ed.): Focus on South Africa. 1996.
G16. SCHNEIDER, Edgar W. (ed.): Focus on the USA. 1996.
G17. PETER PATRICK: Linguistic Variation in Urban Jamaican Creole. A sociolinguistic
study of Kingston, Jamaica. 1999.
G18. SCHNEIDER, Edgar W. (ed.): Englishes around the World, Volume 1. General studies,
British Isles, North America. Studies in honour of Manfred Görlach. 1997.
G19. SCHNEIDER, Edgar W. (ed.): Englishes around the World, Volume 2. Carribbean,
Africa, Asia, Australasia. Studies in honour of Manfred Görlach. 1997.
G20. MACAULAY, Ronald K.S.: Standards and Variation in Urban Speech: Examples
from Lowland Scots. 1997.
G21. KALLEN, Jeffrey L. (ed.): Focus on Ireland. 1997.
G22. GÖRLACH, Manfred: Even More Englishes. 1998.
G23. HUNDT, Marianne: New Zealand English Grammar - Fact or Fiction? A corpus-based
study in morphosyntactic variation. 1998.
G24. HUBER, Magnus: Ghanaian Pidgin English in its West African Context. A sociohistorical and structural analysis. 1999.
G25. BELL, Allan and Koenraad KUIPER (eds.): New Zealand English. n.y.p.
G26. BLAIR, David and Peter COLLINS (eds.): English in Australia. 2001.
G27. LANEHART, Sonja L. (ed.): Sociocultural and Historical Contexts of African
American English. n.y.p.
T1. TODD, Loreto: Cameroon. Heidelberg (Groos), 1982. Spoken examples on tape (ca. 56
22/05/01, 10:02
T2. HOLM, John: Central American English. Heidelberg (Groos), 1982. Spoken examples
on tape (ca. 92 min.)
T3. MACAFEE, Caroline: Glasgow. 1983. Spoken examples on tape (ca. 60 min.)
T4. PLATT, John, Heidi WEBER & Mian Lian HO: Singapore and Malaysia. 1983.
T5. WAKELIN, Martyn F.: The Southwest of England. 1986. Spoken examples on tape (ca.
60 min.)
T6. WINER, Lise: Trinidad and Tobago. 1993. Spoken examples on tape.
T7. MEHROTRA, Raja Ram: Indian English. Texts and Interpretation. 1998.
22/05/01, 10:02
Без категории
Размер файла
2 057 Кб
publishing, series, general, around, benjamin, varieties, world, 2001, collins, English, john, blair, david, peter, australia
Пожаловаться на содержимое документа